There are more than 150 artificial intelligence applications approved by the U.S. Food and Drug Administration currently in use in a medical imaging setting, but widespread integration remains a pipe dream for many.
Dr. Mathias Goyen, Chief Medical Officer for Europe, the Middle East and Africa at GE HealthCare, offered Health Imaging his insight on what needs to happen in order to bridge the gap between where we are with AI in radiology and where we want to be.
Goyen’s outlook on the future of AI in the field is an optimistic one. The doctor believes that the technology will benefit both hospitals and patients while simultaneously addressing healthcare disparities affecting some of the world’s most vulnerable populations.
To get to that point, Goyen suggests keeping it simple.
“In order for healthcare professionals to use AI-powered devices and applications effectively, AI models need to be straightforward in terms of their features and functionality,” Goyen said.
This is especially important when it comes to integrating the technology into current workflows, Goyen noted. Implementing a tool that solves one problem at a time—an algorithm that detects pneumothorax on radiographs, for example—is helpful for that specific situation, but it is a drop in the bucket of a patient’s diagnostic journey. Goyen said having a platform that can integrate numerous tools into workflows would better serve its users.
“Instead of depending on independent tools to solve one problem at a time and building custom solutions to integrate one app with another on a case-by-case basis, it is more beneficial to use a platform that can embrace multiple tools in an umbrella,” he said.
He said that these platforms need the flexibility to support multiple vendors and to connect and adapt each of their offerings interchangeably. He used Edison AI Orchestrator as an example of a workflow management system that offers its users this sort of ease of use, as it can be integrated into radiologists’ current workflows. Platforms like this that do not require extra steps, clicks, logins, etc. to achieve a single function for a single exam are more likely to gain traction among users.
“To drive adoption, it is important that the technology designed to help productivity doesn’t add more work and complexity,” Goyen said.
Goyen is far from alone in his desire for AI simplicity. Several vendors at the annual RSNA conference in 2022 showcased platforms intended to address a widely expressed call for interchangeability in AI solutions.
The bottom line of these vendors, Goyen and any radiology professional whose career will inevitably be impacted by AI: integration should help solve radiology workflow problems, not create more.
In addition to her background in journalism, Hannah also has patient-facing experience in clinical settings, having spent more than 12 years working as a registered rad tech. She joined Innovate Healthcare in 2021 and has since put her unique expertise to use in her editorial role with Health Imaging.