Deep Bayesian segmentation for colon polyps: Well-calibrated predictions in medical imaging

TL;DR

This paper introduces Bayesian neural network models for colon polyp segmentation that provide both high accuracy and reliable uncertainty estimates, improving early detection and clinical decision-making in medical imaging.

Contribution

It develops and evaluates Bayesian segmentation models using MNF and reparameterization on various architectures, achieving state-of-the-art performance with well-calibrated uncertainty estimates.

Findings

FPN + EfficientnetB7 with MNF achieves IOU of 0.94 and ECE of 0.004

Bayesian models outperform deterministic counterparts in uncertainty quantification

Models effectively identify difficult-to-detect polyps, aiding early diagnosis

Abstract

Colorectal polyps are generally benign alterations that, if not identified promptly and managed successfully, can progress to cancer and cause affectations on the colon mucosa, known as adenocarcinoma. Today advances in Deep Learning have demonstrated the ability to achieve significant performance in image classification and detection in medical diagnosis applications. Nevertheless, these models are prone to overfitting, and making decisions based only on point estimations may provide incorrect predictions. Thus, to obtain a more informed decision, we must consider point estimations along with their reliable uncertainty quantification. In this paper, we built different Bayesian neural network approaches based on the flexibility of posterior distribution to develop semantic segmentation of colorectal polyp images. We found that these models not only provide state-of-the-art performance on the segmentation of this medical dataset but also, yield accurate uncertainty estimates. We applied multiplicative normalized flows(MNF) and reparameterization trick on the UNET, FPN, and LINKNET architectures tested with multiple backbones in deterministic and Bayesian versions. We report that the FPN + EfficientnetB7 architecture with MNF is the most promising option given its IOU of 0.94 and Expected Calibration Error (ECE) of 0.004, combined with its superiority in identifying difficult-to-detect colorectal polyps, which is effective in clinical areas where early detection prevents the development of colon cancer.