AI-Powered Mammography Screening Doubles Early Detection in Rural Clinics

A large-scale implementation study has demonstrated that an AI-assisted mammography interpretation system, developed and validated at the Cancer Institute, doubled the rate of early-stage breast cancer detection when deployed in rural screening clinics across Tamil Nadu. The system was evaluated in 12 primary health centres over 18 months, analysing over 15,000 screening mammograms.

The AI model, trained on a diverse dataset of 120,000 mammograms from Indian women — whose breast density patterns differ significantly from Western populations — achieved a sensitivity of 92% for invasive cancers, outperforming standalone radiologist interpretation in the community setting. Crucially, it reduced false-positive rates by 30%, decreasing unnecessary biopsies and patient anxiety.

The programme is now being expanded to 50 additional rural centres, with plans to integrate the AI system into India's national breast cancer screening guidelines.