Mobile Ultrasound Devices Give URMC Clinicians a New Tool for Decision Support

The University of Rochester Medical Center will soon be deploying AI-enhanced devices to help clinicians improve point-of-care management

The University of Rochester Medical Center (URMC) will be deploying AI-enhanced mobile ultrasound devices later this year to help clinicians improve point-of-care treatment.

The upstate New York health system is partnering with Connecticut-based Butterfly Network on the program, which seeks to improve clinical decision support (CDS) and eliminate time spent sending patients off for imaging by giving clinicians the tools they need to do the work themselves.

URMC will begin training staff on how to use the handheld digital health devices this summer, with plans to roll them out to primary care providers, home care nurses, and second- and third-year medical students in the fall. Officials are also planning to include the devices in research projects and enterprise-wide strategies.

"Oftentimes, practitioners call upon imaging modalities to confirm a hypothetical diagnosis and treatment plan," says Michael F. Rotondo, URMC's CEO and vice dean for clinical affairs at the university's School of Medicine and Dentistry. "With Butterfly, care teams can introduce point-of-care ultrasound as part of the initial assessment. A small, stapler-sized imaging probe paired with a smart phone becomes an advanced bed-side imaging tool. Care teams can use this information to confirm or exclude their diagnosis improving the overall care."

The project points to a growing interest among healthcare organizations in digital health tools and platforms that give front-line care providers on-demand access to CDS resources, so they can make better decisions at the point of care instead of scheduling tests and unnecessarily extending the time between examination and diagnosis (as well as diagnosis and treatment).

Portable Point Of Care Ultrasound (PPOCUS) "is undergoing a revolution similar to what computing technology experienced in the 20th century, an acceleration in the development of portable, efficient, and affordable systems," wrote Patrick Lindsay, MB, BS; Lauren Gibson, MD; Edward A. Bittner, MD, PhD; Marvin G. Chang, MD, PhD, all of Massachusetts General Hospital, in a 2020 research article in the newsletter of the Anesthesia Patient Safety Foundation.

"Ultrasound technology has evolved to the point that portable ultrasound devices can now fit seamlessly in a clinician's back scrub pocket at a price point as low as $2,000 USD," they wrote. "Furthermore, the imaging capabilities far exceed many of the best ultrasound machines of decades ago."

It also hits upon a lesser-known aspect to the digital health movement. While the nation is enamored with technology that allows people to access healthcare from their homes, cars, workplace, and other locations, they're paying less attention to the tools and platforms being developed for use in the health system.

In places like URMC, which serves a wide swath of rural New York, those devices and services can literally be lifesaving, especially if access to care is limited and it takes days or weeks to gather all the information necessary for an accurate diagnosis.

"While highly valuable, traditional ultrasound equipment can be rather unwieldy," says Rotondo. "It's cart-based and requires training whereas this much smaller bedside tool is highly portable and AI-empowered, to support caregivers of all kinds, including paramedics, nurses, physicians, and home healthcare workers. Our vision is that this will greatly aid in early detection and diagnosis and will accelerate treatment."

"Our intent is for Butterfly to integrate fully with our existing enterprise systems, workflows, and EMR vs. operating as a capability in a silo," he adds. "This is different and exciting; we're approaching imaging from an enterprise-wide lens—integrating it, optimizing it, and scaling it to improve care delivery across the institution."

The differences between digital health adoption outside the health system and within the health system are clear. Whereas consumer-facing technology focuses more on engagement and ease of use, tools and platforms for in-patient use must meet strict protocols associated with clinical care, particularly with regards to accuracy and reliability. Clinicians have long been wary of consumer-facing technology because they don't trust the data, and they're even more critical of technology they're using inside the four walls of the hospital.

Mobile clinical-grade devices, meanwhile, were in many cases first developed for uses outside the hospital setting, such as remote locations, disaster sites, and underdeveloped countries, where they could bring clinical services to first responders or medical workers. With the onset of the pandemic and the need to reduce contact between infected patients and care providers, those devices are now moving into the healthcare setting.

They're also being enhanced with AI technology.

"AI is a tool to support informed decision-making," says Rotondo. "The physician or care worker is ultimately charged with using her judgment before issuing any diagnosis and developing a plan. A diagnosis always needs to be done with judgment and consideration of likely alternate possibilities. In many cases, particularly in the early stages, it's tough to know exactly what's going on with a patient. In these cases, the best you can do is suggest a probable diagnosis and plan of action that might involve further tests or simply keeping an eye on the patient's progress."

Digital health advocates say AI has the potential to help clinicians make those decisions better and more quickly by automating the numbers-crunching and connecting the dots.

"The combination of cloud-based computing with advanced artificial intelligence has unearthed unprecedented opportunities for healthcare, and we believe the potential for AI in medical diagnostics is in its early phases," says Todd Fruchterman, MD, Butterfly Health's president and CEO. "We believe the combination of bedside imaging, powered by AI can help solve the problem of inadequately informed medical decisions. This is what we're passionate about. As with all applications of technology in healthcare, it needs to be done thoughtfully."