Introduction
Deep learning and machine learning have been effectively utilised to identify tuberculosis (TB) and to classify the patient's drug-resistance type in order to propose the most effective treatment plan to limit the spread of the bacillus. Due to the fact that there is no published model that can diagnose tuberculosis and classify the type of drug resistance in a single model presented in prior literatures. In this study, we will construct a model that can determine if a patient is afflicted with tuberculosis and, if so, what form of medication resistance that patient has. Although, the contribution of our proposed web application is as following. Web application was developed using HTML, JavaScript , and Tensorflow , under the Responsive design concept , making them available on a wide range of devices including computer, mobile phone, or tablet computer. Users can upload image files that they want to analyze. This image file is forwarded to a server to diagnose the disease and to suggest the appropriate drugs and medications for the predicted diseases. On the server side, once the image is obtained, the deep learning model is applied to predict the type of disease. Once the type of disease is obtained, it will search for detailed information and recommendations for the medication used in that particular disease and send it back to the user's side.
