Domain and Geometry Agnostic CNNs for Left Atrium Segmentation in 3D Ultrasound

TL;DR

This paper presents a deep learning approach for automatic 3D ultrasound segmentation of the left atrium, incorporating shape priors and adversarial learning to improve accuracy and domain adaptation.

Contribution

It introduces a novel combined CNN method that integrates shape priors and adversarial training for improved left atrium segmentation in 3D ultrasound.

Findings

Shape prior inclusion enhances domain adaptation.

Adversarial learning further improves segmentation accuracy.

Method outperforms existing approaches in accuracy.

Abstract

Segmentation of the left atrium and deriving its size can help to predict and detect various cardiovascular conditions. Automation of this process in 3D Ultrasound image data is desirable, since manual delineations are time-consuming, challenging and observer-dependent. Convolutional neural networks have made improvements in computer vision and in medical image analysis. They have successfully been applied to segmentation tasks and were extended to work on volumetric data. In this paper we introduce a combined deep-learning based approach on volumetric segmentation in Ultrasound acquisitions with incorporation of prior knowledge about left atrial shape and imaging device. The results show, that including a shape prior helps the domain adaptation and the accuracy of segmentation is further increased with adversarial learning.