Machine learning powered Mueller matrix microscope for collagen and elastin visualization in the mouse cervix
23 January 2022 • 10:00 AM - 10:20 AM PST | Room 151 (Upper Mezzanine South)
One million infants will die of complications from reduced gestation before the age of five. Monitoring elastin and collagen arrangement in the uterine cervix interests researchers studying preterm birth. We have utilized a Self-validating Mueller matrix micro-mesoscope (SAMMM) with convolutional neural networks and K-nearest neighbor for classification of the fibers in the mouse cervix. A polarized microscope can be used for fiber classification leveraging the developed classifier. The Mueller matrix and decomposition parameters from excised cervical tissues are fed to the previously developed classifier. This low cost tool, powered with machine learning, could improve access to collagen and elastin imaging.
Florida International Univ. (United States)