Review of quantitative multiscale imaging of breast cancer

Michael A. Pinkert; Lonie R. Salkowski; Patricia J. Keely; Timothy J. Hall; Walter F. Block; Kevin W. Eliceiri

doi:10.1117/1.JMI.5.1.010901

22 January 2018 Review of quantitative multiscale imaging of breast cancer

Michael A. Pinkert, Lonie R. Salkowski, Patricia J. Keely, Timothy J. Hall, Walter F. Block, Kevin W. Eliceiri

Author Affiliations +

Journal of Medical Imaging, Vol. 5, Issue 1, 010901 (January 2018). https://doi.org/10.1117/1.JMI.5.1.010901

Abstract

Breast cancer is the most common cancer among women worldwide and ranks second in terms of overall cancer deaths. One of the difficulties associated with treating breast cancer is that it is a heterogeneous disease with variations in benign and pathologic tissue composition, which contributes to disease development, progression, and treatment response. Many of these phenotypes are uncharacterized and their presence is difficult to detect, in part due to the sparsity of methods to correlate information between the cellular microscale and the whole-breast macroscale. Quantitative multiscale imaging of the breast is an emerging field concerned with the development of imaging technology that can characterize anatomic, functional, and molecular information across different resolutions and fields of view. It involves a diverse collection of imaging modalities, which touch large sections of the breast imaging research community. Prospective studies have shown promising results, but there are several challenges, ranging from basic physics and engineering to data processing and quantification, that must be met to bring the field to maturity. This paper presents some of the challenges that investigators face, reviews currently used multiscale imaging methods for preclinical imaging, and discusses the potential of these methods for clinical breast imaging.

1. Introduction

Breast cancer is the most common cancer among women and is the second leading cause of cancer deaths worldwide.¹^,² Treatment of breast cancer is difficult because breast cancer encompasses many genotypes and phenotypes that affect risk, diagnosis, prognosis, and treatment response,³^–⁵ and these different cancer types can be hard to quantify with current medical imaging methods.⁴^–⁶ Quantifying breast cancer can ultimately assist researchers in answering difficult questions on why such variation exists in patients in the prevalence, aggressiveness, treatment, and outcomes of primary and metastatic disease.⁷^,⁸ This problem is relevant throughout all stages of the disease, as a patient’s tumor can evolve into new variants through environmental pressure and epigenetics, creating a tumor with multiple regions that respond differently to therapy.⁵^,⁹^,¹⁰ Addressing this problem requires new imaging methods to measure the cellular, structural, and morphological differences expressed by different types of cancer. Researchers then need to incorporate these measurements into new quantitative cancer models, for use in classifying different cancer types. As such, improved imaging methods for quantifying breast tissue are desirable from a research and clinical perspective, offering improved understanding of breast cancer and patient care.⁵^,¹¹^–¹³

However, there is a major barrier to quantitative characterization of breast cancer; biomedical imaging has an inverse relation between the volume any imaging modality can cover (field of view and penetration depth) and the size of details it can visualize (spatial resolution).¹⁴^–¹⁶ Due to this, clinical imaging modalities are often restricted to a single range for the combined resolution, field of view, and penetration depth (the spatial scale). In practical terms, this means that a given imaging modality usually acquires information on either the cellular scale, tissue composition scale, or organ/animal level. In several fields, including neuroscience and oncology, these corresponding spatial scales have been referred to as the microscale, mesoscale, and macroscale, respectively (Table 1).¹⁷^–²¹ These scales and definitions are still evolving, and by no means absolute, but can be useful groupings for categorizing imaging technology.

In the context of a problem such as breast cancer imaging, each imaging scale can yield different and useful insights into the disease process.

Microscale imaging reveals the cellular composition of the tumor, its extracellular matrix, and benign tissue surrounding it; all of which affect disease risk, development, progression, and metastasis.¹⁰^,¹¹^,²²^–²⁵ However, microscale imaging requires invasive procedures and does not characterize the entire tumor.²⁴^,²⁶ Mesoscale imaging can provide real-time information on cancer extent during surgery, but mesoscale imaging in general is in development at the preclinical stage and is not widespread.¹⁹^,²⁷ Macroscale imaging is predominant clinically and can obtain metrics over an entire tumor or organ. These macroscale metrics are used by prospective computer aided detection (CADe) and computer aided diagnosis (CADx) systems that would assist physicians in detecting and characterizing cancer.⁷ Similarly, metrics based on breast density and patterns of the fibroglandular tissue are important risk factors for developing breast cancer,²⁸^–³⁰ play a role in diagnosis of cancer,³¹^,³² and can serve as prognostic biomarkers for certain treatments.³³^–³⁷ Despite this, macroscale imaging cannot resolve small variations and so fails to detect differences in many cancer phenotypes. Imaging at multiple scales can overcome some of these limitations and can offer a more complete understanding of breast cancer biology and development.

Table 1

Three imaging scales for multiscale breast imaging: the cellular microscale, the tissue mesoscale, and the organ macroscale. These definitions are based on usage in prior literature.17–21 The thresholds for defining these scales are unique to this study and meant to be illustrative and not restrictive.

Scale	Resolution (μm)	Penetration depth (mm)	Field of view (cm2)
Micro	$< 2$	$< 1$	$< 1 \times 1$
Meso	2 to 100	1 to 100	$1 \times 1 - 10 \times 10$
Macro	$> 100$	$> 100$	$> 10 \times 10$

Quantitative multiscale imaging of the breast (QMIB) is an emerging field of research which encompasses systems and techniques capable of imaging the breast at multiple spatial scales. A large number of modalities exist to image the breast, but until recently there have been few modalities or techniques to make quantitative comparisons between scales.¹⁷^,¹⁹^,³⁸^–⁴³ The field’s relative youth means that the quantitative links among the three scales are not well defined, leaving large gaps in our knowledge.⁴⁴^,⁴⁵ For example, breast density is measured by clinical macroscale imaging, but the biological reasons it contributes to breast cancer risk and progression are still unclear. Fully characterizing breast density will require quantitative links to the micro and mesoscale, where the biological processes take place.¹³^,⁴⁵ However, many useful tools to do so are currently limited to research modalities.²²^,⁴⁶^–⁴⁸ QMIB seeks to address these problems, but QMIB is in its relative infancy. These factors create a fertile ground for the development and application of multiscale imaging research in the near future, making it essential to forward the dialogue on the technology, methods, and overall direction of QMIB.

The authors hope that this review can serve as a useful introduction to this area of research and will spark further discussion on multiscale imaging of the breast. This review shows that QMIB is a growing research topic and that the technologies involved make it relevant to many areas of breast cancer research. In addition, while the primary focus of the review is on breast cancer, many of the topics are relevant to multiscale imaging of other diseases and organs. The review starts with a discussion on quantitative imaging and the challenges of quantification in multiscale imaging. We then move into preclinical imaging modalities used in QMIB research. We mention, but do not extensively cover, clinical imaging, as there are many excellent reviews on clinical imaging of the breast.¹²^,⁴⁶^,⁴⁹^–⁵¹ Finally, the review describes issues in multiscale data analysis and image processing. We then discuss the current state of the multiscale breast imaging research and possible future directions.

2. Quantification and Quantitative Multiscale Imaging of the Breast

Quantitative imaging is “the extraction of quantifiable features from medical images for the assessment of normal or the severity, degree of change, status of a disease, injury, or chronic condition relative to normal.”⁵²^,⁵³ Based on this definition, quantitative imaging can be divided into two categories. The first is the quantitative analysis of the data in an image. For example, quantitative analysis of a standard mammogram can give a value known as the percent mammographic density (PMD). PMD is a measure related to the extent of fibroglandular tissue and is correlated with breast cancer risk.⁵⁴ However, the standard mammography PMD can only approximate the actual volume and proportion of fibroglandular tissue in the breast.⁵⁵ The second method of quantitative imaging is making quantitative measurements of biology. Volumetric breast density, obtained using a quantitative three-dimensional (3-D) modality or through supplemental mammographic techniques, is a direct measure of the fibroglandular tissue and so can be quantitative by both definitions.⁵⁶

There are many challenges associated with quantitative imaging, and addressing these challenges is a major issue in medical imaging. There are detailed reviews of this subject published by the quantitative imaging biomarkers alliance.⁸^,⁵²^,⁵³ In brief, quantitative data acquisition is difficult because a measurement must be based on a physical value, can be affected by many sources, and must be considered statistically. Thus, a measurand should be traceable to a reference value, be repeatable, reproducible, have known components of estimate variance, and should have known estimate bias. The reference, typically obtained with a digital or physical object with known properties (a phantom) connects the measurement to a physical value. A repeatable measurement is one that yields the same result under the same conditions. A reproducible measurement is one that can be acquired by a different observer using different equipment, yet still achieve a similar result. Yet, no measurement is perfectly repeatable and reproducible. The difficulty of these challenges varies with each imaging modality and should be recognized for every quantitative study.⁵³

QMIB faces all the normal challenges associated with quantitative imaging but also introduces other difficulties due to comparisons across spatial scales. QMIB is frequently multimodal, utilizing multiple imaging modalities. Multimodal QMIB faces all the inherent difficulties specific for each modality in addition to their integration into a combined imaging framework across spatial scales.

2.1.

Image Acquisition

The multimodal nature of QMIB further complicates image acquisition due to the technical and procedural requirements for all modes. Among other variations, the images can be taken at separate time points, under tissue deformations that must be corrected for, may use different contrast agents, or are in vivo in one mode and ex vivo in another. The conditions may not be held constant from one session to the next and human error or processing artifacts can introduce unknown changes to the setup. The imaging time also becomes a large concern for in vivo applications. The longest scan time of all modalities limits the time resolution of studies. Encouragingly, QMIB has made much recent progress due to developing quantitative imaging technology and methods that can address these challenges.

2.2.

Data Analysis

The fundamental disparity of spatial scale in QMIB complicates data analysis. QMIB can require orders of magnitude in higher processing time than single-scale imaging due to large datasets and a need for multivariate analysis. This imposes constraints on real-time imaging and currently makes many QMIB methods impractical for widespread use. For multimodal QMIB, a single voxel in a macroscale image can represent several whole microscale images. This causes partial volume artifacts and makes it difficult to delineate the boundary on the microscale image that corresponds to the macroscale voxel, contributing uncertainty further down the data analysis pipeline. Additionally, in multimodal QMIB the modalities may not have the same biophysical contrast mechanism, e.g., tissue acoustic scattering for acoustic imaging verses molecular composition for optical imaging. This makes multimodal QMIB well suited to quantitative studies where it can measure different components of tissue models and how they interact, but characterizing the ground truth of interactions between those sources of contrast is a research area in and of itself.⁵³

3. Quantitative Multiscale Imaging of the Breast Modalities

This review focuses on preclinical imaging modalities (Table 2), as preclinical modalities drive QMIB research. Multiscale imaging usually combines multiple imaging modalities, with each modality operating over a single spatial scale (Fig. 1). Each scale contains preclinical breast imaging modalities; however, the major clinical modalities are at the macroscale and need to be combined with a preclinical modality for multiscale imaging. Thus, a discussion of preclinical modalities covers the instances where clinical modalities are used for QMIB (Table 3). In addition, many clinical modalities are mentioned in sections for related preclinical modalities. Readers interested in more detail on these clinical modalities may reference several other reviews dedicated to clinical breast imaging.¹²^,⁴⁶^,⁴⁹^–⁵¹

Table 2

Preclinical multiscale breast imaging modalities (order listed as presented in this review). A practical assessment of imaging modalities for multiscale imaging of the breast. The characteristic measured column describes the information acquired from the tissue. The form factor describes the imaging equipment; the breast or sample is placed inside a cylindrical bore, examined using an external probe, or in the case of tissue placed on a microscope stage. The prospective clinical use describes proposed uses for the technology in patient care, based on current literature.

Imaging modality	Characteristic measured	Form factor	Advantages	Disadvantages	Prospective clinical use
Standard microcomputed tomography ( $μ CT$ )	Density	Cylindrical bore	Mature technology, inexpensive, developing dedicated breast systems	Ionizing radiation, macroscale resolution clinically, geometric artifacts, electronic noise artifacts	Rapid ex vivo tumor margin detection,⁵⁷ tumor staging,⁵⁸ biopsy analysis⁵⁹
Spectral and photon counting (SPC)- $μ CT$	Density	Cylindrical bore	Whole breast FOV, no geometric artifacts, no electronic noise artifacts, developing dedicated breast systems	Ionizing radiation, slow imaging speed, quantum noise	Diagnostic screening⁶⁰
PhC- $μ CT$ , synchrotron source	Refractive index	Cylindrical bore	Whole breast FOV, developing dedicated breast systems	Ionizing radiation, high expense, and limited availability	Diagnostic screening⁶¹
PC- $μ CT$ , x-ray tube source	Refractive index	Cylindrical bore	Whole breast FOV, inexpensive x-ray source, developing dedicated breast systems	Ionizing radiation	Diagnostic screening,⁶² breast density quantification⁶³
HF-US	Mechanical properties	External probe	Noninvasive, inexpensive, commercial preclinical systems	Can be subject to operator artifacts and sensitive to instrumentation differences	Computer aided detection⁶⁴ and classification⁶⁵ Image guided biopsy,⁶⁶ treatment response imaging⁶⁷
MRM	Molecular environment of hydrogen and other resonant elements	Cylindrical bore	Noninvasive, multicontrast	Long imaging time, high expense, preclinical only	Ex vivo IMA⁶⁸
3D-QHP	Various, based on the stain used	Microscope stage	Multicontrast, qualitative HP is the gold standard	Ex vivo only, slide artifacts, destructive to tissue, long processing time	Computer aided detection or prognosis⁴⁶
LSM	Various; modality dependent	Microscope stage or external probe	Noninvasive, multicontrast	Preclinical only, slow imaging time, submillimeter penetration depth	N/A
WFM	Various; modality dependent	Microscope stage or external probe	Rapid imaging speed	Millimeter penetration depth	IMA⁶⁹
OCT	Refractive index, optical scattering properties, mechanical properties	External probe	Mature technology, inexpensive, noninvasive, rapid imaging, endoscopy and biopsy needle compatible probes	Millimeter penetration depth	IMA,⁷⁰ image guided biopsy⁷⁰
PAT	Fluorophore concentration, optical scattering parameters	External probe	Noninvasive, multicontrast, intrinsically multiscale, commercial preclinical systems	Requires separate probes to image at multiple scales, significant noise	Treatment response imaging⁷¹
DOT	Fluorophore concentration, optical scattering parameters	External probe or cylinder bore	Noninvasive, multicontrast	Very low resolution, no commercial systems, variety of implementations	Supplemental screening,⁷² treatment response imaging,⁷³ breast density assessment⁷⁴
FMT	Fluorophore concentration, optical scattering parameters	Cylindrical bore	Noninvasive, multicontrast, commercial preclinical systems	Quantification artifacts, preclinical only	N/A
DLIT	Cellular luciferase production	Cylindrical bore	Noninvasive, high specificity, commercial preclinical systems	Quantification artifacts, requires transgenic mice or pathologies, preclinical only	N/A

Fig. 1

Multiscale imaging uses multiple imaging modalities to operate across two or more spatial scales. This typically requires preclinical modalities, as most clinical modalities are at the macroscale. The breast imaging modalities are colored by their predominant use in the literature. Green modalities are clinical, and blue are preclinical. The limits for each modality were determined through breast cancer imaging literature and do not reflect performance in other applications.⁶⁶^,⁷¹^,⁷⁵^–⁸²

Table 3

Resolution and multimodal combinations in QMIB (order listed as presented in this review). The resolutions listed in this table are for demonstrated breast imaging applications and do not necessarily reflect the capability of the modality in other applications. All modalities have been combined and validated with qualitative histopathology, and many with 2-D quantitative histopathology, so these combinations were not included in the table.

Imaging modality	Resolution (μm) (small animal/whole breast)	Microscale FOV	Mesoscale FOV	Macroscale FOV
Standard microcomputed tomography ( $μ CT$ )	40⁸³/N/A	N/A	MRM,⁸⁴ FMT,⁸⁵ OCT, ⁷⁰	$μ$ PET,⁸⁶ $μ$ SPECT,⁸⁷ MRI,¹⁷ US, radiography, PAT⁸⁸
SPC- $μ CT$	100⁸⁹^,c/100b^,⁶⁰	N/A	N/A	N/A
PhC- $μ CT$ , synchrotron source	3.25c^,⁹⁰/30b^,⁷⁸	N/A	N/A	N/A
PhC- $μ CT$ , x-ray tube source	100⁹¹/80b^,⁶²	N/A	FMT⁹¹	N/A
HF-US	5⁶⁶/N/A	Transmission electron microscopy ⁹²	PAT⁷¹	$μ PET$ ,⁹³ US⁶⁶
MRM	75⁸²/N/A	FM⁸²	$μ CT$ ,⁸⁴ PAT	MRI⁹⁴
3D-QHP	$< 1$ a/N/A	N/A	WFM⁴⁰^,⁹⁵	MRI⁴⁰
LSM	$< 1$ a/N/A	N/A	WFM⁹⁶	Radiography⁹⁷
WFM	Various mesoscale/N/A	3D-QHP⁴⁰^,⁹⁵	MRM⁴⁰	MRI⁴⁰
OCT	12⁴³/N/A	Optical coherence microscopy⁴³	WFM⁹⁸	US⁹⁹
PAT	45⁷¹/1900⁷⁷	FM⁷¹	HF-US⁷¹	MRI,¹⁰⁰ PET,¹⁰⁰ DOT⁷³
DOT	N/A/2000⁷⁹	N/A	N/A	PAT,⁷³ MRI,¹⁰¹ x-ray tomosynthesis,¹⁰² US⁷²
FMT	20⁸⁰^,¹⁰³/N/A	N/A	$μ CT$ ,⁸⁵ DLIT⁸⁰	MRI,¹⁰⁴ $μ$ SPECT¹⁰⁵
DLIT	20⁸⁰^,¹⁰³/N/A	N/A	$μ CT$ ,¹⁰⁶ FMT⁸⁰	MRI,¹⁰⁶ $μ SPECT$ ⁸⁰

^aStandard implementations of microscale optical modalities are limited by the optical diffraction limit, which is dependent on the wavelength of light used and the numerical aperture of the objective.

