Leveraging Sequential CNNs for Tuberculosis Detection in Chest X-Rays: Employing Convolutional Neural Networks to Spot Tuberculosis in Radiographs

Tuberculosis (TB) is a substantial global health concern, with millions of new cases reported each. Timely identification is essential for effective handling and containment of the disease's transmission. Our research is centered on developing and evaluating a Convolutional Neural Network (CNN) for automatically classifying tuberculosis using chest X-ray pictures. The dataset includes 700 TB pictures and 3,500 normal images, as well as 2,800 TB images obtained from the NIAID TB site. The CNN that was recommended was trained and validated, showing strong performance on both datasets. The results showed high precision (98.90%) and recall (97.83%) on the validation set, indicating the model's ability to accurately identify TB-positive cases. Additional testing on a separate dataset confirms the reliability of the model, achieving an overall accuracy of 99.52%. The results emphasize the capacity of deep learning to assist healthcare professionals in diagnosing TB, leading to prompt intervention and enhanced patient results. Our research emphasizes the need of using machine learning to improve TB detection, providing a vital tool for healthcare professionals and public health efforts globally.