Artificial Intelligence in Tuberculosis Detection and Diagnosis: Current Advances and Future Perspectives

Abstract

Tuberculosis (TB) remains one of the world's most significant infectious diseases despite major advances in prevention and treatment. Delayed diagnosis, limited access to diagnostic facilities, and shortages of trained healthcare professionals continue to impede global TB control efforts, particularly in low- and middle-income countries. Artificial intelligence (AI) has emerged as a transformative technology capable of improving TB detection, diagnosis, triage, and surveillance. Machine learning (ML), deep learning (DL), computer vision, and natural language processing are increasingly integrated with radiological imaging, molecular diagnostics, and digital health platforms. AI-assisted chest X-ray interpretation has demonstrated performance comparable to expert radiologists and has been endorsed by global health organizations for TB screening in selected populations. Recent developments include smartphone-based cough analysis, portable AI-enabled devices, predictive analytics, and integration with electronic health systems. Despite encouraging outcomes, challenges related to algorithm bias, data privacy, interpretability, infrastructure limitations, and ethical governance remain. This review discusses current AI applications in TB detection and diagnosis, summarizes recent advances, examines implementation challenges, and explores future directions for equitable and sustainable deployment.