^bAdvanced μCT has been demonstrated on mastectomy samples, but not non-invasively with patients.

^cNo breast specific applications were found and so this resolution is from imaging other organs.

Most current QMIB research features mesoscale imaging modalities (Table 3).¹⁸^,²⁷ In the near term, studies use QMIB to validate mesoscale imaging for clinical use. For example, mesoscale imaging can perform intraoperative margin assessment (IMA; the imaging of tumor boundaries during surgery). IMA can prevent the need for a second surgery, which occurs in $\sim 25 %$ of patients operated for a breast malignancy, and will reduce healthcare costs.¹⁰⁷^–¹¹⁰ In the long term, mesoscale imaging makes multiscale coregistration, the spatial mapping of one image to another, more practical. Currently, it is difficult to correlate data based on location between the microscale and the macroscale. For example, positioning a biopsy or imaging probe within a lesion often requires multiple sampling attempts.¹¹¹ During the sampling process, the breast tissue can be distorted by compression or rolling of tissues to obtain access to a lesion. In addition, the orientation of the biopsied sample to the remaining macroscale tissue is not preserved. These issues can be addressed using mesoscale imaging, which is easier to register to and which can act as an intermediary between the microscale and macroscale. This can allow studies to characterize how biological characteristics express at different scales by building multiresolution maps of tissues. Breast cancer expresses many phenotypes at multiple scales that affect patient treatment, and so such characterization could lead to valuable tools and insights.³^,⁴^,¹¹² However, accomplishing these multiresolution maps will require new data analysis methods. For example, there needs to be new methods to accurately register a sequence of images with potential deformations. These multiresolution maps will also depend on the imaging modalities involved, their technical hurdles, and potential applications.

The following sections of this review cover the current status and future perspectives for QMIB imaging modalities. It gives an overview of their biological basis and describes what quantification means to each modality. It covers how the modalities are currently represented in the peer-reviewed literature (Table 2) and how they may be used in the future. In addition, it highlights the many combinations of modalities, including several promising combinations that could bring QMIB into the clinic (Table 3).²⁷^,⁶⁰^,⁷²^,⁹⁹^,¹¹³

3.1.

High-Resolution Variants of Clinical Modalities

Several breast imaging modalities have high-resolution variants that are used in QMIB. These variants are currently preclinical but follow the same principles as their clinical counterparts. This section covers variants of computed tomography (CT), ultrasound (US), and magnetic resonance imaging (MRI).

3.1.1.

Microcomputed tomography

CT imaging utilizes x-rays passing through tissue, obtaining 3-D anatomical information by imparting radiation dose. CT is still developing for clinical breast imaging, but its two-dimensional (2-D) counterpart, mammography, is the most common breast cancer screening modality.¹¹⁴ Standard CT systems produce a spectrum of x-ray energies, and then measure x-ray attenuation through tissue. This obtains semiquantitative 3-D maps of the tissue attenuation coefficient (radiodensity). It is semiquantitative because using a spectrum of x-ray energies results in measurement that varies by depth. The depth variance effectively adds noise to the measurements, making quantification difficult in low-contrast situations. However, quantification is still possible in high contrast situations, e.g., extracting tumor morphology.⁵⁷ This depth limitation can also be overcome by systems that calculate attenuation based on x-ray energy, also known as spectral CT.¹¹⁵

Most QMIB applications of CT occur at the mesoscale. Systems capable of performing mesoscale CT are labeled micro-CT ( $μ CT$ ). $μ CT$ is a well-established preclinical imaging modality with several commercially available systems.¹¹⁵ By comparison, $μ CT$ has not yet reached breast imaging clinically due to technological and radiation dose limitations.¹¹⁵ Several groups are addressing these issues with new systems that can perform whole-breast $μ CT$ ;⁵⁰^,⁶⁰^,⁶² however, they impart radiation dose $2 \times$ to $3 \times$ that produced from clinical mammography or digital breast tomosynthesis systems.⁶⁰^,¹¹⁶ Thus, preclinical and clinical $μ CT$ may both prove valuable tools for future QMIB studies.

Preclinical $μ CT$ has already been paired with many other imaging modalities for QMIB over a wide range of biomedical applications (Table 3). Some examples include characterizing the biodynamics of molecular imaging agents,⁸³^,⁸⁶^,⁸⁷^,¹¹⁷^,¹¹⁸ the biological effects of therapeutic interventions,⁸⁴^,¹¹⁹^–¹²² rapid ex vivo IMA on resected tumors or tumor morphology analysis,⁵⁷^–⁵⁹^,¹²³^,¹²⁴ and the study of vasculature and angiogenesis.¹⁷^,⁸⁴^,¹²⁰^,¹²²^,¹²⁵ There is still much room to expand the preclinical applications of this technology. An excellent review of $μ CT$ in general showcases many possibilities for future QMIB research.¹²⁶ Two notable opportunities include tissue studies with multiscale nano-CT systems, which have submicron resolutions on par with those obtained through microscopy,¹²⁷ and the use of contrast agents for staining antigens, providing substitutions for some immunohistochemistry (IHC) stains in vivo.¹²⁷ Developments in this area of $μ CT$ will add valuable tools to a researcher’s ability to study breast cancer, especially as they can be combined with the technology to be discussed in the following sections.

$μ CT$ is developing clinical relevance for in vivo imaging. Systems using traditional CT technology cannot feasibly reach mesoscale resolution in the clinic, but dedicated breast CT systems based on spectral and photon counting CT (SPC- $μ CT$ ) or phase contrast $μ CT$ (PhC- $μ CT$ ) may make clinical QMIB with $μ CT$ a possibility in the near future.⁵⁰^,¹¹⁴ Both are prospects for clinical QMIB on a single system, as they can obtain mesoscale resolution over the whole breast. This is unique among all mesoscale modalities in this review, combining broad utility with improved ease of use over most multimodal setups while also being familiar in concept to physicians.

SPC- $μ CT$ removes the depth dependence of standard CT, making radiodensity a quantitative measurement. In addition, SPC- $μ CT$ minimizes geometric and electronic noise, improving contrast and resolution.⁶⁰^,¹¹⁵ Commercial preclinical systems using this technology have been released in the last few years, but clinical systems are somewhat behind due to several issues from upscaling the geometry.¹¹⁵ However, this difficulty is being overcome. For example, Kalender et al.⁶⁰ recently published a functioning whole-breast prototype that achieved a resolution of $100 μ m$ at clinically compatible radiation doses. This system obtained 3-D voxels with higher contrast and resolution than the 2-D clinical standards of digital mammography and breast tomosynthesis. Kalender tested this system on lumpectomy specimens to find small calcium deposits (microcalcifications), the morphology and distribution of which may signify cancerous or precancerous cells. This multiscale system detected more calcifications than digital mammography and breast tomosynthesis and was better able to visualize the size and patterns due to high-resolution 3-D images (Fig. 2). Although this study focused on calcifications, the improved imaging capability may lead to earlier detection of other morphological changes that signify breast cancer. In summary, SPC- $μ CT$ can make quantitative and multiscale measurements over the whole breast, and it has prospects for clinical use.

Fig. 2

Multiscale imaging with SPC- $μ CT$ depicts tissue in 3-D with higher resolution and soft-tissue contrast than 2-D single-scale clinical imaging. Panels (a)–(c) show slices from an SPC- $μ CT$ volume. Panel (d) shows a digital mammogram while panel (e) shows a breast tomosynthesis image. There are microcalcifications on each of these images that are pointed to by the white arrows, and specified in a region of interest. The volume in (a–c) has high contrast and locality due to its mesoscale resolution across a whole-breast 3-D volume. © European Society of Radiology 2016.⁶⁰

PhC- $μ CT$ derives contrast from the phase shift of the x-rays passing through the tissue.¹¹³ There are several different methods for PhC- $μ CT$ that are used in preclinical imaging. However, clinical methods are more limited due to technological constraints.¹²⁸ For example, past implementations of PhC- $μ CT$ have imparted too high radiation doses for clinical trials, but groups have recently demonstrated acceptable doses in phantom models.⁶²^,¹²⁹ Another important caveat to PhC- $μ CT$ systems is that prior to 2013 all systems used a synchrotron as an x-ray source.¹¹³^,¹³⁰ A synchrotron is an expensive facility rarely attached to hospitals, so implementation of such systems would be highly limited. Encouragingly, there have been studies reporting PhC- $μ CT$ using standard x-ray tubes and with acceptable radiation doses, which gives the prospect for a more widespread implementation.⁶³^,¹²⁹^–¹³¹ With the improved resolution of such systems, most recent breast PhC- $μ CT$ studies have had mesoscale resolution.⁷⁸^,⁶¹^–⁶³^,¹²⁹^,¹³¹^,¹³² PhC- $μ CT$ mirrors SPC- $μ CT$ in being an upcoming monomodal multiscale system that might be implemented clinically. Although it is more difficult and costly to implement than SPC- $μ CT$ , it also has several advantages and offers complementary information that may give both technologies a strong future in QMIB.

3.1.2.

High-frequency ultrasound

US, imaging through sound waves, is clinically friendly and is developing strong quantitative imaging capabilities. It is noninvasive, nonionizing, relatively inexpensive, portable, and can be quickly performed. In addition, it is the easiest way to image important biomechanical properties such as stiffness.³¹^,¹³³^,¹³⁴ US can perform quantitative imaging with quantitative US (QUS) and US elastography (USE). QUS can make system-independent estimate of acoustic parameters,¹³⁵^–¹³⁷ such as attenuation, backscatter, and mean scatterer spacing. This removes a large source of variance, which is important for clinical application. USE measures tissue elastic properties by applying a force to the tissue and tracking the deformation.¹³³^,¹³⁸^,¹³⁹

There are several factors that can affect the results of QUS and USE. The parameter estimates can be model-based¹³⁵^,¹³⁶^,¹⁴⁰ or be model-free.¹³⁷^,¹⁴¹ Thus, it is important to consider, and validate, the acoustic model and the assumptions involved. In addition, some parameter estimates (e.g., strain elastography) can be heavily dependent on the user,¹³⁹ whereas others can be user—and even system-independent.¹⁴⁰^–¹⁴⁴ Encouragingly, there are several imaging systems that can reduce user dependence in preclinical research. For example, there are whole-breast US imaging systems in clinical trials that can perform both QUS and USE.¹⁴⁵^–¹⁴⁸ Overall, US is a promising modality for QMIB, but researchers need to validate assumptions and experimental implementations.

US is typically separated into clinical US (2 to 20 MHz) and high-frequency US (HF-US) ( $> 20 MHz$ ). Clinical US images at the macroscale and is common in the clinic for several existing and upcoming applications.⁵¹ HF-US images at the mesoscale and has commercial preclinical instruments but has not yet reached the clinic.⁶⁶^,¹⁴⁹ Both types of US can be incorporated into QMIB studies, but so far there have been few QMIB studies performed with clinical US.¹⁴²

HF-US is common in multiscale imaging studies for several applications (Table 3). One notable application is improving US-guided biopsies. Clinical US systems cannot optimally visualize small microcalcifications, thus preventing accurate US-guided biopsy sampling of lesions containing microcalcifications. By comparison, HF-US does have high enough resolution to visualize microcalcifications. However, HF-US has much lower penetration depth. The lower penetration depth can be overcome by combining HF-US with needle-based probes.¹⁵⁰^–¹⁵³ Cummins et al. developed such a HF probe and performed multiscale imaging with by combing with simultaneous external clinical US.⁶⁶ Other applications include better characterizing phantom tissue,¹⁵⁴ tracking cell death from macro- to submicroscales,⁶⁷^,⁹²^,¹⁵⁵ detecting metastatic regions in lymph nodes,¹⁵⁶ and characterizing contrast agent biodistribution.⁹³

The aforementioned studies are the first few to explore this modality with QMIB, with many more potential opportunities. The parameters that HF-US measures reflect the tissue microstructure,⁴⁷^,¹³⁷^,¹⁴¹^,¹⁵⁷^,¹⁵⁸ which is important in breast cancer development and progression.³⁶^,³⁷^,⁴⁸ Such parameters could be mapped to other modalities, thus quantifying their sensitivity to microscale structure (Fig. 3). This may allow US to detect different regions in a tumor, which may respond differently to therapy.⁵^,⁹^,¹⁰ In addition, the biomechanical information that USE can provide is directly important in cancer imaging, such as tumor heterogeneity,⁴⁸ but can also support other imaging modalities by improving image registration models.¹⁶⁰ Finally, QMIB can also be used to improve the models used by US at all resolutions, by comparing them to modalities that image biology on smaller scales.¹⁵⁹ These factors make it likely that HF-US will be one of the main modalities for QMIB in the future.

Fig. 3

Multiscale imaging with HF-US and second harmonic generation microscopy (SHG) can link quantitative US measurements to different regions of collagen structure.¹⁵⁹ Patterns of collagen alignment are prognostic in breast cancer.²² As such, the combination of HF-US and SHG may lead to clinically relevant HF-US metrics for breast cancer. This figure is an unpublished example of HF-US combined with SHG microscopy on a breast cancer biopsy. It compares an HF-US image in panel (a) to a corresponding SHG image in panel (b). The images from US and SHG are registered to create the multiscale image of panel (c). Three tumor regions with different collagen structure are enlarged in panels (d–f). The data in this figure come from the Laboratory for Optical and Computational Instrumentation and the Hall lab at the University of Wisconsin–Madison.

3.1.3.

Magnetic resonance microscopy

MRI is a noninvasive imaging modality that is highly sensitive to the relaxation rate of many atomic protons and/or neutrons, but particularly hydrogen protons, that are returning to equilibrium after they were perturbed by pulses of radiofrequency energy. The sequence of MRI excitation and signal readout segments can be assembled in varied ways to make the measurement sensitive to different tissue properties. This allows MRI to perform anatomical, functional, and molecular imaging. MRI can be implemented on the macroscale and the mesoscale. The macroscale implementation is becoming a key tool in breast cancer treatment and diagnosis.⁴⁹^,¹⁶¹ As such, there is great interest in making MRI measurements quantitative. Many researchers are tackling this problem, but there are calls for robust multicenter studies to evaluate reproducibility and accuracy.¹⁶²^,¹⁶³ However, macroscale MRI is not used in many multiscale imaging studies. The mesoscale implementation of MRI, also known as magnetic resonance microscopy (MRM),¹⁶⁴ requires high magnetic field strengths, fast switching magnetic field gradients, and/or long imaging times to obtain mesoscale resolution. This makes it unsuitable for clinical imaging, but this requirement can be fulfilled by commercial preclinical systems.

Preclinical MRM has been used in several multiscale imaging applications, with varying degrees of quantification. In one study, researchers combined quantitative MRM with intravital-window microscopy, studying tumor growth in 3-D and mapping it to the cellular and molecular changes that cause the growth (Fig. 4).⁸² This information can aid tumor therapy research, as it links the response of the individual tumor cells to the response of the whole tumor. Other small-animal research groups demonstrated multicontrast characterization of cancer,⁸² monitoring of therapy response,⁹⁴ and visualization of vasculature and angiogenesis.⁸⁴^,¹²⁰ A few groups studied MRM of human tissue samples. They found MRM useful for diagnosis and for ex vivo tumor margin assessment.⁶⁸^,¹⁶⁵ These studies show that MRM has high potential for QMIB.

Fig. 4

Example of combined MRM and multiphoton microscopy of a mouse implanted with a breast cancer cell line and with an optical window. Multiscale imaging that combines MRM and multiphoton microscopy can link quantitative measurements of tumor morphology and growth (MRM) to corresponding cellular and molecular changes (multiphoton microscopy). This imaging can aid tumor therapy research by tracking how cellular interactions lead to whole-tumor response. Panel (a) is taken with multiphoton microscopy (through the window) and panel (B) T1-weighted MRM. The images were coregistered and two perpendicular slices from 3-D volumes are featured in the illustration. A growing tumor vessel is highlighted by the asterisk in (a), which corresponds spatially to the arrow in (b). © 2009 BioTechniques. Used with permission.⁸²

MRM has many prospects for future QMIB studies. There are many quantitative parameters that are not currently used in breast MRM. For example, one group demonstrated quantitative maps of anatomical parameters in the brain.¹⁶⁶ Other studies have focused on improving quantitative data collection and analysis with MRM, to replicate the clinical efforts for robust metrics, e.g., cellularity, vascular properties, and metabolites.¹⁶²^,¹⁶³ Researchers could also look to bring MRM into the clinic. Clinical MRI units can now reach into the low macroscale in the hundreds of microns but are not quite capable of whole breast MRM in vivo.⁶⁸ Seven-Tesla field-strength clinical scanners, a technology being validated in clinical trials, have greatly improved resolution in MRI but the imaging times are too long to obtain mesoscale resolution in patient imaging.⁴⁹ In vivo MRM will either require significant improvements in imaging time or a higher field-strength generation of MRI scanners. If it is accomplished, MRM will become a potent tool for in vivo clinical QMIB. Until then, its preclinical implementation shows promise for many different QMIB applications.

