Applying Graph-neural networks to contour detection in medical ultrasound

Presenter: Sarina Thomas, Postdoctoral researchers, UiO

Abstract: Accurate and consistent predictions of echo cardiography parameters are important for cardiovascular diagnosis and treatment. In particular, segmentations of the left ventricle can be used to derive ventricular volume,  ejection fraction (EF) and other relevant measurements.  This project  proposes a new automated method for predicting ejection fraction and segmenting the left ventricle by detecting anatomical keypoints. Models for direct coordinate regression based on Graph Convolutional Networks (GCNs) are used to detect the keypoints. GCNs can learn to represent the cardiac shape based on local appearance of each keypoint, as well as global spatial and temporal structures of all keypoints combined. We evaluated our model on the EchoNet benchmark dataset. Compared to semantic segmentation, GCNs show accurate segmentation and improvements in robustness and inference run-time. EF is computed simultaneously to segmentations and our method also obtains state-of-the-art ejection fraction estimation.

‍