Detection and Classification of Tuberculosis Disease Using Hybrid Deep Learning Method

Tuberculosis remains a significant global health challenge, necessitating precise diagnostic tools for effective intervention. This research addresses the urgent need for an efficient Tuberculosis Detection System leveraging advanced deep learning techniques. With a diverse dataset of chest X-ray images, Convolutional Neural Networks (CNNs) are employed for automatic feature extraction, resulting in robust performance with high accuracy. Utilizing a diverse array of models InceptionV3, Xception, DenseNet, and CNN .By addressing the urgent need for accurate Tuberculosis diagnosis, this Tuberculosis Detection System contributes to global health efforts, paving the way for improved patient outcomes and disease control strategies. Comparative analysis identifies Xception as the highest overall accuracy of 96.95% with precision, recall, and F1 score values of 97%, 96%, and 97%, respectively. These results highlight the effectiveness of deep learning techniques in achieving high precision and recall crucial for early and reliable Tuberculosis diagnosis. Finally this signifies a leap forward in Tuberculosis detection, highlighting the transformative impact of deep learning in advancing medical diagnostics and public health efforts.