3.2.

Optical Microsopy

Optical microscopy encompasses imaging modalities that are the primary methods for clinical screening and for research of microscale biology.¹⁶⁷^,¹⁶⁸ Quantitative optical microscopy is still a developing field, but it is increasingly used in QMIB studies. There are a large number of optical microscopy imaging modalities, but they can be classified into a few categories based on their usage and application. This section covers QMIB using quantitative histopathology (QHP), laser scanning microscopy (LSM), and wide-field microscopy (WFM).

3.2.1.

Quantitative histopathology

Histopathology is one of the oldest techniques used in cancer imaging yet remains the gold standard for diagnosing breast cancer and for evaluating the capabilities of other imaging modalities.¹⁶⁷ It provides significant knowledge about the tumor through tissue stains examined under light microscopy. Major stains include the hematoxylin and eosin (H&E) stains for epithelial and stromal topology, or IHC stains for molecular markers. Until recently this has been a wholly qualitative practice, with the radiologist analyzing the macroscale image and the pathologist the microscale.⁴⁶ The advent of digital pathology imaging and whole-slide scanning technology over the last decade has lead to QHP.¹⁶⁹ The development of commercial and open-source tools for QHP is also making it more accessible to researchers and pathologists.¹⁷⁰ Several applications of QHP for breast cancer have been approved by the FDA and many others are in various stages of the approval process.⁴⁶ A number of excellent recent reviews cover the subject of QHP for cancer in general and specifically for breast cancer.⁴⁶^,¹⁶⁹^,¹⁷¹^,¹⁷²

The primary use of QHP in QMIB is to validate imaging modalities at other spatial scales by diagnosing pathology, but there are many secondary uses. One use of multiscale QHP is quantifying the biology of MD in terms of the PMD of the whole breast⁴¹^,¹⁷³^–¹⁸⁰ or the local density (LD) of tissue.¹⁸¹^–¹⁸⁵ The relative proportion of highly x-ray attenuating, radio dense, fibroglandular tissue to adipose tissue (MD) is an important risk factor in breast cancer, with up to a 4- to 6-fold difference in risk between women with high and low MD women.¹⁸⁶ It is not known whether MD is a cause or a consequence of this increased risk. If causal, elevated MD could be responsible for up to 26% of breast cancers in younger women and 16% overall.¹⁸⁷ Unlike many other risk factors MD can be modified and is potentially a clinical target for intervention. However, it is not known what biological property of dense breast tissue affects the risk. The increased amount of epithelial cells, where most breast cancers originate, is not a definitive explanation.²⁵^,⁴⁴ The percentage dense area on a mammogram is a stronger risk factor than the absolute dense area, suggesting that other mechanisms are at least partially responsible.¹⁸⁸ Complementary theories posit that dense breasts have differences in the tissue microenvironment that result in causal or correlated changes in cellular signaling pathways, stromal organization, and other biological mechanisms that are known to affect cancer development and metastasis.²⁶^,⁴⁴^,¹⁸⁹ As such, the distinction between PMD and LD is important to note for testing a hypothesis of systemic differences between high- and low-density breasts. There are no strong divergent trends according to current literature.¹⁷⁴ However, future differences may be discovered as multiscale QHP spreads and facilitates a greater number of studies.

Other applications of multiscale QHP are still uncommon, but those that exist show the untapped potential of the modality. A recent study demonstrated the potential of examining macroscale phenotypes. Bae et al.³⁸ quantified tumor phenotypes obtained using MRI and found that tumor roundness is an indicator of molecular traits, with negative correlations to ER and PgR statuses and a positive correlation to the cellular proliferation. Studying tumor and parenchymal pattern phenotypes with QHP is a fertile area for new multiscale research and compliments radiogenomic discoveries that link phenotype to genetic risk factors.¹⁹⁰ There are many more types of macroscale measures, including functional and molecular characteristics that have potential to be examined using multiscale QHP.

There are several near-term challenges for improving QHP and making it a better tool for QMIB. Currently QHP is performed on only a small segment of the tissue in any macroscale voxel and so cannot be easily localized.¹⁹¹ It was not feasible to address this issue in the past, but the advent of high-throughput whole-slide scanning systems has opened new doors. Such scanners rapidly process dozens of adjacent histology slides, allowing reconstruction of 3-D volumes.¹⁹² At least two small animal studies use these systems for QMIB. They developed workflows to register to MRI and widefield fluorescence.⁴⁰^,⁹⁵ There is much room to build on these and other similar methods to make 3-D QHP an accessible and valuable tool for research.¹⁸ One final limitation that should be considered is that QHP is inherently destructive to tissue and can only be done ex vivo. The tissues need to be physically sectioned into slices and placed on slides, and they need to be chemically processed to label what is being imaged. This can introduce artifacts throughout the process that effect the finished slides.¹⁹³ These artifacts are minor problems for qualitative applications but heavily impact quantitative analysis, especially in prospective 3-D applications.

3.2.2.

Laser scanning microscopy

LSM is a category of optical imaging modalities that can perform noninvasive 3-D imaging of microscale tissue composition. Notable LSM modalities include confocal fluorescence microscopy, fluorescence lifetime imaging microscopy (FLIM), multiphoton microscopy (MPM), and second harmonic generation (SHG).²²^,⁸¹^,¹⁹⁴^,¹⁹⁵ LSM imaging derives contrast from intrinsic biology or extrinsic molecular probes.¹⁹⁵^,¹⁹⁶ These contrast methods are nondestructive and, unlike histopathology, can be used for live imaging. In addition, the contrast options allow LSM to measure anatomical and molecular tissue composition. These measurements can be quantitative, though with varying difficulties and limitations based on the LSM implementation.²²^,¹⁹⁷^,¹⁹⁸ LSM can image at depths from 100 to $1000 μ m$ , depending on the tissue composition and the LSM modality. Given long imaging times, LSM technologies can also reach mesoscale field of view by stitching multiple images together into a mosaic. Altogether, these qualities make LSM one of the best research tools to characterize benign tissue and tumor microenvironments.¹⁸⁹^,¹⁹⁹^–²⁰¹

LSM’s capability to characterize tissue microenvironments makes it a valuable tool for breast cancer research with great potential for QMIB. This has been shown in tissue sample imaging and in live small animal research. SHG can obtain prognostic information from human breast cancer biopsies.²⁰²^–²⁰⁴ This prognostic information depends on microscale structure, which some mesoscale imaging modalities are sensitive to (Fig. 3).¹⁵⁹ Thus, future QMIB studies may allow the quicker mesoscale imaging modalities to detect this prognostic information. Other human biopsy studies used SHG to quantify microscale collagen characteristics against the macroscale measure breast density, supplementing the QHP methods mentioned in Sec. 3.2.1.⁹⁷ LSM can also be used to image tumor development in small animals. These studies implant optical windows over the mammary gland, known as intravital windows.⁸²^,²⁰⁵^–²⁰⁷ These windows allow LSM to bypass the skin and image the tumor directly. For example, one group combined MPM and MRI for multiscale imaging of tumor vessel development (Fig. 4).⁸² These studies demonstrate how LSM is valuable for QMIB and hint at its future potential.

There are many ways LSM can be advanced for QMIB research. LSM is widely used in research but is not currently present in the clinic due to equipment complexity, difficulty of application, and long imaging times. However, this may soon change as there are groups working on producing simplified equipment, needle or endoscope compatible imaging probes, and improved imaging speeds.¹⁹⁷^,²⁰⁸^,²⁰⁹ These devices can also be incorporated in multimodal imaging devices that inherently coregister the image, greatly easing multiscale research. Researchers can also tackle the depth limitation in tissue samples using physical sectioning, as is done in 3-D QHP. Physical sectioning with LSM would introduce tissue processing artifacts, but these may be lessened compared with 3-D QHP due to greatly increased section thickness.¹⁹³ Finally, there is a general need for quantitative imaging standards and improvements to adapt LSM for robust imaging research and clinical use.²²^,¹⁹⁷^,¹⁹⁸

3.2.3.

Wide-field microscopy

Many optical microscopy imaging modalities can be implemented in wide field, sacrificing resolution to achieve mesoscale field of view but not improved penetration depth. Examples include fluorescence, polarimetry, FLIM, and near-infrared light (NIR) spectroscopy.⁹⁵^,²¹⁰^–²¹² The penetration depth limit mostly restricts them to 2-D imaging, but 3-D imaging is possible ex vivo with serial sectioning of biopsies to obtain adjacent tissue slices and reconstructing those images.⁴⁰^,⁹⁵ They are being investigated for clinical use with IMA, where QMIB studies use histology for validation.⁶⁹^,²¹³ In addition, recent studies have shown improved resolution in some applications. For example, a lens-free and electronic chip-based technology can achieve mesoscale field of view and microscale resolution in a short timeframe and can be substituted for histology using false-color algorithms.²¹⁴^,²¹⁵ Wide-field imaging’s future QMIB applications, outside of validation against histology, are somewhat limited by their restriction to surface imaging, but they are likely to be seen in more ex vivo tissue studies as 3-D sectioning technologies advance.

3.3.

Biophotonics

Biophotonics, the study of optical and NIR light interactions with biological systems, is a field that has seen explosive growth over the last few decades. This growth is due to the desirable qualities of light at these wavelengths and technical advances in detection and illumination technology. These wavelengths of light are nonionizing. They allow a range of contrast options due to their absorption, scattering, and transmission properties. In addition, they do not require biochemical labels to generate contrast (though several can use contrast agents). Finally, it is relatively easy to generate monoenergetic optical and NIR light. These advantages have resulted in a collection of imaging modalities that operate at different spatial scales.¹⁶⁸^,²¹⁶ These modalities are frequently seen in different multimodal and multiscale combinations. This section covers five biophotonic modalities: optical coherence tomography (OCT), photoacoustic tomography (PAT), diffuse optical tomography (DOT), and fluorescence and luminescence tomographies.

3.3.1.

Optical coherence tomography

OCT is a mesoscale in vivo imaging modality that is noninvasive, free of biochemical labels, images rapidly, and can be fit onto compact probes.⁷⁶ It is analogous to US for optical waves, giving anatomical contrast through optical scattering caused by differences in tissue refractive index. It is used clinically for several surface and endoscopic imaging applications, but is at the preclinical stage for cancer imaging.²¹⁷ The near-term clinical applications of OCT are assessing clinical margins for intraoperative surgery (IMA) and biopsy guidance, which are both benefitted by QMIB. Numerous studies, both qualitative and quantitative, have paired it with histology to validate its capability to differentiate pathologic breast tissue in vivo.⁷⁶ Several studies have developed quantitative diagnosis algorithms for IMA and validated them against OCT, both alone and in combination with other modalities.⁷⁰^,⁹⁸ OCT can be added onto biopsy needle probes and can be used to ensure biopsies are correctly sampling the tumor.⁹⁹^,²¹⁸

OCT has several extensions used in breast imaging research, which could be applied to future QMIB studies. One extension measures attenuation, which can be calculated using automatic algorithms. These attenuation maps can help improve contrast of pathological tissue.²¹⁹ Polarization-sensitive OCT can measure how much light polarization changes as it goes through tissue, a property known as birefringence. Birefringence is primarily influenced by microscale collagen, making it an indirect measure of the microscale structure of tissue.²²⁰ Mechanical OCT, or optical coherence elastography (OCE), can measure tissue strain. Strain is a relative quantity, so it is subject to high variance. However, there is high interest in developing OCE that can measure the elastic modulus, which is a quantitative reliable measure.²²¹ Finally, there are other OCT extensions that have not been used in breast imaging research, such as blood flow imaging, but may be valuable for future QMIB research.

OCT is an exciting QMIB modality because it is well developed but still has room for growth. Its existing applications have broad applicability to breast cancer diagnosis and treatment in vivo (Fig. 5), but still need work to transition to clinical use.⁷⁶ There are other exciting possibilities, for example, minimally invasive needle probes that could reduce the necessity of biopsies, to build on in the future.²¹⁸ The various extensions give it other contrast options that are less well investigated, which makes it likely that new applications will arise from them. There are also many macroscale modalities it can be paired with for other investigations, of which some have been demonstrated outside the breast.²²³

Fig. 5

OCT has high potential for QMIB because it has prospects for patient imaging, where it can detect anatomical features useful in breast cancer diagnosis and treatment. This figure compares OCT (a and c) to corresponding histopathology (b and d) from a normal (a and b) and metastatic (c and d) human lymph node. The anatomical features of the lymph nodes can be seen by both modalities. The OCT images are generated as part of a 3-D volume (right). This work is licensed under a https://creativecommons.org/licenses/by/4.0/ Creative Commons Attribution 4.0 International License and is attributed to Nolan et al.²²²

3.3.2.

Photoacoustic tomography

PAT is one of the more promising frontiers of multiscale imaging. It combines rich optical contrast options with acoustic signal.²⁷^,¹⁴⁶ In PAT, a laser is used to illuminate areas of tissue. The tissue is heated by the absorption of photons, and this causes it to expand rapidly, producing an acoustic signal that is detected by US transducers. This is known as the photoacoustic effect. The number of photons absorbed by the tissue, and thus the signal generated, varies based on the tissue composition and the wavelength of light from the laser. This effect is quantifiable and can target several molecules in tissue, such as hemoglobin or collagen. The signal can also be enhanced by contrast agents. The varied contrast options allow PAT to perform anatomical, functional, and molecular imaging for many biomedical applications, often at the same time. In addition, PAT has several other advantages that make it well suited to multiscale imaging. It is easily combined with US, as US transducers both detect and emit acoustic waves. Finally, PAT can be implemented for any of the three scales, with configurations capable of imaging microscale organelles ranging up to imaging the macroscale breast.¹⁹^,²⁷^,¹⁴⁶^,²²⁴ There are commercial systems for clinical macroscale PAT and preclinical mesoscale PAT, making it more accessible to researchers.²²⁴

Several groups have used the macroscale and mesoscale implementations of PAT for QMIB. Some examples of QMIB with macroscale PAT include detecting micrometastases,⁸⁸ finding and distinguishing between benign and malignant microcalcifications,²²⁵ mapping metastatic sentinel lymph nodes,³⁹^,⁸⁸^,²²⁶ and for tracking tumor angiogenesis.⁷⁷ The mesoscale implementation of PAT has also been used in a few QMIB studies. Two studies examined an in vivo multimodal contrast agent for MRI, PET, and PAT using a reporter gene.¹⁰⁰^,²²⁷ A third study validated an in vivo cell-death contrast agent against the gold standard ex vivo fluorescence imaging.⁷¹ All of these studies demonstrate the potential utility of QMIB with PAT.

There are many future directions for PAT research. One of the largest standing problems in PAT is measuring the native optical fluence, which can introduce significant unquantified noise to some measurements.¹⁴⁶ There are also many opportunities to perform new QMIB research with PAT. We have mentioned several mesoscale and microscale PAT studies, but they are still largely unexplored for QMIB. There are very QMIB few studies using its microscale implementation, photoacoustic microscopy (PAM).²²⁴ Significant advancements could also be made to expand the scale range of individual instruments, which might enable a singular PAT system capable of imaging over multiple scales.¹⁴⁶

3.3.3.

Diffuse optical tomography

DOT is a form of whole-breast imaging based on NIR scattering and absorption. It operates at a lower resolution than most macroscale modalities seen in the clinic but has several advantages that make it well suited to translational and multiscale breast imaging research.⁷⁴^,²¹⁶ The main molecules that interact with the NIR, known as fluorophores, are oxy- and deoxyhemoglobin, water, lipid, and collagen. The fluorophores have different absorption and scattering profiles over the NIR. DOT can map the spatial distribution of these fluorophores by imaging at multiple wavelengths, separating out the absorption and scattering contributions of each fluorophore. Hemoglobin imaging allows vasculature and oxygenation imaging, important subjects for studying angiogenesis or for diagnosis and treatment.²¹⁶ Water, lipids, and collagen are used in many breast density quantification schemes.¹² Collagen composition is also important for diagnostic and prognostic reasons.¹⁸⁹ In addition, the optical scattering parameters are useful on their own, as they change in pathologic tissue.²²⁸ However, the accuracy and resolution of these measurements is limited by light propagation models. Light propagation can vary dramatically by tissue type, so models need large volumes to make accurate calculations. This issue can be overcome using multiscale imaging. Multiscale imaging gives prior knowledge of the tissue composition, allowing models with finer resolution and better accuracy. In summary, DOT can obtain several tissue composition parameters, and these measurements can be improved using multiscale imaging.

DOT is relevant to QMIB because it is a good example of coregistered imaging between significantly different resolutions, though thus far only within the macroscale. Some major modalities it has been paired with include US, x-ray tomosynthesis, MRI, and photoacoustic imaging.⁷²^,¹⁰¹^,¹⁰²^,²¹²^,²²⁹^–²³¹ There are several clinically relevant findings from these studies. Two groups found that using it alongside US could improve diagnostic performance and decrease the amount of biopsies of benign tissue.⁷²^,²³¹ Combining it with x-ray tomosynthesis allowed better differentiation between malignant tumors, benign lesions, cysts, and normal fibroglandular tissue.¹⁰² Similar results were also obtained with MRI.²¹² It was also combined with photoacoustic imaging methods, which are reviewed above, to track the biodistribution of a multimodal contrast agent attached to a cancer treatment drug.⁷³^,²³² DOT is unlikely to be paired directly with mesoscale modalities in the near future, but several other biophotonic modalities share some of its principles for QMIB.

3.3.4.

