Tailored for Real-World: A Whole Slide Image Classification System Validated on Uncurated Multi-Site Data Emulating the Prospective Pathology Workload

TLDR

Microscopic examination of H&E‑stained tissue is the standard for suspected skin cancer, but high inter‑pathologist discordance and rising biopsy rates demand greater efficiency and reproducibility. The study presents and validates a deep‑learning system that classifies digitized dermatopathology slides into four categories. The system was trained on 5,070 images from one laboratory and evaluated on an uncurated set of 13,537 images from three laboratories scanned with three different vendor scanners. With confidence‑based scoring, the system achieved up to 98% accuracy, compared to 78% without scoring, and is poised to accelerate skin‑cancer diagnosis, flag cases for specialist review, and enable targeted classifications.

Abstract

Standard of care diagnostic procedure for suspected skin cancer is microscopic examination of hematoxylin \& eosin stained tissue by a pathologist. Areas of high inter-pathologist discordance and rising biopsy rates necessitate higher efficiency and diagnostic reproducibility. We present and validate a deep learning system which classifies digitized dermatopathology slides into 4 categories. The system is developed using 5,070 images from a single lab, and tested on an uncurated set of 13,537 images from 3 test labs, using whole slide scanners manufactured by 3 different vendors. The system's use of deep-learning-based confidence scoring as a criterion to consider the result as accurate yields an accuracy of up to 98\%, and makes it adoptable in a real-world setting. Without confidence scoring, the system achieved an accuracy of 78\%. We anticipate that our deep learning system will serve as a foundation enabling faster diagnosis of skin cancer, identification of cases for specialist review, and targeted diagnostic classifications.