Published 27 Sep 2020

How Deep Learning Is Used For Tuberculosis Detection In City Of Nagpur



The global fight against Tuberculosis (TB), one of the world's most deadly infectious diseases, has found a strong ally in Artificial Intelligence (AI).
In India, where the disease affects over 27 lakh lives yearly,, a leading healthcare-focused AI firm, has stepped up to tackle this crisis in collaboration with PATH, a non-profit organization.
The initiative's primary target was the Maharashtra city of Nagpur, with 35% of its population suffering from TB. The region largely relies on informal healthcare providers who, while affordable and accessible, lack the necessary tools for early TB detection.
Enter qXR,'s AI-powered chest X-ray interpretation platform. This revolutionary tool can accurately detect 29 different, clinically relevant lung markers, playing a vital role in identifying classic and atypical TB forms. It processes X-rays within minutes, swiftly identifying potential cases and alerting healthcare providers to expedite treatment procedures.
The tool's deep learning algorithms have been trained on a dataset of 3.6 million chest X-rays gathered from around 250 sites globally over four years. This AI-driven system can analyze multiple scans from the same patient sequentially, providing a progress report to track changes in lesions over time.
Dr. Shibu Vijayan, Global TB Technical Director at PATH, reported a 20% increase in the notification of TB cases since the implementation of qXR.
With TB diagnosis drastically expedited, this system promises a promising future for rural healthcare providers seeking to tackle the rampant disease.

Share this story