In the current era of personalized cancer therapy, various drugs need to be rapidly tested to determine their efficacy in killing cancer cells. In this way, the most effective drug can be administered to improve therapy effectiveness. However, since the biopsy specimens are heterogeneous, efficient testing requires the use only of the viable tumor part of the specimen. Therefore, rapid and automated evaluation of specimen heterogeneity is needed. In this paper presents an automated algorithm for tissue heterogeneity estimation based on OCT images. This algorithm use image texture features to differentiate between healthy and cancer tissue.
|