An efficient framework for identification of Tuberculosis and Pneumonia in chest X-ray images using Neural Network

The diagnosis of pulmonary diseases through a chest X-ray is a tough task and needs expertise. Many of the pulmonary diseases mimic each other, and the diagnoses becomes challenging. Discriminating pulmonary tuberculosis (PTB) from other pulmonary disease like pneumonia, lung cancer etc. is a major concern in the diagnosis of tuberculosis. Several cases have been recorded that were misdiagnosed and have faced severe complications. Therefore, in this paper, we have proposed a framework to efficiently classify and discriminate between the PTB, Bacterial pneumonia and Viral Pneumonia from the collection of chest X-ray images. The analysis has been performed by using neural network classifier. For the pre-processing of the data, various data augmentation methods were used that improves the validation and classification accuracy of the proposed model. The proposed framework was able to efficiently classify and discriminate different pulmonary infections and achieves remarkable high accuracy of 99.01%.