Fluorescence and luminescence tomography

Fluorescence molecular tomography (FMT) and diffuse luminescent imaging tomography (DLIT) are small animal imaging modalities that operate in the low-mesoscale and high-macroscale range.¹⁸⁰^,²³³ FMT is also known as diffuse laminar optical tomography. They are mathematically and conceptually similar to DOT, although with several important distinctions. FMT can quantify the same fluorophores as DOT, but can also be used to image fluorescence from molecular probes.¹⁸⁰ DLIT is based on luciferase, which emits light during enzyme reactions in live cells, and DLIT requires a priori knowledge of luciferase production in different cell types.²³³ However, DLIT is also significantly less noisy than fluorescence-based imaging and is sensitive to as few as a thousand tumor cells.²³⁴

These modalities have only recently become capable of sub $100 μ m$ resolution.¹⁸⁰^,²³⁴^,²³⁵ Nonetheless, they have been used in several QMIB applications, which include imaging tumor apoptosis,¹⁰⁵ tracking metastasis with nanoparticles,⁸⁵^,¹⁰⁴ quantifying tumor growth and metastasis parameters,¹⁰⁶ and validating a multimodality genetic contrast agent.⁸⁰ FMT and DLIT still face issues in precise quantification, though new methods are being developed to handle these issues.¹⁸⁰ Both have great potential use in QMIB as they mature and are used in new combinations. It should also be noted that there are nonbreast examples of their use for multiscale or multimodal imaging, such as a combination with OCT for phantom imaging.²³⁶

3.4.

Quantitative Multiscale Imaging Outside the Breast

The previous sections covered the existing QMIB modalities, but future researchers might take inspiration from biomedical QMI that has been demonstrated outside the breast. Reusch et al.¹⁵⁹ combined SHG and US to do preclinical imaging of the nonpregnant uterine cervix. Future QMIB studies could benefit from similar methods, as collagen alignment is prognostic in breast cancer.²⁰² Liang et al.²²³ built an MRI compatible OCT probe for intraoperative surgery. They showed that the probe can gather complementary information from tissue samples, and that such information could improve the efficiency and accuracy of surgeries. This surgical probe has obvious applications in breast cancer research, as intraoperative surgery is a common topic. Hipwell et al.¹⁶⁰^,²⁰³ developed an optomechanical device that synchronized SHG imaging with tissue deformation, mapping mechanical properties to microscale structure. This is relevant to breast research, as mechanical properties are a risk factor for breast cancer.³¹ Many research groups have conducted multiscale brain imaging with cranial windows.²³⁷^–²³⁹ These methods are similar to the intravital windows used in QMIB research and could be applied to future studies. In summary, these selected examples have relevance to our current knowledge of breast cancer and could be potent tools for QMIB. Readers interested in more information of biomedical QMI in general might reference recent reviews of multiscale imaging.²⁴⁰^–²⁴⁴

4. Data Analysis Challenges

Data analysis and processing is the key component of quantitative imaging. It acts as a gatekeeper to the clinic, as a quantitative imaging modality needs effective and efficient analysis to facilitate clinical adoption.²⁴⁵ This section covers some of the unique challenges that QMIB faces. It is not meant to be exhaustive but rather to highlight important aspects of the developing field. Readers interested in more general information on data analysis methods might reference some excellent reviews for data analysis in breast cancer,⁷^,⁴⁶ general oncology,¹⁶⁷^,²⁴⁶ biomedical imaging informatics,¹⁴ translational imaging,²⁴⁷ and big healthcare data management.²⁴⁸

4.1.

Hardware and Software Limitations

Multiscale imaging produces large multidimensional datasets that in turn require exponentially more processing power to analyze than typical radiological images.²⁴⁸ Advances in computational power over the last decade have been staggering, but there remains a great deal of ground to be covered before some QMIB modalities become practical for a clinical setting.²⁴⁷ There are medically relevant engineering and programming solutions that can be investigated to speed this transition. Dedicated analysis hardware, such as chipsets that replace parts of algorithms, can be several times more efficient than general computing algorithms. Guerra et al. demonstrated the value of this process, as it allowed them to make the first handheld time-domain OCT probe.²⁴⁹ This approach is particularly relevant where there is little preexisting hardware and can be included in basic design developments. Optimization of software for efficiency is another major area of development. In many cases, software for QMIB is being developed alongside the systems, and there will be room for improvements in speed and architecture at all steps of the process.

4.2.

Registration and Biomechanical Modeling

Many multiscale imaging systems are multimodal and produce independent images, so there is a strong need for image registration to fuse these independent images. Image registration is the process of mapping spatial points on one image to those on another image, often resulting in a fusion image that displays information from both. The breast is a difficult organ to register, as it is made of soft tissue that deforms nonlinearly, thus altering landmarks. As such, breast image registration is an active field of research and only a few methods have been applied clinically.¹⁶⁰

QMIB occupies an interesting area of the field, as it both requires unique registration solutions and can contribute to developing better registration algorithms. Multiscale registration needs to take into account the different size scales of images, which makes it difficult to accurately map the high-resolution image to the low-resolution image. In addition, multiscale registration frequently contends with different imaging geometries, which can alter landmarks and deform the tissue. However, other forms of QMIB offer solutions to registration problems. For example, QMIB can help address the breast deformation problem. Accurately modeling breast deformation requires good multiscale biomechanical models.¹⁶⁰ QMIB can improve these models by accurately measuring anatomic changes and tissue mechanical properties across scales with one set of modalities. This information can then be incorporated into registration algorithms for another set of modalities. In summation, QMIB depends on registration but can also be used to improve existing registration algorithms.

4.3.

Segmentation

Segmentation is the task of delineating regions within an image based on some biological parameter, and it is essential for quantitative image analysis.²⁴⁷ Multiscale systems can complicate this problem by producing several images of distinct types with different resolutions and which do not necessarily have the same biological contrast. Multiple resolutions can cause misalignment due to partial volume effects, where the size of the macroscale pixel or voxel leads to errors in position of the low-resolution images. If the images have different biological contrast, they may not depict the same separations in biology that are used to segment the image initially. Accurate registration can mitigate this problem and turn it into an advantage, combining information on different traits to better segment the image. QMIB can also help advance segmentation. High-resolution multiscale data can improve the atlases and models that many segmentation methods use, or validate such methods with more accurate depictions of biology.²⁵⁰

4.4.

Imaging Biomarkers

Imaging biomarkers (IB) are defined imaging characteristics, which indicate biological processes or response to interventions.⁵³^,²⁵¹ IB provide objective measures to test hypotheses and can become tools for clinical decision making.²⁴⁵ Strong biomarkers, which can be diagnostic or prognostic, are essential elements providing utility to new systems or techniques and justifying their translation to the clinic. The breast imaging-reporting and data system (BI-RADS) represents a set of qualitative IB based on categorization and physician interpretation.³² For example, physicians commonly classify breasts by the BI-RADS breast density categories using 2-D mammograms (Fig. 6). This visual assessment is clinically useful, but it can vary based on the physician’s training and can be inaccurate compared with 3-D measures.²⁵²^–²⁵⁴ This leaves greater chance for error in patient care.²⁵²^–²⁵⁴ By comparison, a quantitative imaging biomarker (QIB) is an objective characteristic derived from an in vivo image measured on a ratio or interval.⁸ CADe and CADx algorithms use QIB, and trials have demonstrated that they can improve radiologist performance.⁷ Biomarker development is a broad field, and interested readers may want to reference several excellent recent reviews covering basic definitions,⁸ metrology,⁵³ and translation (Fig. 6).²⁴⁵

Fig. 6

MD as assessed by BI-RADS is a qualitative imaging biomarker, as it is visually assessed by physicians. These images show breasts classified by density under BI-RADS as (a) almost entirely fatty, (b) scattered areas of fibroglandular density, (c) heterogeneously dense, and (d) extremely dense.

There are two main ways that QMIB can lead to biomarker development: informed biomarkers and multiscale biomarkers. Informed biomarkers are developed using multiscale imaging, but do not use multiscale imaging when imaging the patient and assessing the biomarker. For example, OCT tumor margin detection is an informed biomarker. The biomarker was developed by correlation to the gold standard of histology, but in practice, only OCT is used.²⁵⁵ By contrast, multiscale biomarkers depend on multiscale information taken from the patient. Multiscale biomarkers are a much rarer method, as they require clinically friendly multiscale systems.⁵² One preclinical example of such a system involves dual-modality probes for fluorescence and radiographic imaging. It is difficult to characterize the probe’s biological interactions with only fluorescence imaging, as fluorescence imaging has a small field of view. This difficulty can be overcome by adding radiographic images to provide context, as they measure the probe over a much larger volume.²⁵⁶ Overall, both informed and multiscale biomarkers should become more common as the field of QMIB matures and are important to its ultimate clinical relevance.

4.5.

Computational Cancer Modeling

Computational modeling of cancer development and progression is a growing area relevant for both basic understanding and clinical application. For example, in one study the multiscale modeling of tumor growth indicated that some therapies used in breast cancer treatment could negatively impact long-term survival by selecting more dangerous phenotypes with environmental pressure.²⁵⁷ This was corroborated by another multiscale study, which found that environmental pressure encouraged predictable phenotypes, first with models and then experimentally verified with breast cancer models. Readers interested in more comprehensive knowledge of this type of modeling might refer to a recent review by Simmons et al.²⁵⁸ The extant examples of computational modeling only scratch the surface a field that is becoming increasingly accessible, and QMIB will be essential to validating such promising multiscale models in the future.

4.6.

Radiomics, Multiomics, and Precision Medicine

Radiomics is the process of building searchable medical imaging databases that can be mined for high-dimensional quantitative data.²⁵⁹ This collaborative effort yields data that can be analyzed and used in studies beyond the original, often in ways that were not previously possible. Building these databases involves turning images from a vast number of imaging modalities and their various applications into cross-institution and cross-modality quantitative information.²⁵⁹ QMIB generates cross-modality quantitative information and can help build these datasets. In addition, the datasets can help existing QMIB applications. For example, large cross-institution datasets would help address the sample size issues with breast density composition measurements (Sec. 3.2.1).

Radiomics is a subset of the Big Healthcare Data problem, where large amounts of information from various omic sources are being standardized, quantified, placed in computer archives, and processed to improve patient care.²⁴⁸ Integrating QMIB with multiomic research is another major path forward. For example, many studies are looking into radiogenomics, where genetic information is compared with imaging phenotypes.²⁶⁰ While few have been done with QMIB, the same principles could be applied. One of the main goals of these Big Healthcare Data initiatives is precision medicine, where previously unnoticed trends in these large datasets are used to develop methods for selecting and targeting treatments based on patient specific abnormalities.¹⁴ QMIB has great potential to contribute to precision medicine in breast cancer, contributing rich quantitative datasets on multiple biophysical characteristics. Doing so will rely on researchers to integrate radiomic concerns into their QMIB research and for all involved to build a collaborative data sharing spirit.

5. Discussion

Quantitative multiscale imaging of breast cancer is an area well posed for growth in both the research and clinical regimes. The relative ease of imaging the breast makes it a good testing ground for multiscale imaging technology, and this pairing could address many breast cancer research and clinical needs. In the past, multiscale imaging was largely performed using independent imaging modalities and had high skill and time barriers to entry. The most common use was validation, comparing the gold standard of histopathology to images from new diagnostic modalities. In the present, all-in-one preclinical multiscale systems and simplified multiscale workflows are becoming more common.¹⁷^,⁹¹^,⁹⁵^,²²³^,²²⁴^,²³⁶ Developments in data acquisition methods are starting to simplify quantitative imaging with historically qualitative modalities, such as MRI or US. Improvements in hardware and software are making quantitative data analysis more accessible. Many of the modalities have been integrated into multimodal systems. Some clinical applications are approaching viability, for example, rapid tumor margin imaging that use macroscale modalities for needle guidance and mesoscale for detecting the margins.

Still, there remains a great deal of work to be done in terms of both basic research or validation and the development of new systems. Multiscale imaging of the breast involves a wide range of modalities in various stages of development. A handful of applications are in or are nearly ready for clinical trials. Research-wise there are unexplored quantitative metrics that could be investigated in a multiscale fashion immediately. Other modalities will require the development of new metrics and the derivation of corresponding biomarkers to make them meaningful. Multiscale methods have been demonstrated that suffer from strict constraints, such as slow imaging speed or high cost, which render them impractical. Attention also needs to be paid to areas in data analysis and handling including new algorithms to transform metrics and biomarkers into an end-user friendly format and hardware for larger storage capacities and quicker processing speeds. Finally, and perhaps most importantly, researchers should develop sets of accepted standards and conventions for facilitating interstudy comparisons and moving through the translational research process, where they do not already exist.

6. Conclusion

Quantitative multiscale imaging of the breast is a rapidly evolving field that includes both the technology to enable investigations as well as the informational and clinical needs that drive them. It intersects with many breast imaging modalities and analysis approaches, with relevance to both clinicians and researchers. This review focuses largely on the technological challenges and benefits of QMIB. However, there are many additional issues that face any QMIB technology before clinical adoption. As with any new imaging technology, QMIB modalities being considered for clinical use are subject to rigorous FDA testing and evaluation. Other important clinical concerns for QMIB include ratio of cost effectiveness to outcome, availability, and ease of use. Finally, while this review is focused on breast cancer imaging, many of these same technologies, clinical needs, and research problems apply to multiscale imaging in other organ sites and for other pathologies.

Disclosures

The authors declare that there are no conflicts of interest related to this article.

Acknowledgments

We acknowledge funding from the Laboratory for Optical and Computational Instrumentation, the Morgridge Institute for Research and NSF CBET Award #1429045 (TJH and KWE). Research reported in this publication was also supported by the National Cancer Institute of the National Institutes of Health under Award #T32 CA009206 (MAP) and the National Institute of General Medical Sciences of the National Institutes of Health under Award #T32 GM008349 (MAP). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the National Science Foundation. We thank Eric Nordberg and Joseph Szulczewski for their assistance in the imaging done for Fig. 3. We also thank John Huston for his graphics design assistance.

References

1.

C. Fitzmaurice et al., “The global burden of cancer 2013,” JAMA Oncol., 1 505 –527 (2015). http://dx.doi.org/10.1001/jamaoncol.2015.0735 Google Scholar

2.

R. L. Siegel, K. D. Miller and A. Jemal, “Cancer statistics, 2017,” CA-Cancer J. Clin., 67 7 –30 (2017). http://dx.doi.org/10.3322/caac.21387 Google Scholar

3.

M. E. Barnard, C. E. Boeke and R. M. Tamimi, “Established breast cancer risk factors and risk of intrinsic tumor subtypes,” Biochim. Biophys. Acta, Rev. Cancer, 1856 73 –85 (2015). http://dx.doi.org/10.1016/j.bbcan.2015.06.002 Google Scholar

4.

J. D. C. Hon et al., “Breast cancer molecular subtypes: from TNBC to QNBC,” Am. J. Cancer Res., 6 1864 –1872 (2016). Google Scholar

5.

J. P. O’Connor et al., “Imaging intratumor heterogeneity: role in therapy response, resistance, and clinical outcome,” Clin. Cancer Res., 21 249 –257 (2015). http://dx.doi.org/10.1158/1078-0432.CCR-14-0990 Google Scholar

6.

M. J. Gerdes et al., “Emerging understanding of multiscale tumor heterogeneity,” Front. Oncol., 5 366 (2014). http://dx.doi.org/10.3389/fonc.2014.00366 Google Scholar

7.

M. L. Giger, N. Karssemeijer and J. A. Schnabel, “Breast image analysis for risk assessment, detection, diagnosis, and treatment of cancer,” Ann. Rev. Biomed. Eng., 15 (1), 327 –357 (2013). http://dx.doi.org/10.1146/annurev-bioeng-071812-152416 ARBEF7 1523-9829 Google Scholar

8.

L. G. Kessler et al., “The emerging science of quantitative imaging biomarkers terminology and definitions for scientific studies and regulatory submissions,” Stat. Methods Med. Res., 24 9 –26 (2015). http://dx.doi.org/10.1177/0962280214537333 Google Scholar

9.

P. M. Enriquez-Navas et al., “Exploiting evolutionary principles to prolong tumor control in preclinical models of breast cancer,” Sci. Transl. Med., 8 327ra24 (2016). http://dx.doi.org/10.1126/scitranslmed.aad7842 Google Scholar

10.

M. C. Lloyd et al., “Darwinian dynamics of intratumoral heterogeneity: not solely random mutations but also variable environmental selection forces,” Cancer Res., 76 3136 –3144 (2016). http://dx.doi.org/10.1158/0008-5472.CAN-15-2962 CNREA8 0008-5472 Google Scholar

11.

R. A. Gatenby, O. Grove and R. J. Gillies, “Quantitative imaging in cancer evolution and ecology,” Radiology, 269 8 –14 (2013). http://dx.doi.org/10.1148/radiol.13122697 RADLAX 0033-8419 Google Scholar

12.

M. A. Sak et al., “Current and future methods for measuring breast density: a brief comparative review,” Breast Cancer Manage., 5 209 –221 (2015). http://dx.doi.org/10.2217/bmt.15.13 Google Scholar

13.

M. J. Sherratt, J. C. McConnell and C. H. Streuli, “Raised mammographic density: causative mechanisms and biological consequences,” Breast Cancer Res., 18 45 (2016). http://dx.doi.org/10.1186/s13058-016-0701-9 BCTRD6 Google Scholar

14.

W. Hsu, M. K. Markey and M. D. Wang, “Biomedical imaging informatics in the era of precision medicine: progress, challenges, and opportunities,” J. Am. Med. Inf. Assoc., 20 1010 –1013 (2013). http://dx.doi.org/10.1136/amiajnl-2013-002315 Google Scholar

15.

