Deep Learning-based Tuberculosis Detection from Chest X-rays: An Indonesian Multi-center Dataset

Hidayat, Farhan

doi:10.99999/jhi.v4i2.1

Original Research

Deep Learning-based Tuberculosis Detection from Chest X-rays: An Indonesian Multi-center Dataset

Farhan Hidayat¹ *

1Institut Teknologi Bandung, Indonesia

* Corresponding author: farhan@example.com

Published 01 June 2025 · Vol. 4 No. 2 (2023) · Pages 1–15

DOI: 10.99999/jhi.v4i2.1

DOI link ↗ View Issue

Abstract

We present a deep learning model trained on 12,847 chest X-ray images from five Indonesian hospitals for tuberculosis screening. The model achieves 94.2% sensitivity and 91.8% specificity, comparable to senior radiologists. We discuss deployment challenges in district hospitals with limited radiologists.

Keywords: tuberculosis, deep learning, chest X-ray, computer-aided diagnosis

784Views

225Downloads

DOIRegistered

How to cite this article Hidayat, F. (2025). Deep Learning-based Tuberculosis Detection from Chest X-rays: An Indonesian Multi-center Dataset. Journal of Health Informatics, 4(2), 1–15. https://doi.org/10.99999/jhi.v4i2.1