AI X-Rays Detect More TB Cases, Offering New Hope for Early Diagnosis
In the space of AI-assisted chest X-rays, India possesses a powerful technology for screening presumptive tuberculosis (TB) cases. Qure.ai, a Mumbai-based company, has developed the AI algorithm qXR, which can swiftly detect individuals with presumptive TB in less than a minute. The 2019-2021 National TB prevalence survey in India revealed that nearly 43% of TB cases would have remained undetected without chest X-rays. When deployed at scale for population-based or targeted screening, qXR, combined with molecular TB confirmation tests, has the potential to significantly boost detection rates.
To systematically screen for TB and facilitate early diagnosis, India”s “End TB” strategy considers systematic screening as a crucial element. Recently, India”s drug regulator granted clearance for qXR, which aligns with WHO requirements, boasting over 90% sensitivity and over 70% specificity for individuals aged 15 and above. However, states await a directive from the Central TB Division to adopt qXR, as the absence of policy guidance has hindered widespread adoption.
The Central TB Division, in turn, is awaiting approval from the Health Technology Assessment. Achieving India”s ambitious goal of “eliminating” TB by 2025 hinges on realising early diagnosis and prompt care for millions of TB patients. The widespread utilisation of AI-assisted chest X-rays for screening serves as the inaugural step toward attaining this objective.
A successful precedent exists in Vietnam, where community-wide screening of individuals aged 15 and above, using molecular tests from 2014 to 2017, led to a lower prevalence of pulmonary TB in 2018 compared to standard passive case detection. Contrastingly, in India, employing qXR to interpret digital X-rays before molecular testing within community screening protocols could reduce TB prevalence and minimise the number of molecular tests required for TB detection.
The qXR algorithm has already been adopted in more than 50 countries, with 24 Indian states employing qXR at approximately 150 sites. However, the challenge lies in scaling up its use across India. In certain states, qXR software is deployed at just one site, exemplified by Kerala. Dr. Shibu Vijayan, Medical Director-Global Health at Qure.ai, emphasises that a digital solution lacking scale will fail to make a substantial impact.
Chennai deploys around six outreach mobile vans that use qXR, while nine government hospitals and one mobile van in Mumbai are equipped with the software for surveillance. As of December 2022, over 100,000 X-rays from the nine Mumbai hospitals were screened for TB. Dr. Vijayan reports that an assessment in select facilities revealed that 30% of individuals diagnosed with TB were not originally undergoing TB referral X-rays but rather routine X-rays for other purposes. This finding implies that the exercise helped detect an additional 13% of TB cases.
The use of qXR in screening X-rays resulted in an 18-27% increase in the positivity of molecular tests. Qure.ai has collaborated with Mylab Discovery Solutions to integrate qXR into Mylab”s portable chest X-ray device, MyBeam. This development aims to screen presumptive TB cases in rural areas while minimizing X-ray exposure substantially, making it a valuable asset for public health initiatives.
(Disclaimer: This story is based on a Press Release. Timeline. is not responsible for the claims made by the company)
