A Pathology Foundation Model for Gastric Cancer with Real-World Validation

2026-06-03

AI Summary

Gastric cancer exhibits substantial histological and molecular heterogeneity, limiting the performance of general-purpose pathology foundation models in fine-grained clinical tasks and lacking prospective validation. To address this, we developed GRACE, the first gastric cancer–specific pathology foundation model, trained and evaluated on 48,364 H&E-stained whole-slide images from 37,493 patients across 28 clinical tasks. Through multicenter real-world validation, randomized cross-reading trials, and a safety-gating mechanism, GRACE achieved a macro-AUC of 0.9188, improving diagnostic accuracy from 82.0% to 89.9% and reducing diagnostic time by 14.9%. Furthermore, it enabled non-inferior automated triage for 60.7% of atrophic gastritis and 82.7% of intestinal metaplasia cases, significantly enhancing pathologists’ diagnostic efficiency.

Abstract

Gastric cancer remains a major cause of cancer mortality, yet its histological and molecular heterogeneity complicates diagnosis and risk stratification. General-purpose pathology foundation models (PFMs) often plateau on fine-grained endpoints central to gastric cancer care, and few have undergone rigorous prospective validation or clinical reader studies. We present GRACE, a Gastric-specific foundation model for Real-world Assessment and Clinical dEcision support. GRACE was developed from multicenter gastric pathology datasets totaling 48,364 primarily HE-stained whole-slide images from 37,493 patients. When evaluated on 28 clinically relevant tasks, GRACE consistently outperformed representative pancancer PFMs, achieving a macro-AUC of 0.9188, with strong performance for precancerous lesion diagnosis (macro-AUC 0.9322), tumor histopathological assessment (macro-AUC 0.9119), molecular profiling (macro-AUC 0.8682), and prognostic prediction. Beyond benchmarking, GRACE's translational value was substantiated through a rigorous evidence chain. Under safety-gated criteria requiring 100% NPV for rule-out and 100% PPV for rule-in, GRACE streamlined review for up to 69.6% of malignancy-diagnosis cases and triaged 46.8% of MMR-IHC follow-up requests. This translational feasibility was further strengthened by a randomized crossover reader study of pathologist-AI collaboration. With GRACE assistance, diagnostic accuracy improved from 82.0% to 89.9%, yielding nearly twofold higher adjusted odds of a correct diagnosis (OR 1.987) alongside concurrent gains in sensitivity and specificity. AI assistance also reduced diagnostic time by 14.9%, elevated diagnostic confidence by 9.0%, and markedly improved inter-rater agreement. When calibrated to maintain non-inferior performance to senior pathologists, the AI-assisted workflow could triage 60.7% of atrophy and 82.7% of intestinal metaplasia cases.

Problem

• gastric cancer
• histological heterogeneity
• molecular heterogeneity
• diagnosis
• risk stratification

Innovation

• foundation model
• gastric cancer
• real-world validation
• pathologist-AI collaboration
• clinical decision support