P. Lambin et al., “Radiomics: extracting more information from medical images using advanced feature analysis,” Eur. J. Cancer, 48 441 –446 (2012). http://dx.doi.org/10.1016/j.ejca.2011.11.036 Google Scholar

16.

A. M. Rutman and M. D. Kuo, “Radiogenomics: creating a link between molecular diagnostics and diagnostic imaging,” Eur. J. Radiol., 70 232 –241 (2009). http://dx.doi.org/10.1016/j.ejrad.2009.01.050 EJRADR 0720-048X Google Scholar

17.

J. Cebulla et al., “Multiscale and multi-modality visualization of angiogenesis in a human breast cancer model,” Angiogenesis, 17 695 –709 (2014). http://dx.doi.org/10.1007/s10456-014-9429-2 Google Scholar

18.

V. Ntziachristos, “Going deeper than microscopy: the optical imaging frontier in biology,” Nat. Methods, 7 603 –614 (2010). http://dx.doi.org/10.1038/nmeth.1483 1548-7091 Google Scholar

19.

A. Taruttis, G. M. van Dam and V. Ntziachristos, “Mesoscopic and macroscopic optoacoustic imaging of cancer,” Cancer Res., 75 1548 –1559 (2015). http://dx.doi.org/10.1158/0008-5472.CAN-14-2522 CNREA8 0008-5472 Google Scholar

20.

C. Debbaut et al., “Analyzing the human liver vascular architecture by combining vascular corrosion casting and micro-CT scanning: a feasibility study,” J. Anat., 224 509 –517 (2014). http://dx.doi.org/10.1111/joa.12156 JOANAY 0021-8782 Google Scholar

21.

J. W. Bohland et al., “A proposal for a coordinated effort for the determination of brainwide neuroanatomical connectivity in model organisms at a mesoscopic scale,” PLoS Comput. Biol., 5 (3), e1000334 (2009). http://dx.doi.org/10.1371/journal.pcbi.1000334 Google Scholar

22.

P. Campagnola, “Second harmonic generation imaging microscopy: applications to diseases diagnostics,” Anal. Chem., 83 3224 –3231 (2011). http://dx.doi.org/10.1021/ac1032325 ANCHAM 0003-2700 Google Scholar

23.

C. S. Curran et al., “Collagen density regulates xenobiotic and hypoxic response of mammary epithelial cells,” Environ. Toxicol. Pharmacol., 39 114 –124 (2015). http://dx.doi.org/10.1016/j.etap.2014.10.017 Google Scholar

24.

J. Insua-Rodrguez and T. Oskarsson, “The extracellular matrix in breast cancer,” Adv. Drug Delivery Rev., 97 41 –55 (2016). http://dx.doi.org/10.1016/j.addr.2015.12.017 ADDREP 0169-409X Google Scholar

25.

P. J. Keller et al., “Defining the cellular precursors to human breast cancer,” Proc. Natl. Acad. Sci. U. S. A., 109 2772 –2777 (2012). http://dx.doi.org/10.1073/pnas.1017626108 Google Scholar

26.

J. Ursini-Siegel and P. M. Siegel, “The influence of the pre-metastatic niche on breast cancer metastasis,” Cancer Lett., 380 281 –288 (2016). http://dx.doi.org/10.1016/j.canlet.2015.11.009 CALEDQ 0304-3835 Google Scholar

27.

Y. Liu, L. Nie and X. Chen, “Photoacoustic molecular imaging: from multiscale biomedical applications towards early-stage theranostics,” Trends Biotechnol., 34 420 –433 (2016). http://dx.doi.org/10.1016/j.tibtech.2016.02.001 TRBIDM 0167-7799 Google Scholar

28.

H. Li et al., “Computerized analysis of mammographic parenchymal patterns for assessing breast cancer risk: effect of ROI size and location,” Med. Phys., 31 549 –555 (2004). http://dx.doi.org/10.1118/1.1644514 MPHYA6 0094-2405 Google Scholar

29.

V. A. McCormack and I. dos Santos Silva, “Breast density and parenchymal patterns as markers of breast cancer risk: a meta-analysis,” Cancer Epidemiol. Biomarkers Prev., 15 (6), 1159 –1169 (2006). http://dx.doi.org/10.1158/1055-9965.EPI-06-0034 Google Scholar

30.

A. T. Wang et al., “Breast density and breast cancer risk: a practical review,” Mayo Clin. Proc., 89 548 –557 (2014). http://dx.doi.org/10.1016/j.mayocp.2013.12.014 MACPAJ 0025-6196 Google Scholar

31.

N. F. Boyd et al., “Evidence that breast tissue stiffness is associated with risk of breast cancer,” PLoS One, 9 e100937 (2014). http://dx.doi.org/10.1371/journal.pone.0100937 POLNCL 1932-6203 Google Scholar

32.

C. D’Orsi et al., ACR BI-RADS Atlas, Breast Imaging Reporting and Data System, American College of Radiology, Reston, Virginia (2013). Google Scholar

33.

J. Kim et al., “Breast density change as a predictive surrogate for response to adjuvant endocrine therapy in hormone receptor positive breast cancer,” Breast Cancer Res., 14 R102 (2012). http://dx.doi.org/10.1186/bcr3221 BCTRD6 Google Scholar

34.

J. Li et al., “Mammographic density reduction is a prognostic marker of response to adjuvant tamoxifen therapy in postmenopausal patients with breast cancer,” J. Clin. Oncol., 31 (18), 2249 –2256 (2013). http://dx.doi.org/10.1200/JCO.2012.44.5015 JCONDN 0732-183X Google Scholar

35.

S. J. Nyante et al., “Prognostic significance of mammographic density change after initiation of tamoxifen for ER-positive breast cancer,” J. Natl. Cancer Inst., 107 (3), (2015). http://dx.doi.org/10.1093/jnci/dju425 JNCIEQ Google Scholar

36.

H. Tadayyon et al., “Quantitative ultrasound characterization of locally advanced breast cancer by estimation of its scatterer properties,” Med. Phys., 41 012903 (2014). http://dx.doi.org/10.1118/1.4852875 MPHYA6 0094-2405 Google Scholar

37.

L. Sannachi et al., “Non-invasive evaluation of breast cancer response to chemotherapy using quantitative ultrasonic backscatter parameters,” Med. Image Anal., 20 224 –236 (2015). http://dx.doi.org/10.1016/j.media.2014.11.009 Google Scholar

38.

M. S. Bae et al., “Quantitative MRI morphology of invasive breast cancer: correlation with immunohistochemical biomarkers and subtypes,” Acta Radiol., 56 (3), 269 –275 (2015). http://dx.doi.org/10.1177/0284185114524197 Google Scholar

39.

A. Garcia-Uribe et al., “Dual-modality photoacoustic and ultrasound imaging system for noninvasive sentinel lymph node detection in patients with breast cancer,” Sci. Rep., 5 15748 (2015). http://dx.doi.org/10.1038/srep15748 Google Scholar

40.

L. Jiang et al., “Combined magnetic resonance, fluorescence, and histology imaging strategy in a human breast tumor xenograft model,” NMR Biomed., 26 285 –298 (2013). http://dx.doi.org/10.1002/nbm.v26.3 NMRBEF 0952-3480 Google Scholar

41.

J.-M. B. Pang et al., “Breast tissue composition and immunophenotype and its relationship with mammographic density in women at high risk of breast cancer,” PLoS One, 10 e0128861 (2015). http://dx.doi.org/10.1371/journal.pone.0128861 POLNCL 1932-6203 Google Scholar

42.

A. Sztrkay et al., “High-resolution breast tomography at high energy: a feasibility study of phase contrast imaging on a whole breast,” Phys. Med. Biol., 57 (10), 2931 –2942 (2012). http://dx.doi.org/10.1088/0031-9155/57/10/2931 PHMBA7 0031-9155 Google Scholar

43.

C. Zhou et al., “Integrated optical coherence tomography and microscopy for ex vivo multiscale evaluation of human breast tissues,” Cancer Res., 70 10071 –10079 (2010). http://dx.doi.org/10.1158/0008-5472.CAN-10-2968 CNREA8 0008-5472 Google Scholar

44.

C. W. Huo et al., “Mammographic density-a review on the current understanding of its association with breast cancer,” Breast Cancer Res. Treat., 144 479 –502 (2014). http://dx.doi.org/10.1007/s10549-014-2901-2 BCTRD6 Google Scholar

45.

A. J. Ironside and J. L. Jones, “Stromal characteristics may hold the key to mammographic density: the evidence to date,” Oncotarget, 7 (21), 31550 –31562 (2016). http://dx.doi.org/10.18632/oncotarget.v7i21 Google Scholar

46.

M. Veta et al., “Breast cancer histopathology image analysis: a review,” IEEE Trans. Biomed. Eng., 61 1400 –1411 (2014). http://dx.doi.org/10.1109/TBME.2014.2303852 IEBEAX 0018-9294 Google Scholar

47.

T. Liu et al., “Inferring spatial variations of microstructural properties from macroscopic mechanical response,” Biomech. Model. Mechanobiol., 16 479 –496 (2017). http://dx.doi.org/10.1007/s10237-016-0831-0 BMMICD 1617-7959 Google Scholar

48.

T. Liu et al., “Noninvasive in-vivo quantification of mechanical heterogeneity of invasive breast carcinomas,” PLoS One, 10 e0130258 (2015). http://dx.doi.org/10.1371/journal.pone.0130258 POLNCL 1932-6203 Google Scholar

49.

G. L. Menezes et al., “Magnetic resonance imaging in breast cancer: a literature review and future perspectives,” World J. Clin. Oncol., 5 61 –70 (2014). http://dx.doi.org/10.5306/wjco.v5.i2.61 Google Scholar

50.

A. M. O'Connell, A. Karellas and S. Vedantham, “The potential role of dedicated 3D breast CT as a diagnostic tool: review and early clinical examples,” Breast J., 20 592 –605 (2014). http://dx.doi.org/10.1111/tbj.12327/full BRJOFK 1075-122X Google Scholar

51.

J. R. Eisenbrey, J. K. Dave and F. Forsberg, “Recent technological advancements in breast ultrasound,” Ultrasonics, 70 183 –190 (2016). http://dx.doi.org/10.1016/j.ultras.2016.04.021 ULTRA3 0041-624X Google Scholar

52.

“Quantitative imaging biomarkers alliance,” (2016) https://www.rsna.org/QIBA/ December ). 2016). Google Scholar

53.

D. C. Sullivan et al., “Metrology standards for quantitative imaging biomarkers,” Radiology, 277 813 –825 (2015). http://dx.doi.org/10.1148/radiol.2015142202 RADLAX 0033-8419 Google Scholar

54.

L. Yaghjyan et al., “Mammographic breast density and breast cancer risk: interactions of percent density, absolute dense, and non-dense areas with breast cancer risk factors,” Breast Cancer Res. Treat., 150 181 –189 (2015). http://dx.doi.org/10.1007/s10549-015-3286-6 BCTRD6 Google Scholar

55.

D. B. Kopans, “Basic physics and doubts about relationship between mammographically determined tissue density and breast cancer risk,” Radiology, 246 348 –353 (2008). http://dx.doi.org/10.1148/radiol.2461070309 RADLAX 0033-8419 Google Scholar

56.

M. Andr et al., “Quantitative volumetric breast imaging with 3D inverse scatter computed tomography,” in Annual Int. Conf. of the IEEE Engineering in Medicine and Biology Society (EMBC 2012), 1110 –1113 (2012). http://dx.doi.org/10.1109/EMBC.2012.6346129 Google Scholar

57.

R. Tang et al., “Intraoperative micro-computed tomography (micro-CT): a novel method for determination of primary tumour dimensions in breast cancer specimens,” Br. J. Radiol., 89 20150581 (2015). http://dx.doi.org/10.1259/bjr.20150581 Google Scholar

58.

W. M. Sarraj et al., “Prediction of primary breast cancer size and T-stage using micro-computed tomography in lumpectomy specimens,” J. Pathol. Inf., 6 60 (2015). http://dx.doi.org/10.4103/2153-3539.170647 Google Scholar

59.

I. Willekens et al., “High-resolution 3D micro-CT imaging of breast microcalcifications: a preliminary analysis,” BMC Cancer, 14 9 (2014). http://dx.doi.org/10.1186/1471-2407-14-9 BCMACL 1471-2407 Google Scholar

60.

W. A. Kalender et al., “Technical feasibility proof for high-resolution low-dose photon-counting CT of the breast,” Eur. Radiol., 27 1081 –1086 (2017). http://dx.doi.org/10.1007/s00330-016-4459-3 EURAE3 1432-1084 Google Scholar

61.

S. Pacil et al., “Clinical application of low-dose phase contrast breast CT: methods for the optimization of the reconstruction workflow,” Biomed. Opt. Express, 6 3099 –3112 (2015). http://dx.doi.org/10.1364/BOE.6.003099 BOEICL 2156-7085 Google Scholar

62.

A. Sarno et al., “Imaging performance of phase-contrast breast computed tomography with synchrotron radiation and a CdTe photon-counting detector,” Phys. Med., 32 681 –690 (2016). http://dx.doi.org/10.1016/j.ejmp.2016.04.011 PHYME2 1120-1797 Google Scholar

63.

M. Willner et al., “Quantitative three-dimensional imaging of lipid, protein, and water contents via x-ray phase-contrast tomography,” PLoS One, 11 (3), e0151889 (2016). http://dx.doi.org/10.1371/journal.pone.0151889 POLNCL 1932-6203 Google Scholar

64.

C. Mortensen et al., “Ultrasound tissue characterization of breast biopsy specimens: expanded study,” Ultrasonic Imaging, 18 (3), 215 –230 (1996). http://dx.doi.org/10.1177/016173469601800304 ULIMD4 0161-7346 Google Scholar

65.

L. Sannachi et al., “Non-invasive evaluation of breast cancer response to chemotherapy using quantitative ultrasonic backscatter parameters,” Med. Image Anal., 20 (1), 224 –236 (2015). http://dx.doi.org/10.1016/j.media.2014.11.009 Google Scholar

66.

T. Cummins et al., “High-frequency ultrasound imaging for breast cancer biopsy guidance,” J. Med. Imaging, 2 047001 (2015). http://dx.doi.org/10.1117/1.JMI.2.4.047001 JMEIET 0920-5497 Google Scholar

67.

H. Tadayyon et al., “Quantification of ultrasonic scattering properties of in vivo tumor cell death in mouse models of breast cancer,” Transl. Oncol., 8 463 –473 (2015). http://dx.doi.org/10.1016/j.tranon.2015.11.001 Google Scholar

68.

B. Z. Dashevsky et al., “The potential of high resolution magnetic resonance microscopy in the pathologic analysis of resected breast and lymph tissue,” Sci. Rep., 5 17435 (2015). http://dx.doi.org/10.1038/srep17435 Google Scholar

69.

J. Q. Brown et al., “Optical spectral surveillance of breast tissue landscapes for detection of residual disease in breast tumor margins,” PLoS One, 8 e69906 (2013). http://dx.doi.org/10.1371/journal.pone.0069906 POLNCL 1932-6203 Google Scholar

70.

J. G. Sun et al., “Segmentation and correlation of optical coherence tomography and x-ray images for breast cancer diagnostics,” J. Innovative Opt. Health Sci., 6 1350015 (2013). http://dx.doi.org/10.1142/S1793545813500156 Google Scholar

71.

M. A. Stammes et al., “The necrosis-avid small molecule HQ4-DTPA as a multimodal imaging agent for monitoring radiation therapy-induced tumor cell death,” Front. Oncol., 6 221 (2016). http://dx.doi.org/10.3389/fonc.2016.00221 Google Scholar

72.

B. La Yun et al., “Does adding diffuse optical tomography to sonography improve differentiation between benign and malignant breast lesions? Observer performance study,” J. Ultrasound Med., 34 749 –757 (2015). http://dx.doi.org/10.7863/ultra.34.5.749 Google Scholar

73.

C. Xu et al., “Indocyanine green enhanced co-registered diffuse optical tomography and photoacoustic tomography,” J. Biomed. Opt., 18 126006 (2013). http://dx.doi.org/10.1117/1.JBO.18.12.126006 JBOPFO 1083-3668 Google Scholar

74.

P. Taroni, “Diffuse optical imaging and spectroscopy of the breast: a brief outline of history and perspectives,” Photochem. Photobiol. Sci., 11 241 –250 (2012). http://dx.doi.org/10.1039/C1PP05230F Google Scholar

75.

N. J. Pelc, “Recent and future directions in CT imaging,” Ann. Biomed. Eng., 42 260 –268 (2014). http://dx.doi.org/10.1007/s10439-014-0974-z ABMECF 0090-6964 Google Scholar

76.

L. Scolaro et al., “A review of optical coherence tomography in breast cancer,” Photonics Lasers Med., 3 (3), 225 –240 (2014). http://dx.doi.org/10.1515/plm-2014-0013 Google Scholar

77.

Z. Xie et al., “Combined photo-acoustic and acoustic imaging of human breast speciments in the mammographic geometry,” Ultrasound Med. Biol., 39 2176 –2184 (2013). http://dx.doi.org/10.1016/j.ultrasmedbio.2013.05.018 Google Scholar

78.

R. Longo et al., “Towards breast tomography with synchrotron radiation at Elettra: first images,” Phys. Med. Biol., 61 (4), 1634 –1649 (2016). http://dx.doi.org/10.1088/0031-9155/61/4/1634 PHMBA7 0031-9155 Google Scholar

79.

