ADeS: a deep learning based Apoptosis Detection System for live cell imaging

Abstract

AbstractLive-cell imaging allows the study of apoptosis at cellular level, highlighting morphological hallmarks such as nuclear shrinkage, membrane blebbing, and cell disruption. Identifying the exact location and timing of this process is essential to foster the understanding of its spatial-temporal regulation. However, the analysis of live-cell imaging datasets is complex, whereas computational tools tailored to this task are yet scarce. Therefore, we developed ADeS, an Apoptosis Detection System based on deep learning and activity recognition. ADeS uses morpho-dynamic hallmarks to detect the exact location and timing of apoptotic events in different cell types, reaching an accuracy above 97% in the classification of our validation datasets acquiredin vitroandin vivo. Moreover, ADeS is the first successful implementation of a deep learning network for the automatic detection of apoptotic cells in full microscopy movies in an end-to-end fashion, outperforming human in the same task. As a case study, we employed ADeS for the analysis of cell survivalin vitro, and for tissue damage assessmentin vivo, showing its potential application for toxicity assays, treatments evaluation and measuring of tissue dynamics.