Automatic Endocardial Cushion Segmentation and EMT Cell Density Characterization Using Optical Coherence Tomography

Optical coherence tomography (OCT) has been applied for understanding heart development because of its capability of imaging both the structure and function of tiny beating embryonic hearts. Labeling endocardial cushions is necessary for quantifying morphological characteristics of the looping hearts. Since manual segmentation is time-consuming and prone to subjectivity, this study aims to use V-net to automatically segment endocardial cushions from OCT images. This will benefit research on heart development, especially studies requiring large cohorts of embryos, for example those investigating the teratogenic effects of ethanol or drugs and the prevention of these effects on the developing hearts.