Loss Design and Architecture Selection for Long-Tailed Multi-Label Chest X-Ray Classification

TL;DR

This paper systematically evaluates loss functions, architectures, and strategies for long-tailed multi-label chest X-ray classification, achieving competitive results and providing practical insights for handling class imbalance in clinical imaging.

Contribution

It presents a comprehensive empirical study on loss functions and architectures for long-tailed CXR classification, highlighting LDAM-DRW and ConvNeXt-Large as effective choices.

Findings

LDAM-DRW outperforms standard losses for rare class recognition

ConvNeXt-Large achieves top performance with 0.5220 mAP

Our method ranks 5th on the official leaderboard

Abstract

Long-tailed class distributions pose a significant challenge for multi-label chest X-ray (CXR) classification, where rare but clinically important findings are severely underrepresented. In this work, we present a systematic empirical evaluation of loss functions, CNN backbone architectures and post-training strategies on the CXR-LT 2026 benchmark, comprising approximately 143K images with 30 disease labels from PadChest. Our experiments demonstrate that LDAM with deferred re-weighting (LDAM-DRW) consistently outperforms standard BCE and asymmetric losses for rare class recognition. Amongst the architectures evaluated, ConvNeXt-Large achieves the best single-model performance with 0.5220 mAP and 0.3765 F1 on our development set, whilst classifier re-training and test-time augmentation further improve ranking metrics. On the official test leaderboard, our submission achieved 0.3950 mAP, ranking 5th amongst all 68 participating teams with total of 1528 submissions. We provide a candid analysis of the development-to-test performance gap and discuss practical insights for handling class imbalance in clinical imaging settings. Code is available at https://github.com/Nikhil-Rao20/Long_Tail.