Detecting Hypo-plastic Left Heart Syndrome in Fetal Ultrasound via Disease-Specific Atlas Maps

Fetal ultrasound screening during pregnancy plays a vital role in the early detection of fetal malformations which have potential long-term health impacts. The level skill required to diagnose such from live examination is high and resources for are often limited. We present an interpretable, atlas-learning segmentation method automatic diagnosis Hypo-plastic Left Heart Syndrome (HLHS) single ‘4 Chamber Heart’ view image. propose extend recently introduced Image-and-Spatial Transformer Networks (Atlas-ISTN) into framework that enables sensitising atlas generation disease. In this we can jointly learn image segmentation, registration, construction disease prediction while providing maximum clinical interpretability compared direct classification methods. As result our allows diagnoses competitive with expert-derived manual yields AUC-ROC 0.978 (1043 cases training, 260 validation 325 testing).