Improving the Reliability of Semantic Segmentation of Medical Images by Uncertainty Modeling with Bayesian Deep Networks and Curriculum Learning

TL;DR

This paper introduces a novel approach combining Bayesian deep networks and curriculum learning to improve the reliability of semantic segmentation in medical images by focusing training on uncertain regions.

Contribution

It presents a new method that uses uncertainty estimates from Bayesian networks to guide curriculum learning, enhancing segmentation reliability.

Findings

Significant increase in model reliability for iPS cell colony segmentation

Effective use of uncertainty feedback to resample training data

Improved segmentation performance in high-uncertainty areas

Abstract

In this paper we propose a novel method which leverages the uncertainty measures provided by Bayesian deep networks through curriculum learning so that the uncertainty estimates are fed back to the system to resample the training data more densely in areas where uncertainty is high. We show in the concrete setting of a semantic segmentation task (iPS cell colony segmentation) that the proposed system is able to increase significantly the reliability of the model.