An adapted & improved validation protocol for digital pathology implementation

作者：Ying-Han R. Hsu, Iman I. Ahmed, Juliana Phlamon, Charlotte Carment-Baker, Joyce Yin Tung Chan, Ioannis Prassas, Karen Weiser, Shaza Zeidan, Blaise Alexander Clarke, George Makram Yousef · 发表于：Seminars in Diagnostic Pathology · 年份：2025 · DOI：10.1016/j.semdp.2025.150905 · 被引用次数：8 · 研究领域：AI in cancer detection、Radiomics and Machine Learning in Medical Imaging、Biomedical Text Mining and Ontologies

Digital Pathology (DP) is transforming disease diagnosis by providing rapid and efficient analysis of tissue samples. However, ensuring the accuracy and reliability of diagnoses is crucial. This manuscript outlines University Health Network (UHN)'s journey towards the development of a customized validation protocol for implementing a digital workflow for primary clinical assessment. Drawing on guidelines from the Royal College of Pathologists (RCPath) UK and the College of American Pathologists (CAP), UHN has tailored its approach to accommodate the unique needs of its 14 subspecialty groups. Our protocol emphasizes pathologist-led self-validation, integration of diverse subspecialty cases, and a phased rollout with continuous monitoring. Additionally, the use of change management principles inspired by Leeds University (CCP) played a critical role in guiding the process, ensuring pathologists' comfort with digital workflows, and addressing subspecialty-specific challenges. This comprehensive validation protocol supports UHN's broader goals of leveraging DP for clinical practice while ensuring patient safety and data integrity.