| Popis: |
Tuberculosis is an infectious bacterial disease that mainly impacts the lungs. But these bacteria could attack any part of the body such as the kidney, brain, and spine. Therefore, if tuberculosis is not detected early, it could damage the body parts may also lead to death. In contrast to other infectious diseases, the diagnosis of tuberculosis is highly doughty. Numerous checks are usually needed where chest X-rays are mostly used for TB screening. This paper proposes a deep learning-based method to detect tuberculosis from a chest X-ray. For this, we have proposed a VGG16-based model for tuberculosis detection. For training, the model used two publically available datasets Shenzhen and Montgomery datasets [7]. We have achieved 91% and 89% accuracy, respectively, on the Shenzhen and Montgomery datasets. The results we have obtained show promising results in tuberculosis detection. |
