Abstract
Background Tuberculosis (TB) is an infectious disease caused by the bacterium Mycobacterium tuberculosis. It primarily affects the lungs but can also affect other organs, such as the kidneys, bones, and brain. TB is transmitted through the air when an infected individual coughs, sneezes, or speaks, releasing tiny droplets containing the bacteria. Despite significant efforts to combat TB, challenges such as drug resistance, co-infection with human immunodeficiency virus (HIV), and limited resources in high-burden settings continue to pose obstacles to its eradication. TB remains a significant global health challenge, necessitating accurate and timely detection for effective management. Methods This study aimed to develop a TB detection model using chest X-ray images obtained from Kaggle.com, utilizing Google's Collaboration Platform. Over 1196 chest X-ray images, comprising both TB-positive and normal cases, were employed for model development. The model was trained to recognize patterns within the TB chest X-rays to efficiently recognize TB within patients in order to be treated in time. Results The model achieved an average precision of 0.934, with precision and recall values of 94.1% each, indicating its high accuracy in classifying TB-positive and normal cases. Sensitivity and specificity values were calculated as 96.85% and 91.49%, respectively. The F1 score was also calculated to be 0.941. The overall accuracy of the model was found to be 94%. Conclusion These results highlight the potential of machine learning algorithms for TB detection using chest X-ray images. Further validation studies and research efforts are needed to assess the model's generalizability and integration into clinical practice, ultimately facilitating early detection and improved management of TB.