Analysing the effectiveness of a generative model for semi-supervised medical image segmentation

TL;DR

This paper evaluates the effectiveness of generative models like SemanticGAN for semi-supervised medical image segmentation, comparing their performance, robustness, and disparities to discriminative models on large chest X-ray datasets.

Contribution

It provides a comprehensive analysis of generative versus discriminative models for semi-supervised medical image segmentation, highlighting their strengths and limitations.

Findings

Generative models show competitive segmentation performance.

Robustness varies between generative and discriminative approaches.

Potential subgroup disparities identified in model performance.

Abstract

Image segmentation is important in medical imaging, providing valuable, quantitative information for clinical decision-making in diagnosis, therapy, and intervention. The state-of-the-art in automated segmentation remains supervised learning, employing discriminative models such as U-Net. However, training these models requires access to large amounts of manually labelled data which is often difficult to obtain in real medical applications. In such settings, semi-supervised learning (SSL) attempts to leverage the abundance of unlabelled data to obtain more robust and reliable models. Recently, generative models have been proposed for semantic segmentation, as they make an attractive choice for SSL. Their ability to capture the joint distribution over input images and output label maps provides a natural way to incorporate information from unlabelled images. This paper analyses whether deep generative models such as the SemanticGAN are truly viable alternatives to tackle challenging medical image segmentation problems. To that end, we thoroughly evaluate the segmentation performance, robustness, and potential subgroup disparities of discriminative and generative segmentation methods when applied to large-scale, publicly available chest X-ray datasets.