Classification of Breast Cancer Images in Mice Utilizing Mueller Matrix Transformation

A framework to discriminate between polarization characteristics of pathological and normal breast tissues utilizing Mueller matrix decomposition was proposed. This study uses the method indicated to diagnose breast cancer in vivo by the data collection, extraction, and evaluation of the polarimetric characteristics of breast cancer tumors on mouse models. Ten estrogen- and DMBA-induced mouse model lines of Swiss albino mice were exposed to various cancer lesions, such as skin cancer and early-stage breast cancer, to produce a total of 40 malignant samples and 42 healthy samples. Along with statistical analyses, Mueller matrix images and parameters of malignant and healthy lesions from the in vivo measurement are figured out. These findings suggest that the m41, m42, and m44 elements from the Mueller matrix transformation can distinguish between two different sample types. This research proposed a novel non-invasively method for non-cancerous and cancerous tissue classification.