Proscia, a provider of artificial intelligence (AI)-enabled digital pathology software, has released the findings of a new study on the first deep-learning system with proven accuracy in real laboratory environments. Published in Scientific Reports, a journal from Nature Research, the study is the largest AI validation study conducted in pathology to date and supports the growing impact of AI in cancer diagnosis.
The Proscia’s study, entitled “Tailored for Real-World: A Whole Slide Image Classification System Validated on Uncurated Multi-Site Data Emulating the Prospective Pathology Workload”, describes a deep-learning system that achieves 98% accuracy in classifying whole-slide images of skin biopsies in real laboratory settings. To achieve this consistent performance, the system was designed using high-quality and diverse de-identified data. It was developed using thousands of images from the Dermatopathology Laboratory of Central States, one of the largest dermatopathology laboratories in the USA, and tested on an uncurated set of 13,537 images from Cockerell Dermatopathology, Thomas Jefferson University Hospital and University of Florida to account for the wide variety of diseases seen in practice. Pictured is a juvenile xanthogranuloma showing distinctive touton giant cells.
Proscia conducted this study to validate its DermAI application and bring AI into the pathology laboratory. Launched in June 2019, DermAI is the first in a series of AI applications on Proscia’s Concentriq platform. By using deep learning to classify hundreds of variants of skin diseases automatically, DermAI is driving much-needed confidence, quality and efficiency gains with capabilities including intelligent workload balancing, case prioritisation, automated QA, and 100% AI re-review. This first-of-its-kind pathology solution can reduce costly errors and process-added volume to meet the global cancer burden.