Development of AI Model for Tuberculosis Detection in Chest CT Images

Abstract

Tuberculosis (TB) remains a major public health issue in Indonesia, characterized by challenges such as delayed early detection and low patient adherence to long-term treatment. Current TB prevention and control efforts still rely heavily on healthcare facilities, while support for technology-based patient self-management has not yet been optimally integrated. Advances in Artificial Intelligence (AI), specifically deep learning, offer opportunities to develop intelligent models that support sustainable patient self-empowerment. The objective of this study is to develop an AI-based application for the early detection of tuberculosis using CT scan images. The research employs an AI model development approach for the early detection of tuberculosis (TB) based on Vision Transformers and CNN-Transformer hybrids, as well as multimodal deep learning that integrates medical imaging and clinical data. A key aspect of this study is the integration of Explainable AI to enhance model transparency and interpretability, thereby fostering trust within the digital health context. The study’s novelty lies in the development of an AI-based TB detection model designed to support patient self-management. The findings are expected to serve as a scientific foundation for the future development of digital health technologies for TB control.