JURJ

Digital Image Processing application to predict the TB using CXR images: An case study for optimization of medical images

Author: Pawan Kumar Jatwaa, Suraj Yadav

Abstract:

TB is a lethal malignant infectious illness in most parts of the world. Image Processing analysis and categorization of chest X-rays (CXRs) into TB and non-TB can be a valid alternative to predict the TB disease. The work proposes an automated TB detection technique employing sophisticated IP (image processing) models. A large section of a CXR picture is black, preventing diagnosis and misleading IP models. Therefore, in the proposed approach, we use advanced segmentation networks to extract the region of interest from multi-modal CXRs. IP models get segmented images to create the model development.

Researcher employ explainable machine learning to depict TBinfected lungs for the subjective judgment. In present study various techniques are compare with various techniques to find the best prediction method. ImageJ software was used for the present study. Images uses for the present investigation was collected from public domain like kaggle.