Application of an open-source AI tool for quantitative quality control in whole slide images of prostate needle core biopsies - a retrospective study.

Abstract

The quality of Whole Slide Images (WSI) is a determining factor for proper diagnosis and prognosis, and for enhanced performance of Digital and Computational Pathology. In a context where diagnoses are increasingly quantitative, an automated, precise, effective, and rapid quality control is of paramount importance. PathProfiler is a deep learning-based software trained on prostatic tissue that provides a 'usability score' of WSI, evaluating its suitability for diagnosis. The Centro de Anatomia Patológica Germano de Sousa receives around 2500 prostate biopsies a year, distributed to Pathologists remotely. Hence, it becomes crucial to investigate PathProfiler's viability for automated and quantitative WSI quality control and its monitoring for diagnostic purposes. Thus, in the last 3 months of 2024, 226 H&E WSI from prostate needle core biopsies were retrospectively analysed by PathProfiler. Usability score, focus, and H&E quality were registered numerically. WSIs had a usability score mean of 0.6, focus score of 9.6, and H&E quality score of 9.1. A usability score of 0.5 or higher was obtained for 70% of the WSIs, while 30% required revision. Artefacts such as mounting media, dust, and folded tissue were the major artifacts detected; extra-prostatic tissue was also recognised as 'other artifacts,' since the algorithm provides a lower rating in our cohort. PathProfiler is a valuable tool in the automatic quality control of prostate biopsies, allowing quick evaluation and identification of cases requiring review before being handed over to the pathologist, and promotes recognition of opportunities to improve laboratory and clinical quality.