In this research, an automated system is developed with the help of ANN and image processing methods. MATLAB provides access to these toolboxes. The MATLAB environment is used to create the entire system. As a classifier, the ANN toolbox employs Feed Forward Back Propagation (FFBP). X-ray pictures are used as input for PTB detection. These X-ray images are used in the application of segmentation and enhancement techniques. Features such as shape and texture are retrieved from the resulting image. The neural network is trained using these features as input. These characteristics are taken into account alongside the outcome of a clinical examination (sputum). After the ANN has been trained, it is put through its paces with a test X-ray picture. Segmentation and enhancement, the first two processes that occurred during training the ANN, will likewise occur during testing. Comparisons are made between the test image's extracted form and texture features and the trained features. The ANN classifier determines whether or not a given case involves tuberculosis. In addition to the classification, a severity check is performed. Architecture (6-2) is used for form characteristics, while architecture (128-40-10-2) is used for gabour features, making this ANN multi-level. The system's output, intermediate steps, etc., are presented to the user via a graphical user interface (GUI).