Quality control stress test for deep learning-based diagnostic model in digital pathology

TLDR

Digital pathology enables computational analysis of histological slides, yet slide heterogeneity from staining, thickness, and processing artifacts complicates routine tasks. The study aims to digitally reproduce major artifact types and stress‑test deep‑learning diagnostic models with synthetic artifacts to assess their impact during clinical validation. Using six datasets from four institutions scanned with different systems, the authors systematically evaluate how these artifacts affect the accuracy of a pre‑trained deep‑learning model for prostate cancer detection. Artifacts of any severity substantially degrade model performance, underscoring the need for strategies to prevent diagnostic accuracy losses.

Abstract

Digital pathology provides a possibility for computational analysis of histological slides and automatization of routine pathological tasks. Histological slides are very heterogeneous concerning staining, sections' thickness, and artifacts arising during tissue processing, cutting, staining, and digitization. In this study, we digitally reproduce major types of artifacts. Using six datasets from four different institutions digitized by different scanner systems, we systematically explore artifacts' influence on the accuracy of the pre-trained, validated, deep learning-based model for prostate cancer detection in histological slides. We provide evidence that any histological artifact dependent on severity can lead to a substantial loss in model performance. Strategies for the prevention of diagnostic model accuracy losses in the context of artifacts are warranted. Stress-testing of diagnostic models using synthetically generated artifacts might be an essential step during clinical validation of deep learning-based algorithms.