PROJECT

Role of Artificial intelligence using XRAY-CAD Software in detection and differentiation of TB using a phased approach

2022 - -

Objectives:

To develop a computer-aided detection (CAD) system for detecting abnormalities in Chest Xray potentially intensifying TB screening and characterisation of Chest X-rays for triaging to reduce Mortality in Lung diseases . A chatbot is being developed as a clinical decision support system for TB care tailored to low-resource settings in India, guiding the health worker, taking patient care to the doorsteps.

Study Overview:

This study aims to develop an artificial intelligence-based software tool that can help doctors detect Chest Xray abnormalities instantly suggestive of lung disease including pulmonary tuberculosis. The software is being trained using large collections of de-identified X-rays from Surveys, microbiologically proven pulmonary TB cases and finally “TB mimics”

Study Approach:

The study is a phased prospective observational validation study using de-identified chest X-rays collected from TB surveys, clinical trials, and hospitals. Stage 1 of the project involves developing an AI-based chest X-ray interpretation platform to identify PTB and differentiate normal chest radiographs variants through detailed clinician-led annotation and interpretation. Stage 2 focuses on training the system to recognize various radiographic abnormalities such as opacities, cavitary lesions, fibrosis etc which are likely to be TB . Stage 3 aims to expand the application of the platform beyond tuberculosis to identify and differentiate other thoracic diseases. Deep learning and Convoluted neural network models developed by CDAC, Chennai is being developed and validated.

Expected Public Health Impact:

The study is expected to support early and accurate TB screening, especially in high-burden and low-resource settings apart from triaging for reducing Mortality due to lung ailment . This tool if successful would be used as a POINT of care test for better and timely referral for treatment of TB , reducing mortality for the patient and transmission in the community