Artificial intelligence and digital pathology: where are we now and what are the implementation barriers?

Abstract

Artificial intelligence (AI) has garnered significant public attention over the last few years and has become part of our everyday lives. This is no different in histopathology, with vast amounts of private and public sector investment into digital reporting and AI development. There are now several companies with commercially available AI decision support tools ready for adoption by pathologists in digitally mature laboratories. Many more pathology-oriented AI tools, with a variety of uses across the patient pathway, are on the horizon and include the possibility of multimodal integration. However, for AI in pathology to be used effectively and its benefits realized, there needs to be widespread ‘business as usual’ use of digital laboratory workflows and digital reporting. The NHS is currently undergoing a national pathology transformation programme that includes targets for digital reporting and AI ready capabilities for all laboratories in England. In this article, we provide an update on the current progress of AI development in histopathology including the AI tools currently available and potential future applications. We also discuss the ongoing implementation of digital pathology services in the NHS and highlight the barriers to building a strong foundation for AI tool deployment. This builds upon, and complements, our previous article on these issues.