3D Deep Affine-Invariant Shape Learning for Brain MR Image Segmentation

TL;DR

This paper introduces a novel shape-aware segmentation method for brain MRI images that effectively incorporates shape priors, resulting in improved accuracy over existing methods.

Contribution

The paper proposes a new shape integration technique that enhances segmentation performance without adding extra parameters to the network.

Findings

Lower Hausdorff distance compared to state-of-the-art methods

Higher Dice coefficient indicating better segmentation accuracy

Effective incorporation of shape priors in deep learning models

Abstract

Recent advancements in medical image segmentation techniques have achieved compelling results. However, most of the widely used approaches do not take into account any prior knowledge about the shape of the biomedical structures being segmented. More recently, some works have presented approaches to incorporate shape information. However, many of them are indeed introducing more parameters to the segmentation network to learn the general features, which any segmentation network is able learn, instead of specifically shape features. In this paper, we present a novel approach that seamlessly integrates the shape information into the segmentation network. Experiments on human brain MRI segmentation demonstrate that our approach can achieve a lower Hausdorff distance and higher Dice coefficient than the state-of-the-art approaches.