Exploring Weakly Supervised Semantic Segmentation Ensembles for Medical Imaging Systems

TL;DR

This paper introduces a novel ensemble framework that enhances weakly supervised semantic segmentation in medical imaging by combining multiple low-threshold CAMs, resulting in significant accuracy improvements on challenging datasets.

Contribution

The proposed framework effectively utilizes low-quality CAM predictions to improve segmentation accuracy in medical images, addressing the limitations of existing weakly supervised methods.

Findings

Achieved up to 8% dice score improvement on BRATS dataset.

Achieved up to 6% dice score improvement on DECATHLON dataset.

Demonstrated robustness across multi-modal medical imaging datasets.

Abstract

Reliable classification and detection of certain medical conditions, in images, with state-of-the-art semantic segmentation networks, require vast amounts of pixel-wise annotation. However, the public availability of such datasets is minimal. Therefore, semantic segmentation with image-level labels presents a promising alternative to this problem. Nevertheless, very few works have focused on evaluating this technique and its applicability to the medical sector. Due to their complexity and the small number of training examples in medical datasets, classifier-based weakly supervised networks like class activation maps (CAMs) struggle to extract useful information from them. However, most state-of-the-art approaches rely on them to achieve their improvements. Therefore, we propose a framework that can still utilize the low-quality CAM predictions of complicated datasets to improve the accuracy of our results. Our framework achieves that by first utilizing lower threshold CAMs to cover the target object with high certainty; second, by combining multiple low-threshold CAMs that even out their errors while highlighting the target object. We performed exhaustive experiments on the popular multi-modal BRATS and prostate DECATHLON segmentation challenge datasets. Using the proposed framework, we have demonstrated an improved dice score of up to 8% on BRATS and 6% on DECATHLON datasets compared to the previous state-of-the-art.