Clinically-Inspired Hierarchical Multi-Label Classification of Chest X-rays with a Penalty-Based Loss Function

2025-02-05

AI Summary

Addressing the challenge of balancing clinical interpretability and model simplicity in multi-label chest X-ray classification, this paper proposes a unified, single-model, single-inference hierarchical framework. Methodologically: (1) it constructs a medical-knowledge-informed hierarchical label taxonomy; (2) it introduces a Hierarchical Binary Cross-Entropy (HBCE) loss function, enabling flexible modeling of diagnostic dependencies—either predefined or data-driven; and (3) it integrates Grad-CAM-based visual attribution with predictive uncertainty estimation. Evaluated on the CheXpert dataset and semantically enriched VisualCheXbert labels, the method achieves a mean test AUROC of 0.903. All code, configurations, and experimental details are publicly released. This work establishes a lightweight, reliable, and clinically friendly paradigm for interpretable AI in chest radiography.

Technology Category

• Computer Vision: Interpretability, Explainability, and Transparency
• Machine Learning: Calibration & Uncertainty Quantification
• Humans and AI: Explainable AI (XAI) for Human Understanding

Application Category

• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

In this work, we present a novel approach to multi-label chest X-ray (CXR) image classification that enhances clinical interpretability while maintaining a streamlined, single-model, single-run training pipeline. Leveraging the CheXpert dataset and VisualCheXbert-derived labels, we incorporate hierarchical label groupings to capture clinically meaningful relationships between diagnoses. To achieve this, we designed a custom hierarchical binary cross-entropy (HBCE) loss function that enforces label dependencies using either fixed or data-driven penalty types. Our model achieved a mean area under the receiver operating characteristic curve (AUROC) of 0.903 on the test set. Additionally, we provide visual explanations and uncertainty estimations to further enhance model interpretability. All code, model configurations, and experiment details are made available.

Problem

• Enhances clinical interpretability in chest X-ray classification
• Uses hierarchical label groupings for diagnosis relationships
• Introduces custom loss function for label dependencies

Innovation

• Hierarchical Multi-Label Classification
• Custom Hierarchical Loss Function
• Clinical Interpretability Enhancement