Comparative Benchmarking of Failure Detection Methods in Medical Image Segmentation: Unveiling the Role of Confidence Aggregation

TL;DR

This paper presents a benchmarking framework for failure detection in medical image segmentation, emphasizing confidence aggregation and identifying the pairwise Dice score as a robust baseline under distribution shifts.

Contribution

It introduces a comprehensive benchmarking framework and evaluates failure detection methods, highlighting the effectiveness of confidence aggregation and the pairwise Dice score.

Findings

Pairwise Dice score outperforms other methods in failure detection.

Risk-coverage analysis provides a holistic evaluation approach.

Confidence aggregation improves failure detection performance.

Abstract

Semantic segmentation is an essential component of medical image analysis research, with recent deep learning algorithms offering out-of-the-box applicability across diverse datasets. Despite these advancements, segmentation failures remain a significant concern for real-world clinical applications, necessitating reliable detection mechanisms. This paper introduces a comprehensive benchmarking framework aimed at evaluating failure detection methodologies within medical image segmentation. Through our analysis, we identify the strengths and limitations of current failure detection metrics, advocating for the risk-coverage analysis as a holistic evaluation approach. Utilizing a collective dataset comprising five public 3D medical image collections, we assess the efficacy of various failure detection strategies under realistic test-time distribution shifts. Our findings highlight the importance of pixel confidence aggregation and we observe superior performance of the pairwise Dice score (Roy et al., 2019) between ensemble predictions, positioning it as a simple and robust baseline for failure detection in medical image segmentation. To promote ongoing research, we make the benchmarking framework available to the community.