Dreyer also serves as Vice President of Radiology at Massachusetts General Hospital, Director of Data and Information Science for the Departments of Radiology at Massachusetts General Hospital and Brigham and Women’s Hospital, and Associate Professor of radiology at Harvard Medical School.
The development of AI algorithms for medical imaging has been very piecemeal with applications developed by many separate companies. This forces hospitals or healthcare systems to contract with multiple small businesses to access these applications, which can be a nightmare for hospital IT departments. This has led several large radiology software companies to adopt an app store type of approach, offering multiple AI products from many vendors in one place that meet certain basic integration criteria.
This was the trend in how to access AI starting in 2019, but in the two years since COVID hit, this trend seemed to be changing at the 2021 HIMSS and RSNA meetings. The trend appears to be the transition by several radiology vendors from the app store concept to full partnerships with key AI developers to embed their algorithms directly into their PACS or enterprise imaging systems. This allows for a seamless interface and the AI coin then becomes just another tool in the options submenus without requiring another login or separate report or storage of the AI rated images.
Dreyer said it makes sense for PACS vendors to integrate the AI tool directly into their specific systems, but for healthcare systems that don’t use those vendors’ systems, there’s still a need for a “app store” like the market. He believes this model will be the way of the future, as long as AI can easily integrate into existing PACS with industry standards for health informatics.
“There are over 200 AI algorithms and it seems to follow Moore’s Law,” Dreyer said. “But it will be unmanageable to deploy 20 or 30 algorithms from 20 to 30 different vendors.”
He said it was a similar situation that existed with MP3 players and music download services in the early 2000s, and that it was unmanageable to download all those pieces onto a hard drive. But, the industry quickly changed and embraced the technology when it was standardized by Apple and its iPod concept and Apple Store. He said something similar will be needed in AI radiology.
“I really think that’s the only way a lot of AI will work. The other part is built-in AI. I think PACS systems will have built-in AI, EHR systems will have built-in AI , imaging modalities will have their own AI, but if you want independent AI developers solving problems in the radiology diagnostic cockpit, it has to go through an app store or two at some point. .
He said that’s a beauty of the Apple Store concept is that independent software developers can build their own apps to industry standards and then easily and make them available through a very public outlet. This is what he said is needed for the future of AI algorithms for healthcare.