Calibrated Bayesian Deep Learning for Explainable Decision Support Systems Based on Medical Imaging

TL;DR

This paper introduces a Bayesian deep learning framework with novel calibration techniques to improve the reliability and interpretability of AI models in medical imaging decision support systems.

Contribution

It proposes a new probabilistic training loss and a post-hoc calibration method to enhance uncertainty quantification in medical imaging AI models.

Findings

Achieves improved calibration across multiple medical imaging tasks

Maintains performance in data-scarce and imbalanced datasets

Enhances model reliability for clinical decision support

Abstract

In critical decision support systems based on medical imaging, the reliability of AI-assisted decision-making is as relevant as predictive accuracy. Although deep learning models have demonstrated significant accuracy, they frequently suffer from miscalibration, manifested as overconfidence in erroneous predictions. To facilitate clinical acceptance, it is imperative that models quantify uncertainty in a manner that correlates with prediction correctness, allowing clinicians to identify unreliable outputs for further review. In order to address this necessity, the present paper proposes a generalizable probabilistic optimization framework grounded in Bayesian deep learning. Specifically, a novel Confidence-Uncertainty Boundary Loss (CUB-Loss) is introduced that imposes penalties on high-certainty errors and low-certainty correct predictions, explicitly enforcing alignment between prediction correctness and uncertainty estimates. Complementing this training-time optimization, a Dual Temperature Scaling (DTS) strategy is devised for post-hoc calibration, further refining the posterior distribution to improve intuitive explainability. The proposed framework is validated on three distinct medical imaging tasks: automatic screening of pneumonia, diabetic retinopathy detection, and identification of skin lesions. Empirical results demonstrate that the proposed approach achieves consistent calibration improvements across diverse modalities, maintains robust performance in data-scarce scenarios, and remains effective on severely imbalanced datasets, underscoring its potential for real clinical deployment.