Semantic-Aware Contrastive Learning for Multi-object Medical Image Segmentation

TL;DR

This paper introduces a semantic-aware contrastive learning method that improves multi-object medical image segmentation by embedding different objects into separate clusters, leading to significant performance gains across multiple datasets.

Contribution

It proposes a novel contrastive learning approach using attention masks to enhance multi-object segmentation, addressing challenges of multiple objects with different semantics in medical images.

Findings

Improves Dice score by over 5% on medical datasets

Achieves 2.75% mIoU improvement on PASCAL VOC 2012

Demonstrates statistical significance with p-value<0.01

Abstract

Medical image segmentation, or computing voxelwise semantic masks, is a fundamental yet challenging task to compute a voxel-level semantic mask. To increase the ability of encoder-decoder neural networks to perform this task across large clinical cohorts, contrastive learning provides an opportunity to stabilize model initialization and enhance encoders without labels. However, multiple target objects (with different semantic meanings) may exist in a single image, which poses a problem for adapting traditional contrastive learning methods from prevalent 'image-level classification' to 'pixel-level segmentation'. In this paper, we propose a simple semantic-aware contrastive learning approach leveraging attention masks to advance multi-object semantic segmentation. Briefly, we embed different semantic objects to different clusters rather than the traditional image-level embeddings. We evaluate our proposed method on a multi-organ medical image segmentation task with both in-house data and MICCAI Challenge 2015 BTCV datasets. Compared with current state-of-the-art training strategies, our proposed pipeline yields a substantial improvement of 5.53% and 6.09% on Dice score for both medical image segmentation cohorts respectively (p-value<0.01). The performance of the proposed method is further assessed on natural images via the PASCAL VOC 2012 dataset, and achieves a substantial improvement of 2.75% on mIoU (p-value<0.01).