Clinical decision support in hematological malignancies using a case-grounded AI agent.
AI agent matched tumor board decisions in 81.8% of external validation cases
Clinical decision support in hematological malignancies using a case-grounded AI agent.
Multidisciplinary tumor boards integrate longitudinal treatment histories, molecular profiling and rapidly evolving evidence to guide decisions in hematological malignancies, yet access to this level of subspecialty deliberation is increasingly uneven.
External validation on 555 independent cases from a second academic center yielded 81.8% concordance across 47 entities, and a prospective 1-month silent trial on 64 consecutive, unselected cases achieved 82.8% concordance.
AI recommendations matched a second hospital's tumor board most of the time
In expert-blinded benchmarking on 45 high-complexity cases across six foundation models, HemaGuide substantially improved concordance with tumor board decisions.
A systematic ablation study across 11 layers confirmed that performance gains were routing-type-dependent, with no single component sufficient across case types.
Automated classification of 70 clinically relevant missense variants showed high concordance with expert standards; no oncogenic variant was downgraded to benign and the whole workflow was completed under real-time conditions on commodity hardware with a median latency of 39 s rather than the hours typically required for manual molecular board workflows.
Together these data provide evidence that locally deployable, case-grounded large language model agents can deliver auditable clinical decision support across hematological malignancies, with concordance maintained across institutions and under real-time conditions on commodity hardware.