Reproducing and Improving CheXNet: Deep Learning for Chest X-ray Disease Classification

TL;DR

This paper reproduces CheXNet for chest X-ray disease classification and introduces improved models that outperform the original baseline, demonstrating enhanced performance metrics on a large public dataset.

Contribution

The study reproduces CheXNet and explores new algorithms that surpass its baseline performance on the NIH ChestX-ray14 dataset.

Findings

Best model achieved an AUC-ROC of 0.85

F1 score improved to 0.39

Outperformed baseline CheXNet model

Abstract

Deep learning for radiologic image analysis is a rapidly growing field in biomedical research and is likely to become a standard practice in modern medicine. On the publicly available NIH ChestX-ray14 dataset, containing X-ray images that are classified by the presence or absence of 14 different diseases, we reproduced an algorithm known as CheXNet, as well as explored other algorithms that outperform CheXNet's baseline metrics. Model performance was primarily evaluated using the F1 score and AUC-ROC, both of which are critical metrics for imbalanced, multi-label classification tasks in medical imaging. The best model achieved an average AUC-ROC score of 0.85 and an average F1 score of 0.39 across all 14 disease classifications present in the dataset.