Pulse.

a daily field guide to health research that matters

◆ Console

‹ Fri · 24 Apr 2026
Early cancer detection or prevention

Canine Olfaction Combined With Bayesian Modeling for Multicancer Detection From Breath Samples: A Phase II Study in India

Trained dogs paired with AI can detect multiple cancers from breath with high accuracy in early stages, offering a low-cost screening option for resource-limited settings.

This Phase II study in 1,502 participants across 6 Indian hospitals showed that trained detection dogs combined with a Bayesian fusion algorithm can detect multiple cancer types from breath samples with over 90% sensitivity and specificity. Notably, sensitivity was equally high for early-stage disease (stages I-II), supporting potential use as a low-cost triage screen in low- and middle-income country settings.

What the study was

Study design
Phase II multicenter assessor-masked case-control study
Population
Treatment-naïve biopsy-confirmed cancer patients (7 major cancer types) and controls in India; 6 hospitals in Karnataka
Sample size
1502
Category
Early Detection
Maturity
Validated
Journal
Journal of Clinical Oncology

Why it surfaced

Novel multicancer breath triage system achieving >90% sensitivity across 7 cancer types including stage I-II in a Phase II study, with direct relevance to LMIC screening where low-cost approaches are essential. Journal: JCO. Multicenter design with 1,502 test participants.

A plain-language summary of published research — not medical advice. Talk to a clinician about your own care.