Detecting the cancers in early stage plays the crucial role in preventing mortality from them. Although the conventional clinical tests are reliable, they could not offer the early access in case of the biopsy is invasive and requires specialists and specific facilities to be performed. Meanwhile, polarimetric imaging systems has been widely used recently for many biological applications, such as tissue morphological characterization, cancer staging, and/or detection.
A new approach is proposed for detecting and classifying cancerous human tissue based on the characteristics of the Mueller matrix elements / histopathological images and machine learning technique. Machine learning provides a powerful tool for performing the objective and precise diagnosis of cancer through its use of statistics, probabilistic algorithms and massive computational power. Accordingly, the present study explores the feasibility for using a machine learning technique to discriminate between normal tissue and cancer based on the images of 16 elements of the Mueller matrix of a biomedical sample (specifically in skin / breast / liver / blood cancer …) in which all of the optical effects may appear simultaneously. Along with statistical analyses, Mueller matrix images / histopathological images and parameters of malignant and healthy lesions from the mouse model in vitro measurement are figured out. Overall, the results show that the proposed framework has a promising potential for the development of machine learning approaches for automated cancer tissue screening and diagnosis.
Keywords: Polarimetric imaging, breast cancer classification, Mueller matrix image, machine learning, cancerous tissue.
