Chapter 8 - A Set of Texture-Based Methods for Breast Cancer Response Prediction in Neoadjuvant Chemotherapy Treatment

Abstract

Texture analysis of medical images is a useful way of allowing to increase the information obtainable from medical images. It represents a very useful task for pathological tissue characterization and classification. The aim of this chapter is to prove that texture-based feature techniques such as co-occurrence matrix, fractal analysis, and local binary pattern features extracted from MRI images can be used to quantify response of tumor treatment. To this aim, we use a dataset composed of two breast MRI examinations for nine patients. Three of them were responders and six nonresponders. The first exam was done before the initiation of the treatment (baseline). The latter was done after the first cycle of the chemo treatment (control). A set of selected texture parameters have been selected and calculated for each exam. These selected parameters are Cluster Shade, dissimilarity, entropy and homogeneity extracted from the gray-level co-occurrence matrix (GLCM), the local binary pattern (LBP) descriptor, and fractal analysis. The P values estimated for the pathologic complete responder pCR and nonpathologic complete responder pNCR patients have proven that in the case of pCR responder patients, we observe an important variation in tumor texture. This variation is used to quantify the efficiency of the treatment used.