SarcGraph for High-Throughput Regional Analysis of Sarcomere Organization and Contractile Function in 2D Cardiac Muscle Bundles.

Timelapse images of human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) provide rich information on cell structure and contractile function. The two-dimensional cardiac muscle bundle (2DMB) platform standardizes tissue geometry, enabling physiologic, uniaxial contractions on elastomeric substrates. However, larger sarcomere displacements in 2DMBs challenge existing tracking pipelines. We present adaptations to SarcGraph, an open-source Python package for sarcomere detection and tracking, enabling automated analysis of high-frame-rate 2DMB videos. Key modifications include frame-by-frame detection with automated segmentation, Gaussian Process Regression for denoising, and automatic contractile phase detection. We provide 130 example movies through Harvard Dataverse, enabling high-throughput analysis and advancing hiPSC-CM research.