Interoperability across health systems could help unleash the power of artificial intelligence (AI) in medical imaging and healthcare overall, according to a keynote talk by Hal Wolf on May 26 at the Society for Imaging Informatics in Medicine (SIIM) annual meeting.
"I believe that this is a once-in-a-century moment, when it's possible to reimagine health and wellness by building on the digital transformation of the past 18 months and integrating a new generation of innovations that are now just beginning to move out of research and in the hospitals and clinics," said Hal Wolf, president and CEO of the Healthcare Information and Management Systems Society (HIMSS). "This is both our greatest opportunity and our most pressing responsibility."
Wolf noted that researchers are finding new ways to apply AI to medical imaging in order to streamline processes, target treatments more accurately, and find indications of anomalies sooner. However, "we are not where we need to be across the ecosystem to take full advantage of this progressing science," he said.
The lack of interoperability represents one of the biggest gaps among healthcare systems today, he said. As a result, one of HIMSS' most important priorities is to drive ecosystem-wide interoperability that enables the secure electronic exchange of health information, according to Wolf.
"This [issue] is particularly acute in medical imaging, where we capture so many different kinds of images from multiple sources in so many different formats," he said. "And so many different specialties and their patients alike have to sort through and ensure the capability of them coming together into a single record. The challenges are immense."
Thorny issues include making images available in a relevant manner within the patient record, standardizing how images and associated metadata are shared with specialists both in and outside the health system, and building systems that avoid the need for patients to bring a CD with their medical images when getting a second opinion or seeing a specialist.
Digital imaging has been estimated to take up as much as 90% of all healthcare data, according to Wolf. And imaging file sizes are growing faster than the cost of storage is falling.
"The enterprise is becoming the endpoint of imaging from a multitude of sources, which is part of the reason why imaging is the most expensive part of almost every healthcare system's IT infrastructure," he said. "This is where the work to promote standards-based interoperability in medical imaging that SIIM and HIMSS have collaborated on is so important."
A common language to describe data in images is important for enabling interoperability, but it may be even more important for unlocking the potential role of AI in healthcare, Wolf said.
"The tremendous increase in the use of digital health solutions combined with rapid advances in computer vision, machine learning, and virtual networks has dramatically accelerated the rate of progress in applying AI to healthcare," he said. "Capabilities that we thought lay decades in the future before the pandemic now seem achievable in a matter of years. Today we are just beginning to see what is possible."
One of the biggest questions that need to be addressed is how to use AI to fundamentally deliver safer and more effective treatments, he said.
"The answer begins by ensuring that AI is incorporated in the processes and workflows to augment your effectiveness as clinicians and specialists by streamlining your work and by providing you with information, insights, and recommendations that you need to make the best possible decisions about diagnosis and care," Wolf said.