L. Xi et al., “Design and evaluation of a hybrid photoacoustic tomography and diffuse optical tomography system for breast cancer detection,” Med. Phys., 39 2584 –2594 (2012). http://dx.doi.org/10.1118/1.3703598 MPHYA6 0094-2405 Google Scholar

80.

N. Hartung et al., “Mathematical modeling of tumor growth and metastatic spreading: validation in tumor-bearing mice,” Cancer Res., 74 6397 –6407 (2014). http://dx.doi.org/10.1158/0008-5472.CAN-14-0721 CNREA8 0008-5472 Google Scholar

81.

S. Paddock, K. Eliceiri, “Laser scanning confocal microscopy: history, applications, and related optical sectioning techniques,” Confocal Microscopy, 9 –47 Springer, New York (2014). Google Scholar

82.

R. Schafer, H. M. Leung and A. F. Gmitro, “Multi-modality imaging of a murine mammary window chamber for breast cancer research,” BioTechniques, 57 45 –50 (2014). http://dx.doi.org/10.2144/000114191 BTNQDO 0736-6205 Google Scholar

83.

Z. Yang et al., “High specific activity is not optimal: 18f-fluoroestradio positron emission tomography-computed tomography results in a breast cancer xenograft,” J. Labelled Compd. Radiopharm., 59 576 –581 (2016). http://dx.doi.org/10.1002/jlcr.v59.13 JLCRD4 1099-1344 Google Scholar

84.

E. Kim et al., “Anti-vascular effects of the cytosolic phospholipase A2 inhibitor AVX235 in a patient-derived basal-like breast cancer model,” BMC Cancer, 16 (1), 191 (2016). http://dx.doi.org/10.1186/s12885-016-2225-1 BCMACL 1471-2407 Google Scholar

85.

D. Vonwil et al., “Validation of fluorescence molecular tomography/micro-CT multimodal imaging in vivo in rats,” Mol. Imaging Biol., 16 350 –361 (2014). http://dx.doi.org/10.1007/s11307-013-0698-8 Google Scholar

86.

J. P. Gambini et al., “Evaluation of 99mtc-glucarate as a breast cancer imaging agent in a xenograft animal model,” Nucl. Med. Biol., 38 255 –260 (2011). http://dx.doi.org/10.1016/j.nucmedbio.2010.08.002 Google Scholar

87.

M. D’Huyvetter et al., “Development of 177lu-nanobodies for radioimmunotherapy of HER2-positive breast cancer: evaluation of different bifunctional chelators,” Contrast Media Mol. Imaging, 7 254 –264 (2012). http://dx.doi.org/10.1002/cmmi.491 Google Scholar

88.

V. Neuschmelting et al., “Lymph node micrometastases and in-transit metastases from melanoma: in vivo detection with multispectral optoacoustic imaging in a mouse model,” Radiology, 280 137 –150 (2016). http://dx.doi.org/10.1148/radiol.2016160191 RADLAX 0033-8419 Google Scholar

89.

A. de Vries et al., “Quantitative spectral K-edge imaging in preclinical photon-counting x-ray computed tomography,” Invest. Radiol., 50 297 –304 (2015). http://dx.doi.org/10.1097/RLI.0000000000000126 INVRAV 0020-9996 Google Scholar

90.

Y. Cui et al., “Dual-targeting magnetic PLGA nanoparticles for codelivery of paclitaxel and curcumin for brain tumor therapy,” ACS Appl. Mater. Interfaces, 8 32159 –32169 (2016). http://dx.doi.org/10.1021/acsami.6b10175 Google Scholar

91.

P. Mohajerani et al., “FMT-PCCT: hybrid fluorescence molecular tomography-x-ray phase-contrast CT imaging of mouse models,” IEEE Trans. Med. Imaging, 33 1434 –1446 (2014). http://dx.doi.org/10.1109/TMI.2014.2313405 ITMID4 0278-0062 Google Scholar

92.

M. M. Pasternak et al., “High-frequency ultrasound analysis of post-mitotic arrest cell death,” Oncoscience, 3 109 –121 (2016). http://dx.doi.org/10.18632/oncoscience.301 Google Scholar

93.

A.-H. Liao et al., “Evaluation of 18f-labeled targeted perfluorocarbon-filled albumin microbubbles as a probe for microUS and microPET in tumor-bearing mice,” Ultrasonics, 53 320 –327 (2013). http://dx.doi.org/10.1016/j.ultras.2012.06.014 ULTRA3 0041-624X Google Scholar

94.

W. Zhu, Y. Kato and D. Artemov, “Heterogeneity of tumor vasculature and antiangiogenic intervention: insights from MR angiography and DCE-MRI,” PLoS One, 9 e86583 (2014). http://dx.doi.org/10.1371/journal.pone.0086583 POLNCL 1932-6203 Google Scholar

95.

M. Koch et al., “Threshold analysis and biodistribution of fluorescently labeled bevacizumab in human breast cancer,” Cancer Res., 77 (3), 623 –631 (2017). http://dx.doi.org/10.1158/0008-5472.CAN-16-1773 CNREA8 0008-5472 Google Scholar

96.

J. Y. Hwang et al., “A multimode optical imaging system for preclinical applications in vivo: technology development, multiscale imaging, and chemotherapy assessment,” Mol. Imaging Biol., 14 431 –442 (2012). http://dx.doi.org/10.1007/s11307-011-0517-z Google Scholar

97.

C. W. Huo et al., “High mammographic density is associated with an increase in stromal collagen and immune cells within the mammary epithelium,” Breast Cancer Res., 17 (1), 79 (2015). http://dx.doi.org/10.1186/s13058-015-0592-1 Google Scholar

98.

R. Patel et al., “Polarization-sensitive multimodal imaging for detecting breast cancer,” Cancer Res., 74 4685 –4693 (2014). http://dx.doi.org/10.1158/0008-5472.CAN-13-2411 CNREA8 0008-5472 Google Scholar

99.

A. Curatolo et al., “Ultrasound-guided optical coherence tomography needle probe for the assessment of breast cancer tumor margins,” Am. J. Roentgenol., 199 W520 –W522 (2012). http://dx.doi.org/10.2214/AJR.11.7284 Google Scholar

100.

H. Feng et al., “TYR as a multifunctional reporter gene regulated by the Tet-on system for multimodality imaging: an in vitro study,” Sci. Rep., 5 15502 (2015). http://dx.doi.org/10.1038/srep15502 Google Scholar

101.

F. S. Azar et al., “Standardized platform for coregistration of nonconcurrent diffuse optical and magnetic resonance breast images obtained in different geometries,” J. Biomed. Opt., 12 (5), 051902 (2007). http://dx.doi.org/10.1117/1.2798630 JBOPFO 1083-3668 Google Scholar

102.

Q. Fang et al., “Combined optical and x-ray tomosynthesis breast imaging,” Radiology, 258 89 –97 (2011). http://dx.doi.org/10.1148/radiol.10082176 RADLAX 0033-8419 Google Scholar

103.

“IVIS Instrument, Spectrum, 120v, Andor C-124262,” (2017). http://www.perkinelmer.com/product/ivis-instrument-spectrum-120v-andor-c-124262 Google Scholar

104.

P. M. Peiris et al., “Imaging metastasis using an integrin-targeting chain-shaped nanoparticle,” ACS Nano, 6 8783 –8795 (2012). http://dx.doi.org/10.1021/nn303833p ANCAC3 1936-0851 Google Scholar

105.

R. Zhang et al., “Annexin A5-conjugated polymeric micelles for dual SPECT and optical detection of apoptosis,” J. Nucl. Med., 52 958 –964 (2011). http://dx.doi.org/10.2967/jnumed.110.083220 Google Scholar

106.

V. P. Baklaushev et al., “Modeling and Integral x-ray, optical, and MRI visualization of multiorgan metastases of orthotopic 4t1 breast carcinoma in BALB/c mice,” Bull. Exp. Biol. Med., 158 581 –588 (2015). http://dx.doi.org/10.1007/s10517-015-2810-3 BEXBAN 0007-4888 Google Scholar

107.

W. M. Allen et al., “Wide-field optical coherence micro-elastography for intraoperative assessment of human breast cancer margins,” Biomed. Opt. Express, 7 4139 –4153 (2016). http://dx.doi.org/10.1364/BOE.7.004139 BOEICL 2156-7085 Google Scholar

108.

T. E. Doyle et al., “High-frequency ultrasound for intraoperative margin assessments in breast conservation surgery: a feasibility study,” BMC Cancer, 11 444 (2011). http://dx.doi.org/10.1186/1471-2407-11-444 BCMACL 1471-2407 Google Scholar

109.

S. J. Erickson-Bhatt et al., “In vivo intra-operative breast tumor margin detection using a portable OCT system with a handheld surgical imaging probe,” Proc. SPIE, 8935 89351C (2014). http://dx.doi.org/10.1117/12.2040315 PSISDG 0277-786X Google Scholar

110.

L. Hughes et al., “Surgeon volume, patient age, and tumor-related factors influence the need for re-excision after breast-conserving surgery,” Ann. Surg. Oncol., 23 656 –664 (2016). http://dx.doi.org/10.1245/s10434-016-5602-8 Google Scholar

111.

M. L. Huang et al., “Stereotactic breast biopsy: pitfalls and pearls,” Tech. Vasc. Interventional Radiol., 17 32 –39 (2014). http://dx.doi.org/10.1053/j.tvir.2013.12.006 Google Scholar

112.

K. Polyak, “Heterogeneity in breast cancer,” J. Clin. Invest., 121 3786 –3788 (2011). http://dx.doi.org/10.1172/JCI60534 JCINAO 0021-9738 Google Scholar

113.

S. D. Auweter et al., “X-ray phase-contrast imaging of the breast-advances towards clinical implementation,” Br. J. Radiol., 87 20130606 (2014). http://dx.doi.org/10.1259/bjr.20130606 Google Scholar

114.

A. Sarno, G. Mettivier and P. Russo, “Dedicated breast computed tomography: Basic aspects,” Med. Phys., 42 2786 –2804 (2015). http://dx.doi.org/10.1118/1.4919441 MPHYA6 0094-2405 Google Scholar

115.

D. Panetta, “Advances in x-ray detectors for clinical and preclinical computed tomography,” Nucl. Instrum. Methods Phys. Res., Sect. A, 809 2 –12 (2016). http://dx.doi.org/10.1016/j.nima.2015.10.034 Google Scholar

116.

L. E. Paulis et al., “Radiation exposure of digital breast tomosynthesis using an antiscatter grid compared with full-field digital mammography,” Invest. Radiol., 50 679 –685 (2015). http://dx.doi.org/10.1097/RLI.0000000000000168 INVRAV 0020-9996 Google Scholar

117.

L. Aranda-Lara et al., “Synthesis and evaluation of Lys¹(

α, γ

-Folate)Lys³(¹⁷⁷Lu-DOTA)-Bombesin(1–14) as a potential theranostic radiopharmaceutical for breast cancer,” Appl. Radiat. Isot., 107 214 –219 (2016). http://dx.doi.org/10.1016/j.apradiso.2015.10.030 ARISEF 0969-8043 Google Scholar

118.

Z. Ortiz-Arzate et al., “Kit preparation and biokinetics in women of 99mtc-EDDA/HYNIC-E-[c(RGDfK)]2 for breast cancer imaging,” Nucl. Med. Commun., 35 423 –432 (2014). http://dx.doi.org/10.1097/MNM.0000000000000065 NMCODC 0143-3636 Google Scholar

119.

S. He et al., “Comparison of 18f-FES, 18f-FDG, and 18f-FMISO PET imaging probes for early prediction and monitoring of response to endocrine therapy in a mouse xenograft model of ER-positive breast cancer,” PLoS One, 11 e0159916 (2016). http://dx.doi.org/10.1371/journal.pone.0159916 POLNCL 1932-6203 Google Scholar

120.

E. Kim et al., “Anti-angiogenic therapy affects the relationship between tumor vascular structure and function: a correlation study between microcomputed tomography angiography and dynamic contrast enhanced MRI,” Magn. Reson. Med., 78 (4), 1513 –1522 (2017). http://dx.doi.org/10.1002/mrm.26547 MRMEEN 0740-3194 Google Scholar

121.

K. McLarty et al., “Micro-SPECT/CT with 111in-DTPA-pertuzumab sensitively detects trastuzumab-mediated HER2 downregulation and tumor response in athymic mice bearing MDA-MB-361 human breast cancer xenografts,” J. Nucl. Med., 50 1340 –1348 (2009). http://dx.doi.org/10.2967/jnumed.109.062224 JNMEAQ 0161-5505 Google Scholar

122.

T. Pschinger et al., “Dynamic contrast-enhanced micro-computed tomography correlates with 3-dimensional fluorescence ultramicroscopy in antiangiogenic therapy of breast cancer xenografts,” Invest. Radiol., 49 445 –456 (2014). http://dx.doi.org/10.1097/RLI.0000000000000038 INVRAV 0020-9996 Google Scholar

123.

H. Gufler et al., “Fine structure of breast tissue on micro computed tomography: a feasibility study,” Acad. Radiol., 18 230 –234 (2011). http://dx.doi.org/10.1016/j.acra.2010.10.006 Google Scholar

124.

R. Tang et al., “Micro-computed tomography (micro-CT): a novel approach for intraoperative breast cancer specimen imaging,” Breast Cancer Res. Treat., 139 311 –316 (2013). http://dx.doi.org/10.1007/s10549-013-2554-6 BCTRD6 Google Scholar

125.

E. Kim et al., “Assessing breast cancer angiogenesis in vivo: which susceptibility contrast MRI biomarkers are relevant?,” Magn. Reson. Med., 70 1106 –1116 (2013). http://dx.doi.org/10.1002/mrm.24530 MRMEEN 0740-3194 Google Scholar

126.

E. L. Ritman, “Current status of developments and applications of micro-CT,” Annu. Rev. Biomed. Eng., 13 (1), 531 –552 (2011). http://dx.doi.org/10.1146/annurev-bioeng-071910-124717 ARBEF7 1523-9829 Google Scholar

127.

C. L. Gregg, A. K. Recknagel and J. T. Butcher, “Micro/nano-computed tomography technology for quantitative dynamic, multi-scale imaging of morphogenesis,” Methods Mol. Biol., 1189 47 –61 (2015). http://dx.doi.org/10.1007/978-1-4939-1164-6 Google Scholar

128.

A. Bravin, P. Coan and P. Suortti, “X-ray phase-contrast imaging: from pre-clinical applications towards clinics,” Phys. Med. Biol., 58 (1), R1 (2013). http://dx.doi.org/10.1088/0031-9155/58/1/R1 PHMBA7 0031-9155 Google Scholar

129.

Y. I. Nesterets et al., “A feasibility study of x-ray phase-contrast mammographic tomography at the imaging and medical beamline of the Australian synchrotron,” J. Synchrotron Radiat., 22 1509 –1523 (2015). http://dx.doi.org/10.1107/S160057751501766X JSYRES 0909-0495 Google Scholar

130.

S. Grandl et al., “Evaluation of phase-contrast CT of breast tissue at conventional x-ray sources presentation of selected findings,” Z. Med. Phys., 23 212 –221 (2013). http://dx.doi.org/10.1016/j.zemedi.2013.02.005 Google Scholar

131.

S. Grandl et al., “Visualizing typical features of breast fibroadenomas using phase-contrast CT: an ex-vivo study,” PLoS One, 9 (5), e97101 (2014). http://dx.doi.org/10.1371/journal.pone.0097101 POLNCL 1932-6203 Google Scholar

132.

S. Grandl et al., “Detection of post-therapeutic effects in breast carcinoma using hard x-ray index of refraction computed tomography a feasibility study,” PLoS One, 11 (6), e0158306 (2016). http://dx.doi.org/10.1371/journal.pone.0158306 POLNCL 1932-6203 Google Scholar

133.

M. Denis et al., “Correlating tumor stiffness with immunohistochemical subtypes of breast cancers: prognostic value of comb-push ultrasound shear elastography for differentiating luminal subtypes,” PLoS One, 11 (10), e0165003 (2016). http://dx.doi.org/10.1371/journal.pone.0165003 POLNCL 1932-6203 Google Scholar

134.

J. Melnikow et al., “Supplemental screening for breast cancer in women with dense breasts: a systematic review for the U.S. preventive services task forcesupplemental breast cancer screening in women with dense breasts,” Ann. Intern. Med., 164 (4), 268 –278 (2016). http://dx.doi.org/10.7326/M15-1789 AIMEAS 0003-4819 Google Scholar

135.

F. L. Lizzi et al., “Relationship of ultrasonic spectral parameters to features of tissue microstructure,” IEEE Trans. Ultrason. Ferroelectr. Freq. Control, 34 (3), 319 –329 (1987). http://dx.doi.org/10.1109/T-UFFC.1987.26950 Google Scholar

136.

M. Insana and T. Hall, “Characterizing the microstructure of random-media using ultrasound,” Phys. Med. Biol., 35 1373 –1386 (1990). http://dx.doi.org/10.1088/0031-9155/35/10/003 PHYSGM 1943-2879 Google Scholar

137.

I. M. Rosado-Mendez et al., “Analysis of coherent and diffuse scattering using a reference phantom,” IEEE Trans. Ultrason. Ferroelectr. Freq. Control, 63 (9), 1306 –1320 (2016). http://dx.doi.org/10.1109/TUFFC.2016.2547341 Google Scholar

138.

