Detection and Classification of Local Ca²⁺ Release Events in Cardiomyocytes Using 3D-UNet Neural Network

Abstract

Global Ca²⁺ increase in the cytosol of cardiomyocytes is crucial for the contraction of the heart. Malfunctioning of proteins involved in this process can trigger local events (e.g., sparks and puffs) and global events (e.g., waves). These are thought to be involved in the development of arrhythmia. Therefore, it is important to detect and classify local Ca²⁺ release events. We present a novel approach, based on a 3D U‐Net architecture, to perform these tasks in a fully automated fashion. We employed data obtained with fast xyt confocal imaging of cardiomyocytes where such subcellular Ca²⁺ events are manually annotated and trained the neural network to infer comparable segmentation as output. Despite the relatively small amount of available data and the challenges that it exhibits, we obtained qualitatively promising results.