Validation of an AI-based end-to-end model for prostate pathology using long-term archived routine samples

arXiv:2605.02614v1 Announce Type: cross Abstract: Artificial intelligence (AI) is becoming a clinical tool for prostate pathology, but generalization across variations in sample preparation and preservation over prolonged time periods remains poorly understood. We evaluated GleasonAI, an end-to-end...