Abstract

Deep Organ: How Machine Learning Can Improve Organs on a Chip Experiments

Davide Di Giuseppe, PHD student, University of Rome "Tor Vergata"

The use of organs-on-a-chip in biology has led to a profound change of paradigm in the way of conceiving experiments and understanding biological phenomena. But, at the same time, all its potentialities are not yet completely exploited. There is a plenty of information that can be extrapolated from the time-lapse microscopy images of OOC experiments using deep learning analysis. Machine learning includes cell localization and tracking, feature extraction, and cell classification thanks to the use of modern video analysis algorithms and deep learning architectures. The latter approach allows not only to improve effectiveness of the analysis but to conduct massive studies in which diverse biological conditions and cell types can be considered at a time, thus reducing time to market and incrementing the available testing rate.