T. Shiina et al., “WFUMB guidelines and recommendations for clinical use of ultrasound elastography: Part 1: basic principles and terminology,” Ultrasound Med. Biol., 41 (5), 1126 –1147 (2015). http://dx.doi.org/10.1016/j.ultrasmedbio.2015.03.009 USMBA3 0301-5629 Google Scholar

139.

R. G. Barr et al., “Wfumb guidelines and recommendations for clinical use of ultrasound elastography: Part 2: breast,” Ultrasound Med. Biol., 41 (5), 1148 –1160 (2015). http://dx.doi.org/10.1016/j.ultrasmedbio.2015.03.008 Google Scholar

140.

M. F. Insana and T. J. Hall, “Parametric ultrasound imaging from backscatter coefficient measurements: image formation and interpretation,” Ultrasonic Imaging, 12 245 –267 (1990). http://dx.doi.org/10.1177/016173469001200402 ULIMD4 0161-7346 Google Scholar

141.

Q. W. Guerrero et al., “Quantifying backscatter anisotropy using the reference phantom method,” IEEE Trans. Ultrason. Ferroelectr. Freq. Control, 64 1063 –1077 (2017). http://dx.doi.org/10.1109/TUFFC.2017.2698832 Google Scholar

142.

K. Nam et al., “Cross-imaging system comparison of backscatter coefficient estimates from a tissue-mimicking material,” J. Acoust. Soc. Am., 132 1319 –1324 (2012). http://dx.doi.org/10.1121/1.4742725 JASMAN 0001-4966 Google Scholar

143.

K. Nam, J. A. Zagzebski and T. J. Hall, “Quantitative assessment of in vivo breast masses using ultrasound attenuation and backscatter,” Ultrasonic Imaging, 35 (2), 146 –161 (2013). http://dx.doi.org/10.1177/0161734613480281 ULIMD4 0161-7346 Google Scholar

144.

H. G. Nasief et al., “Acoustic properties of breast fat,” J. Ultrasound Med., 34 (11), 2007 –2016 (2015). http://dx.doi.org/10.7863/ultra.14.07039 JUMEDA 0278-4297 Google Scholar

145.

N. Duric et al., “Breast density measurements with ultrasound tomography: a comparison with film and digital mammography,” Med. Phys., 40 013501 (2013). http://dx.doi.org/10.1118/1.4772057 MPHYA6 0094-2405 Google Scholar

146.

L. V. Wang and J. Yao, “A practical guide to photoacoustic tomography in the life sciences,” Nat. Methods, 13 627 –638 (2016). http://dx.doi.org/10.1038/nmeth.3925 1548-7091 Google Scholar

147.

M. Andr, J. Wiskin, D. Borup, “Clinical results with ultrasound computed tomography of the breast,” Quantitative Ultrasound in Soft Tissues, 395 –432 Springer, Netherlands (2013). Google Scholar

148.

Y. Wang et al., “Three-dimensional ultrasound elasticity imaging on an automated breast volume scanning system,” Ultrasonic Imaging, 39 (6), 369 –392 (2017). http://dx.doi.org/10.1177/0161734617712238 ULIMD4 0161-7346 Google Scholar

149.

M. L. Oelze et al., “Differentiation and characterization of rat mammary fibroadenomas and 4t1 mouse carcinomas using quantitative ultrasound imaging,” IEEE Trans. Med. Imaging, 23 (6), 764 –771 (2004). http://dx.doi.org/10.1109/TMI.2004.826953 ITMID4 0278-0062 Google Scholar

150.

Q. Zhou et al., “PMN-PT single crystal, high-frequency ultrasonic needle transducers for pulsed-wave Doppler application,” IEEE Trans. Ultrason. Ferroelectr. Freq. Control, 54 668 –675 (2007). http://dx.doi.org/10.1109/TUFFC.2007.290 Google Scholar

151.

R. Chen et al., “In vivo sonothrombolysis of ear marginal vein of rabbits monitored with high-frequency ultrasound needle transducer,” J. Med. Biol. Eng., 33 (1), 103 –110 (2013). http://dx.doi.org/10.5405/jmbe.1219 Google Scholar

152.

H. K. Chiang et al., “Eyes in the needle: novel epidural needle with embedded high-frequency ultrasound transducer-epidural access in porcine model,” Anesthesiology, 114 1320 –1324 (2011). http://dx.doi.org/10.1097/ALN.0b013e31821b5746 ANESAV 0003-3022 Google Scholar

153.

Z. Torbatian et al., “Listening to the cochlea with high-frequency ultrasound,” Ultrasound Med. Biol., 38 2208 –2217 (2012). http://dx.doi.org/10.1016/j.ultrasmedbio.2012.07.015 Google Scholar

154.

L. M. Cannon, A. J. Fagan and J. E. Browne, “Novel tissue mimicking materials for high frequency breast ultrasound phantoms,” Ultrasound Med. Biol., 37 122 –135 (2011). http://dx.doi.org/10.1016/j.ultrasmedbio.2010.10.005 Google Scholar

155.

A. Sadeghi-Naini et al., “Low-frequency quantitative ultrasound imaging of cell death in vivo,” Med. Phys., 40 082901 (2013). http://dx.doi.org/10.1118/1.4812683 MPHYA6 0094-2405 Google Scholar

156.

J. Mamou et al., “Three-dimensional quantitative ultrasound to guide pathologists towards metastatic foci in lymph nodes,” in Annual Int. Conf. of the IEEE Engineering in Medicine and Biology Society, 1114 –1117 (2012). http://dx.doi.org/10.1109/EMBC.2012.6346130 Google Scholar

157.

M. F. Insana and T. J. Hall, “Characterising the microstructure of random media using ultrasound,” Phys. Med. Biol., 35 1373 –1386 (1990). http://dx.doi.org/10.1088/0031-9155/35/10/003 PHMBA7 0031-9155 Google Scholar

158.

S. Gefen et al., “Roc analysis of ultrasound tissue characterization classifiers for breast cancer diagnosis,” IEEE Trans. Med. Imaging, 22 (2), 170 –177 (2003). http://dx.doi.org/10.1109/TMI.2002.808361 ITMID4 0278-0062 Google Scholar

159.

L. M. Reusch et al., “Nonlinear optical microscopy and ultrasound imaging of human cervical structure,” J. Biomed. Opt., 18 031110 (2013). http://dx.doi.org/10.1117/1.JBO.18.3.031110 JBOPFO 1083-3668 Google Scholar

160.

J. H. Hipwell et al., “A review of biomechanically informed breast image registration,” Phys. Med. Biol., 61 (2), R1 (2016). http://dx.doi.org/10.1088/0031-9155/61/2/R1 PHMBA7 0031-9155 Google Scholar

161.

P. J. Bolan, “Magnetic resonance spectroscopy of the breast: current status,” Magn. Reson. Imaging Clin. North Am., 21 625 –639 (2013). http://dx.doi.org/10.1016/j.mric.2013.04.008 Google Scholar

162.

H. Rahbar and S. C. Partridge, “Multiparametric MR imaging of breast cancer,” Magn. Reson. Imaging Clin. North Am., 24 223 –238 (2016). http://dx.doi.org/10.1016/j.mric.2015.08.012 Google Scholar

163.

J. M. Winfield et al., “DCE-MRI, DW-MRI, and MRS in cancer: challenges and advantages of implementing qualitative and quantitative multi-parametric imaging in the clinic,” Top. Magn. Reson. Imaging, 25 245 –254 (2016). http://dx.doi.org/10.1097/RMR.0000000000000103 TMRIEY 0899-3459 Google Scholar

164.

Z. H. Cho et al., “Nuclear magnetic resonance microscopy with 4- m resolution: theoretical study and experimental results,” Med. Phys., 15 815 –824 (1988). http://dx.doi.org/10.1118/1.596287 MPHYA6 0094-2405 Google Scholar

165.

S. A. Jansen et al., “Detection of in situ mammary cancer in a transgenic mouse model: in vitro and in vivo MRI studies demonstrate histopathologic correlation,” Phys. Med. Biol., 53 5481 –5493 (2008). http://dx.doi.org/10.1088/0031-9155/53/19/014 PHMBA7 0031-9155 Google Scholar

166.

N. Weiskopf et al., “Quantitative multi-parameter mapping of R1, PD*, MT, and R2* at 3t: a multi-center validation,” Front. Neurosci., 7 95 (2013). http://dx.doi.org/10.3389/fnins.2013.00095 1662-453X Google Scholar

167.

M. N. Gurcan et al., “Histopathological image analysis: a review,” IEEE Rev. Biomed. Eng., 2 147 –171 (2009). http://dx.doi.org/10.1109/RBME.2009.2034865 Google Scholar

168.

C. Krafft, “Modern trends in biophotonics for clinical diagnosis and therapy to solve unmet clinical needs,” J. Biophotonics, 9 1362 –1375 (2016). http://dx.doi.org/10.1002/jbio.201600290 Google Scholar

169.

F. Ghaznavi et al., “Digital imaging in pathology: whole-slide imaging and beyond,” Ann. Rev. Pathol.: Mech. Dis., 8 331 –359 (2013). http://dx.doi.org/10.1146/annurev-pathol-011811-120902 Google Scholar

170.

C. Magliaro et al., “HisTOOLogy: an open-source tool for quantitative analysis of histological sections,” J. Microsc., 260 260 –267 (2015). http://dx.doi.org/10.1111/jmi.12292 JMICAR 0022-2720 Google Scholar

171.

S. Al-Janabi, A. Huisman and P. J. Van Diest, “Digital pathology: current status and future perspectives,” Histopathology, 61 1 –9 (2012). http://dx.doi.org/10.1111/his.2012.61.issue-1 HISTDD 1365-2559 Google Scholar

172.

L. A. D. Cooper et al., “Digital pathology: data-intensive frontier in medical imaging,” Proc. IEEE, 100 991 –1003 (2012). http://dx.doi.org/10.1109/JPROC.2011.2182074 IEEPAD 0018-9219 Google Scholar

173.

S. Alowami et al., “Mammographic density is related to stroma and stromal proteoglycan expression,” Breast Cancer Res., 5 (5), R129 –R135 (2003). http://dx.doi.org/10.1186/bcr622 Google Scholar

174.

M. Gabrielson et al., “Amount of stroma is associated with mammographic density and stromal expression of oestrogen receptor in normal breast tissues,” Breast Cancer Res. Treat., 158 253 –261 (2016). http://dx.doi.org/10.1007/s10549-016-3877-x BCTRD6 Google Scholar

175.

Y.-P. Guo et al., “Growth factors and stromal matrix proteins associated with mammographic densities,” Cancer Epidemiol. Biomarkers Prev., 10 243 –248 (2001). Google Scholar

176.

J. A. Harvey et al., “Histologic changes in the breast with menopausal hormone therapy use: correlation with breast density, estrogen receptor, progesterone receptor, and proliferation indices,” Menopause, 15 (1), 67 –73 (2008). http://dx.doi.org/10.1097/gme.0b013e318054e29a Google Scholar

177.

Q. J. Khan et al., “Mammographic density does not correlate with Ki-67 expression or cytomorphology in benign breast cells obtained by random periareolar fine needle aspiration from women at high risk for breast cancer,” Breast Cancer Res., 9 R35 (2007). http://dx.doi.org/10.1186/bcr1683 BCTRD6 Google Scholar

178.

X. Sun et al., “Relationship of mammographic density and gene expression: analysis of normal breast tissue surrounding breast cancer,” Clin. Cancer Res., 19 4972 –4982 (2013). http://dx.doi.org/10.1158/1078-0432.CCR-13-0029 Google Scholar

179.

M. Verheus et al., “Mammographic density and epithelial histopathologic markers,” BMC Cancer, 9 182 (2009). http://dx.doi.org/10.1186/1471-2407-9-182 BCMACL 1471-2407 Google Scholar

180.

F. Yang et al., “High-resolution mesoscopic fluorescence molecular tomography based on compressive sensing,” IEEE Trans. Bio-Med. Eng., 62 248 –255 (2015). http://dx.doi.org/10.1109/TBME.2014.2347284 Google Scholar

181.

K. Ghosh et al., “Tissue composition of mammographically dense and non-dense breast tissue,” Breast Cancer Res. Treat., 131 267 –275 (2012). http://dx.doi.org/10.1007/s10549-011-1727-4 BCTRD6 Google Scholar

182.

T. Li et al., “The association of measured breast tissue characteristics with mammographic density and other risk factors for breast cancer,” Cancer Epidemiol. Biomarkers Prev., 14 343 –349 (2005). http://dx.doi.org/10.1158/1055-9965.EPI-04-0490 Google Scholar

183.

S. J. Lin et al., “Image-guided sampling reveals increased stroma and lower glandular complexity in mammographically dense breast tissue,” Breast Cancer Res. Treat., 128 505 –516 (2011). http://dx.doi.org/10.1007/s10549-011-1346-0 BCTRD6 Google Scholar

184.

G. Turashvili et al., “Columnar cell lesions, mammographic density and breast cancer risk,” Breast Cancer Res. Treat., 115 561 –571 (2009). http://dx.doi.org/10.1007/s10549-008-0099-x BCTRD6 Google Scholar

185.

C. M. Vachon et al., “Aromatase immunoreactivity is increased in mammographically dense regions of the breast,” Breast Cancer Res. Treat., 125 243 –252 (2011). http://dx.doi.org/10.1007/s10549-010-0944-6 BCTRD6 Google Scholar

186.

N. F. Boyd, “Mammographic density and risk of breast cancer,” in American Society of Clinical Oncology Educational Book / ASCO. American Society of Clinical Oncology. Meeting, (2013). Google Scholar

187.

N. F. Boyd et al., “Mammographic density and the risk and detection of breast cancer,” N. Engl. J. Med., 356 227 –236 (2007). http://dx.doi.org/10.1056/NEJMoa062790 NEJMBH Google Scholar

188.

A. Pettersson et al., “Mammographic density phenotypes and risk of breast cancer: a meta-analysis,” J. Natl. Cancer Inst., 106 (5), (2014). http://dx.doi.org/10.1093/jnci/dju078 JNCIEQ Google Scholar

189.

Y. Mao et al., “Stroma cells in tumor microenvironment and breast cancer,” Cancer Metastasis Rev., 32 303 –315 (2013). http://dx.doi.org/10.1007/s10555-012-9415-3 Google Scholar

190.

H. Li et al., “Comparative analysis of image-based phenotypes of mammographic density and parenchymal patterns in distinguishing between BRCA1/2 cases, unilateral cancer cases, and controls,” J. Med. Imaging, 1 (3), 031009 (2014). http://dx.doi.org/10.1117/1.JMI.1.3.031009 JMEIET 0920-5497 Google Scholar

191.

K. Glunde, A. P. Pathak and Z. M. Bhujwalla, “Molecularfunctional imaging of cancer: to image and imagine,” Trends Mol. Med., 13 287 –297 (2007). http://dx.doi.org/10.1016/j.molmed.2007.05.002 Google Scholar

192.

N. Farahani and C. E. Monteith, “The coming paradigm shift: a transition from manual to automated microscopy,” J. Pathol. Inf., 7 35 (2016). http://dx.doi.org/10.4103/2153-3539.189698 Google Scholar

193.

S. Chatterjee, “Artefacts in histopathology,” J. Oral Maxillofac. Pathol., 18 S111 –S116 (2014). http://dx.doi.org/10.4103/0973-029X.141346 Google Scholar

194.

A. Kumar, “Fluorescence lifetime-based optical molecular imaging,” Molecular Imaging, 165 –180 Humana Press, Totowa, New Jersey (2011). Google Scholar

195.

E. E. Hoover and J. A. Squier, “Advances in multiphoton microscopy technology,” Nat. Photonics, 7 93 –101 (2013). http://dx.doi.org/10.1038/nphoton.2012.361 NPAHBY 1749-4885 Google Scholar

196.

W. J. Alford, R. D. VanderNeut and V. J. Zaleckas, “Laser scanning microscopy,” Proc. IEEE, 70 641 –651 (1982). http://dx.doi.org/10.1109/PROC.1982.12362 IEEPAD 0018-9219 Google Scholar

197.

W.-L. Chen et al., “Quantitative analysis of multiphoton excitation autofluorescence and second harmonic generation imaging for medical diagnosis,” Comput. Med. Imaging Graphics, 36 519 –526 (2012). http://dx.doi.org/10.1016/j.compmedimag.2012.06.003 CMIGEY 0895-6111 Google Scholar

198.

G. Yellen and R. Mongeon, “Quantitative two-photon imaging of fluorescent biosensors,” Curr. Opin. Chem. Biol., 27 24 –30 (2015). http://dx.doi.org/10.1016/j.cbpa.2015.05.024 COCBF4 1367-5931 Google Scholar

199.

M. Allinen et al., “Molecular characterization of the tumor microenvironment in breast cancer,” Cancer Cell, 6 17 –32 (2004). http://dx.doi.org/10.1016/j.ccr.2004.06.010 Google Scholar

200.

J. A. Joyce and J. W. Pollard, “Microenvironmental regulation of metastasis,” Nat. Rev. Cancer, 9 239 –252 (2009). http://dx.doi.org/10.1038/nrc2618 NRCAC4 1474-175X Google Scholar

201.

T. L. Whiteside, “The tumor microenvironment and its role in promoting tumor growth,” Oncogene, 27 (45), 5904 –5912 (2008). http://dx.doi.org/10.1038/onc.2008.271 ONCNES 0950-9232 Google Scholar

202.

M. W. Conklin et al., “Aligned collagen is a prognostic signature for survival in human breast carcinoma,” Am. J. Pathol., 178 1221 –1232 (2011). http://dx.doi.org/10.1016/j.ajpath.2010.11.076 AJPAA4 0002-9440 Google Scholar

203.

