Photo courtesy of Qure.ai
Indian medical imaging AI company Qure.ai has secured new approval in Europe for the expanded use of its chest X-ray analysis AI tool to diagnose paediatric tuberculosis in toddlers.
It has recently received a new CE Mark under the European Union's Medical Device Regulation for the use of its AI-powered chest X-ray software, qXR, in children aged 0-3 years.
FINDINGS
Last year, Qure.ai further trained qXR with more paediatric data – chest X-rays from children aged 0-15 years paired with corresponding GeneXpert results collected from several high TB-burden countries.
"We gathered thousands of these X-rays and retrained the model to capture the subtle and often non-specific patterns seen in younger children, particularly those under five," Divya Gupta, Chief Business Officer – Global Health at Qure.ai, told Mobihealth News in an email interview.
The model demonstrated a "marked increase in accuracy across all paediatric groups," she noted. Retrospective and prospective studies done after an internal validation and initial deployments at select global sites also showed "highly encouraging results." Qure.ai is set to report findings from a large-scale internal validation at a global conference on lung health in Denmark later in November.
"Preliminary analyses from earlier test sets have already shown a 20% improvement in diagnostic accuracy against GeneXpert results after incorporating paediatric data into the model," Gupta added.
WHY IT MATTERS
Qure.ai noted that children under five have so far been underrepresented in the development of AI tools for detecting TB. "The youngest children have long been the hardest to reach and the most vulnerable," said Dr Shibu Vijayan, the company’s chief medical officer.
Children under five account for more than 75% of all TB-related child deaths worldwide. In 2023 alone, most of the nearly 200,000 children who died from TB were in this age group. The growing prevalence of drug-resistant TB among children further complicates diagnosis and treatment.
Following the CE MDR certification for its AI tool, Qure.ai will "focus on broader deployments and collaborations." It will start securing registrations and approvals in high TB-burden countries across Africa and Southeast Asia to support national-level implementations. It also seeks collaborations on projects with children-focused organisations like Save the Children, Elizabeth Glaser Paediatric AIDS Foundation, and UNICEF.
THE LARGER CONTEXT
Beyond TB, Qure.ai is also addressing other major paediatric health concerns raised by its clients, including TB severity classification and paediatric pneumonia, "which are large causes of under-five mortality today."
Meanwhile, Qure.ai has also recently implemented one of the World Health Organization's treatment decision algorithms within its qTrack care coordination platform. The system enables structured data entry and automated calculation of clinical parameters to support TB diagnosis in children.
Qure.ai said integrating its TB screening tool and care coordination platform into a single workflow allows health professionals to make more informed, guideline-based decisions, even in decentralised or low-resource settings. "Essentially, qXR provides the diagnostic insight, and qTrack ensures the right follow-up and treatment pathway, closing the loop from screening to care," Gupta explained.
The company is also promoting AIRA, a large language model-based AI assistant for community health, launched in May. The tool automates patient data collection, aggregates population-level insights, and supports clinical protocol adherence and decision making.
