A generalized deep learning model for multi-disease Chest X-Ray diagnostics

TL;DR

This paper presents a deep learning model trained on multi-site chest X-ray datasets, demonstrating improved generalization across different patient populations for multi-disease diagnosis.

Contribution

The study introduces a sequential training approach that enhances the generalizability of CNNs for multi-disease chest X-ray classification across diverse datasets.

Findings

Model trained on multiple datasets outperforms single-dataset models.

Significant improvement in prediction accuracy for 3 of 4 disease classes.

Model generalizes well across different hospital data sources.

Abstract

We investigate the generalizability of deep convolutional neural network (CNN) on the task of disease classification from chest x-rays collected over multiple sites. We systematically train the model using datasets from three independent sites with different patient populations: National Institute of Health (NIH), Stanford University Medical Centre (CheXpert), and Shifa International Hospital (SIH). We formulate a sequential training approach and demonstrate that the model produces generalized prediction performance using held out test sets from the three sites. Our model generalizes better when trained on multiple datasets, with the CheXpert-Shifa-NET model performing significantly better (p-values < 0.05) than the models trained on individual datasets for 3 out of the 4 distinct disease classes. The code for training the model will be made available open source at: www.github.com/link-to-code at the time of publication.