J. T. Keyes et al., “Design and demonstration of a microbiaxial optomechanical device for multiscale characterization of soft biological tissues with two-photon microscopy,” Microsc. Microanal., 17 167 –175 (2011). http://dx.doi.org/10.1017/S1431927610094341 MIMIF7 1431-9276 Google Scholar

204.

J. M. Szulczewski et al., “In vivo visualization of stromal macrophages via label-free FLIM-based metabolite imaging,” Sci. Rep., 6 25086 (2016). http://dx.doi.org/10.1038/srep25086 Google Scholar

205.

L. Bonapace et al., “If you dont look, you wont see: intravital multiphoton imaging of primary and metastatic breast cancer,” J. Mammary Gland Biol. Neoplasia, 17 125 –129 (2012). http://dx.doi.org/10.1007/s10911-012-9250-8 Google Scholar

206.

A. J. Walsh et al., “Collagen density and alignment in responsive and resistant trastuzumab-treated breast cancer xenografts,” J. Biomed. Opt., 20 026004 (2015). http://dx.doi.org/10.1117/1.JBO.20.2.026004 JBOPFO 1083-3668 Google Scholar

207.

J. Wyckoff et al., “High-resolution multiphoton imaging of tumors in vivo,” Cold Spring Harbor Protocols, 2011 1167 –1184 (2011). http://dx.doi.org/10.1101/pdb.prot065912 Google Scholar

208.

M. Balu et al., “Rapid mesoscale multiphoton microscopy of human skin,” Biomed. Opt. Express, 7 4375 –4387 (2016). http://dx.doi.org/10.1364/BOE.7.004375 BOEICL 2156-7085 Google Scholar

209.

J. Williams and P. Campagnola, “Wearable second harmonic generation imaging: the sarcomeric bridge to the clinic,” Neuron, 88 1067 –1069 (2015). http://dx.doi.org/10.1016/j.neuron.2015.12.009 NERNET 0896-6273 Google Scholar

210.

N. Ghosh and I. A. Vitkin, “Tissue polarimetry: concepts, challenges, applications, and outlook,” J. Biomed. Opt., 16 (11), 110801 (2011). http://dx.doi.org/10.1117/1.3652896 JBOPFO 1083-3668 Google Scholar

211.

J. M. Jabbour et al., “Fluorescence lifetime imaging and reflectance confocal microscopy for multiscale imaging of oral precancer,” J. Biomed. Opt., 18 (4), 046012 (2013). http://dx.doi.org/10.1117/1.JBO.18.4.046012 JBOPFO 1083-3668 Google Scholar

212.

S. Srinivasan et al., “Image guided near-infrared spectroscopy of breast tissue in vivo using boundary element method,” J. Biomed. Opt., 15 (6), 061703 (2010). http://dx.doi.org/10.1117/1.3499419 JBOPFO 1083-3668 Google Scholar

213.

R. Patel et al., “Multimodal optical imaging for detecting breast cancer,” J. Biomed. Opt., 17 066008 (2012). http://dx.doi.org/10.1117/1.JBO.17.6.066008 JBOPFO 1083-3668 Google Scholar

214.

A. Greenbaum et al., “Wide-field computational imaging of pathology slides using lens-free on-chip microscopy,” Sci. Transl. Med., 6 267ra175 (2014). http://dx.doi.org/10.1126/scitranslmed.3009850 Google Scholar

215.

Y. Zhang et al., “Wide-field pathology imaging using on-chip microscopy,” Virchows Arch., 467 3 –7 (2015). http://dx.doi.org/10.1007/s00428-015-1782-z Google Scholar

216.

D. Grosenick et al., “Review of optical breast imaging and spectroscopy,” J. Biomed. Opt., 21 (9), 091311 (2016). http://dx.doi.org/10.1117/1.JBO.21.9.091311 JBOPFO 1083-3668 Google Scholar

217.

B. J. Vakoc et al., “Cancer imaging by optical coherence tomography: preclinical progress and clinical potential,” Nat. Rev. Cancer, 12 363 –368 (2012). http://dx.doi.org/10.1038/nrc3235 NRCAC4 1474-175X Google Scholar

218.

K. M. Kennedy et al., “Needle optical coherence elastography for the measurement of microscale mechanical contrast deep within human breast tissues,” J. Biomed. Opt., 18 121510 (2013). http://dx.doi.org/10.1117/1.JBO.18.12.121510 JBOPFO 1083-3668 Google Scholar

219.

R. A. McLaughlin et al., “Parametric imaging of cancer with optical coherence tomography,” J. Biomed. Opt., 15 046029 (2010). http://dx.doi.org/10.1117/1.3479931 JBOPFO 1083-3668 Google Scholar

220.

J. F. De Boer et al., “Polarization effects in optical coherence tomography of various biological tissues,” IEEE J. Sel. Top. Quantum Electron., 5 1200 –1204 (1999). http://dx.doi.org/10.1109/2944.796347 IJSQEN 1077-260X Google Scholar

221.

K. V. Larin and D. D. Sampson, “Optical coherence elastography OCT at work in tissue biomechanics,” Biomed. Opt. Express, 8 1172 –1202 (2017). http://dx.doi.org/10.1364/BOE.8.001172 BOEICL 2156-7085 Google Scholar

222.

R. M. Nolan et al., “Intraoperative optical coherence tomography for assessing human lymph nodes for metastatic cancer,” BMC Cancer, 16 (1), 144 (2016). http://dx.doi.org/10.1186/s12885-016-2194-4 BCMACL 1471-2407 Google Scholar

223.

C.-P. Liang et al., “Concurrent multiscale imaging with magnetic resonance imaging and optical coherence tomography,” J. Biomed. Opt., 18 046015 (2013). http://dx.doi.org/10.1117/1.JBO.18.4.046015 JBOPFO 1083-3668 Google Scholar

224.

J. Yao, J. Xia and L. V. Wang, “Multi-scale functional and molecular photoacoustic tomography,” Ultrason. Imaging, 38 44 –62 (2016). http://dx.doi.org/10.1177/0161734615584312 ULIMD4 0161-7346 Google Scholar

225.

J. Kang et al., “Photoacoustic imaging of breast microcalcifications: a validation study with 3-dimensional ex vivo data and spectrophotometric measurement,” J. Biophotonics, 8 71 –80 (2015). http://dx.doi.org/10.1002/jbio.201300100 Google Scholar

226.

W. J. Akers et al., “Multimodal sentinel lymph node mapping with SPECT/CT and photoacoustic tomography,” Transl. Res., 159 175 –181 (2012). http://dx.doi.org/10.1016/j.trsl.2011.09.006 Google Scholar

227.

C. Qin et al., “Tyrosinase as a multifunctional reporter gene for Photoacoustic/MRI/PET triple modality molecular imaging,” Sci. Rep., 3 1490 (2013). http://dx.doi.org/10.1038/srep01490 Google Scholar

228.

S. L. Jacques, “Optical properties of biological tissues: a review,” Phys. Med. Biol., 58 (11), R37 (2013). http://dx.doi.org/10.1088/0031-9155/58/11/R37 PHMBA7 0031-9155 Google Scholar

229.

B. Brooksby et al., “Imaging breast adipose and fibroglandular tissue molecular signatures by using hybrid MRI-guided near-infrared spectral tomography,” Proc. Natl. Acad. Sci. U. S. A., 103 8828 –8833 (2006). http://dx.doi.org/10.1073/pnas.0509636103 Google Scholar

230.

Q. Zhu et al., “Ultrasound-guided optical tomographic imaging of malignant and benign breast lesions: initial clinical results of 19 cases,” Neoplasia, 5 379 –388 (2003). http://dx.doi.org/10.1016/S1476-5586(03)80040-4 Google Scholar

231.

Q. Zhu et al., “Early-stage invasive breast cancers: potential role of optical tomography with us localization in assisting diagnosis,” Radiology, 256 367 –378 (2010). http://dx.doi.org/10.1148/radiol.10091237 RADLAX 0033-8419 Google Scholar

232.

J. F. Arevalo and J. V. Espinoza, “Indocyanine green mediated photothrombosis and high dose intravitreal bevacizumab as adjuvant therapy for isolated choroidal metastasis from breast cancer,” J. Ophthalmic Vision Res., 7 332 –340 (2012). Google Scholar

233.

S. Mollard et al., “In vivo bioluminescence tomography for monitoring breast tumor growth and metastatic spreading: comparative study and mathematical modeling,” Sci. Rep., 6 36173 (2016). http://dx.doi.org/10.1038/srep36173 Google Scholar

234.

M. Keyaerts, V. Caveliers and T. Lahoutte, “Bioluminescence imaging: looking beyond the light,” Trends Mol. Med., 18 164 –172 (2012). http://dx.doi.org/10.1016/j.molmed.2012.01.005 Google Scholar

235.

V. Ntziachristos et al., “Fluorescence molecular tomography resolves protease activity in vivo,” Nat. Med., 8 757 –761 (2002). http://dx.doi.org/10.1038/nm729 1078-8956 Google Scholar

236.

S. Yuan et al., “Three-dimensional coregistered optical coherence tomography and line-scanning fluorescence laminar optical tomography,” Opt. Lett., 34 1615 –1617 (2009). http://dx.doi.org/10.1364/OL.34.001615 OPLEDP 0146-9592 Google Scholar

237.

J. W. Bohland, “Toward a multimodal, multiscale understanding of white matter abnormalities in autism spectrum disorder,” Biol. Psychiatry, 79 e47 –e48 (2016). http://dx.doi.org/10.1016/j.biopsych.2016.02.020 BIPCBF 0006-3223 Google Scholar

238.

S. R. Kantelhardt et al., “In vivo multiphoton tomography and fluorescence lifetime imaging of human brain tumor tissue,” J. Neuro-Oncol., 127 473 –482 (2016). http://dx.doi.org/10.1007/s11060-016-2062-8 JNODD2 0167-594X Google Scholar

239.

Y. Yamada et al., “Chronic multiscale imaging of neuronal activity in the awake common marmoset,” Sci. Rep., 6 35722 (2016). http://dx.doi.org/10.1038/srep35722 Google Scholar

240.

X. L. Deán-Ben et al., “Advanced optoacoustic methods for multiscale imaging of in vivo dynamics,” Chem. Soc. Rev., 46 2158 –2198 (2017). http://dx.doi.org/10.1039/C6CS00765A CSRVBR 0306-0012 Google Scholar

241.

F. Ostadhossein and D. Pan, “Functional carbon nanodots for multiscale imaging and therapy,” Wiley Interdiscip. Rev. Nanomed. Nanobiotechnol., 9 (3), e1436 (2017). http://dx.doi.org/10.1002/wnan.1436 Google Scholar

242.

A. A. Poundarik and D. Vashishth, “Multiscale imaging of bone microdamage,” Connect. Tissue Res., 56 87 –98 (2015). http://dx.doi.org/10.3109/03008207.2015.1008133 Google Scholar

243.

D. Rousseau et al., “Multiscale imaging of plants: current approaches and challenges,” Plant Methods, 11 (1), 6 (2015). http://dx.doi.org/10.1186/s13007-015-0050-1 Google Scholar

244.

X. Wu et al., “Upconversion nanoparticles: a versatile solution to multiscale biological imaging,” Bioconjugate Chem., 26 166 –175 (2015). http://dx.doi.org/10.1021/bc5003967 BCCHES 1043-1802 Google Scholar

245.

J. P. B. O’Connor et al., “Imaging biomarker roadmap for cancer studies,” Nat. Rev. Clin. Oncol., 14 (3), 169 –186 (2016). http://dx.doi.org/10.1038/nrclinonc.2016.162 Google Scholar

246.

S. M. Brady et al., “Oncological image analysis,” Med. Image Anal., 33 7 –12 (2016). http://dx.doi.org/10.1016/j.media.2016.06.012 Google Scholar

247.

T. M. Deserno (n Lehmann) et al., “Viewpoints on medical image processing: from science to application,” Curr. Med. Imaging Rev., 9 79 –88 (2013). http://dx.doi.org/10.2174/1573405611309020002 Google Scholar

248.

I. D. Dinov, “Methodological challenges and analytic opportunities for modeling and interpreting big healthcare data,” Gigascience, 5 (1), 12 (2016). http://dx.doi.org/10.1186/s13742-016-0117-6 Google Scholar

249.

P. Guerra et al., “Real time signal processing and data handling with dedicated hardware in handheld OCT device,” J. Instrum., 10 (11), C11001 (2015). http://dx.doi.org/10.1088/1748-0221/10/11/C11001 JIONAS 1748-0221 Google Scholar

250.

J. Weese and C. Lorenz, “Four challenges in medical image analysis from an industrial perspective,” Med. Image Anal., 33 44 –49 (2016). http://dx.doi.org/10.1016/j.media.2016.06.023 Google Scholar

251.

BEST (Biomarkers, EndpointS, and other Tools) Resource, Food and Drug Administration (US)/National Institutes of Health (US), Maryland (2016). Google Scholar

252.

H. Sartor et al., “Measuring mammographic density: comparing a fully automated volumetric assessment versus European radiologists qualitative classification,” Eur. Radiol., 26 (12), 4354 –4360 (2016). http://dx.doi.org/10.1007/s00330-016-4309-3 EURAE3 1432-1084 Google Scholar

253.

H.-J. Eom et al., “Comparison of variability in breast density assessment by BI-RADS category according to the level of experience,” Acta Radiol., (2017). http://dx.doi.org/10.1177/0284185117725369 Google Scholar

254.

C. C. Gard et al., “Misclassification of breast imaging reporting and data system (BI-RADS) mammographic density and implications for breast density reporting legislation,” Breast J., 21 481 –489 (2015). http://dx.doi.org/10.1111/tbj.12443 BRJOFK 1075-122X Google Scholar

255.

S. A. Boppart et al., “Optical coherence tomography: feasibility for basic research and image-guided surgery of breast cancer,” Breast Cancer Res. Treat., 84 85 –97 (2004). http://dx.doi.org/10.1023/B:BREA.0000018401.13609.54 BCTRD6 Google Scholar

256.

R. R. Zhang et al., “Beyond the margins: real-time detection of cancer using targeted fluorophores,” Nat. Rev. Clin. Oncol., 14 347 –364 (2017). http://dx.doi.org/10.1038/nrclinonc.2016.212 Google Scholar

257.

M. Robertson-Tessi et al., “Impact of metabolic heterogeneity on tumor growth, invasion, and treatment outcomes,” Cancer Res., 75 1567 –1579 (2015). http://dx.doi.org/10.1158/0008-5472.CAN-14-1428 CNREA8 0008-5472 Google Scholar

258.

A. Simmons et al., “Environmental factors in breast cancer invasion: a mathematical modelling review,” Pathology, 49 172 –180 (2017). http://dx.doi.org/10.1016/j.pathol.2016.11.004 PTLGAX 1465-3931 Google Scholar

259.

R. J. Gillies, P. E. Kinahan and H. Hricak, “Radiomics: images are more than pictures, they are data,” Radiology, 278 563 –577 (2015). http://dx.doi.org/10.1148/radiol.2015151169 RADLAX 0033-8419 Google Scholar

260.

M. D. Kuo and N. Jamshidi, “Behind the numbers: decoding molecular phenotypes with radiogenomicsguiding principles and technical considerations,” Radiology, 270 320 –325 (2014). http://dx.doi.org/10.1148/radiol.13132195 RADLAX 0033-8419 Google Scholar

Biography

Michael A. Pinkert is a graduate student of medical physics at the University of Wisconsin at Madison. He studied physics as an undergraduate at Rensellaer Polytechnic University. He is working on his dissertation at the Laboratory for Optical and Computation Instrumentation.

Lonie R. Salkowski is a professor of radiology at the University of Wisconsin at Madison and has received her doctorate in curriculum and instruction through the UW School of Education. She is a member of the Anatomy Task Force and Year 1 Curriculum Development Committees.

Patricia J. Keely was the Jan and Kathryn Ver Hagen Professor of translational research at the University of Wisconsin at Madison. Her research interests were in understanding how molecular level cellular interactions with the extracellular matrix are altered during carcinogenesis to result in invasive, metastatic carcinoma. She passed away on Saturday, June 24, 2017.

Timothy J. Hall is a professor and vice chair for faculty of medical physics at the University of Wisconsin at Madison. His research interests include acoustic scattering, tissue elasticity, breast imaging, cervical assessment, and ultrasound phantom development.

Walter F. Block is a professor of medical physics at the University of Wisconsin at Madison. His research interests include magnetic resonance (MR) interventional procedures, MR angiography and cardiac imaging, MR contrast mechanisms, signal and image processing, and distributed computing.

Kevin W. Eliceiri is the director of the Laboratory for Optical and Computational Instrumentation at the University of Wisconsin at Madison and a principal investigator in the Morgridge Institute for Research. His current research focuses on the development of novel optical and computational methods for investigating cell signaling and cancer progression, and the development of multiscale imaging methods.

Citation Download Citation

Michael A. Pinkert, Lonie R. Salkowski, Patricia J. Keely, Timothy J. Hall, Walter F. Block, and Kevin W. Eliceiri "Review of quantitative multiscale imaging of breast cancer," Journal of Medical Imaging 5(1), 010901 (22 January 2018). https://doi.org/10.1117/1.JMI.5.1.010901

Received: 3 June 2017; Accepted: 19 December 2017; Published: 22 January 2018

Access the abstract

JOURNAL ARTICLE
20 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

CITATIONS

Cited by 18 scholarly publications.

Explore citations on Lens.org

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Breast cancer

Breast

Imaging systems

Tissues

Mammography

Tumors

Magnetic resonance imaging

1.

Introduction