Long-tailed medical diagnosis with relation-aware representation learning and iterative classifier calibration

• Published in: Computers in biology and medicine Vol. 188; p. 109772
• Main Authors: Pan, Li, Zhang, Yupei, Yang, Qiushi, Li, Tan, Chen, Zhen
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.04.2025; Elsevier Limited
• Subjects: Algorithms; Calibration; Categories; Class imbalance; Classification; Classifier calibration; Coders; Data augmentation; Databases, Factual; Datasets; Decoupling; Diagnosis; Diagnosis, Computer-Assisted - methods; Humans; Image classification; Internal Medicine; Learning; Machine Learning; Medical diagnosis; Medical image diagnosis; Medical imaging; Optimization; Other; Representation learning; Representations; Source code; Representation learning; Classifier calibration; Medical image diagnosis; Class imbalance
• ISSN: 0010-4825; 1879-0534; 1879-0534
• DOI: 10.1016/j.compbiomed.2025.109772

Recently computer-aided diagnosis has demonstrated promising performance, effectively alleviating the workload of clinicians. However, the inherent sample imbalance among different diseases leads algorithms biased to the majority categories, leading to poor performance for rare categories. Existing works formulated this challenge as a long-tailed problem and attempted to tackle it by decoupling the feature representation and classification. Yet, due to the imbalanced distribution and limited samples from tail classes, these works are prone to biased representation learning and insufficient classifier calibration. To tackle these problems, we propose a new Long-tailed Medical Diagnosis (LMD) framework for balanced medical image classification on long-tailed datasets. In the initial stage, we develop a Relation-aware Representation Learning (RRL) scheme to boost the representation ability by encouraging the encoder to capture intrinsic semantic features through different data augmentations. In the subsequent stage, we propose an Iterative Classifier Calibration (ICC) scheme to calibrate the classifier iteratively. This is achieved by generating a large number of balanced virtual features and fine-tuning the encoder using an Expectation-Maximization manner. The proposed ICC compensates for minority categories to facilitate unbiased classifier optimization while maintaining the diagnostic knowledge in majority classes. Comprehensive experiments on three public long-tailed medical datasets demonstrate that our LMD framework significantly surpasses state-of-the-art approaches. The source code can be accessed at https://github.com/peterlipan/LMD. •We present a novel method named LMD to tackle long-tailed problems in medical image diagnosis.•The proposed RRL module efficiently enhances the representation learning of medical image encoders.•The proposed ICC module iteratively calibrates the classifier with abundant balanced virtual features.•Experiments on three long-tailed medical image datasets illustrates the effectiveness of our LMD method.