AI-driven LPA analysis accelerates DR-TB diagnosis with clinical-grade accuracy, ready for widescale lab deployment
The challenge & Our AI Solution
Diagnosing drug-resistant TB (DR-TB) using LPA tests is crucial but manually intensive and prone to errors, delaying life-saving treatment.
We’ve automated the analysis of LPA test strips using AI, dramatically reducing errors and speeding up the diagnostic process for DR-TB.
An AI model accurately interprets the complex band patterns on LPA strips.
Real-World Impact
Validated by ICMR with outstanding accuracy (97.3% for First-Line, 100% for Second-Line LPA). Deployed nationwide in over 97 CDST labs, set to impact 400,000 tests annually through Ni-kshay integration.