Related Experiment Video
Updated: Jun 9, 2025
07:46
Author Spotlight: Advancing Reproductive Immunology with a Protocol for the Quantitative Evaluation of Endometrial Immune Cells
Published on: October 13, 2023
1.2K
Public evidence on AI products for digital pathology.
Gillian A Matthews1, Clare McGenity2,3, Daljeet Bansal2
1Leeds Teaching Hospitals NHS Trust, Leeds, UK. gillian.matthews13@nhs.net.
NPJ Digital Medicine
|October 25, 2024
Summary
Public evidence for artificial intelligence (AI) in digital pathology is scarce. A review found most AI products lack robust validation studies, highlighting a need for transparency in this rapidly evolving field.
Area of Science:
- Digital pathology
- Artificial intelligence (AI) in healthcare
- Medical device regulation
Background:
- Novel artificial intelligence (AI)-based products for digital pathology offer potential benefits but often lack sufficient publicly available evidence.
- Assessing the independent evidence for these AI tools is crucial for informed adoption and decision-making.
Purpose of the Study:
- To identify and examine the publicly available evidence for AI-based products in the digital pathology market within the European Economic Area/Great Britain (EEA/GB).
- To assess key product features, including regulatory status and the availability of validation studies.
Main Methods:
- A systematic review was conducted to identify AI-based digital pathology products available in the EEA/GB market.
- Data extraction focused on regulatory approval, intended use, and the presence of peer-reviewed internal and external validation studies.
- An online register was created to compile and present the gathered public evidence.
Main Results:
- Twenty-six AI-based digital pathology products met the inclusion criteria.
- A majority (24/26) received regulatory approval as General in vitro diagnostic (IVD) medical devices through self-certification.
- A significant gap exists in published validation evidence, with only 38% having internal and 42% having external peer-reviewed studies.
Conclusions:
- While many AI digital pathology products are approved, there is a notable lack of independent, peer-reviewed validation data.
- The developed online register aims to enhance transparency and support informed decision-making regarding these novel medical devices.
