Improving Access to Ultrasound at the Point of Care
By Matt Phillion
To a layperson, ultrasound as a tool generally calls pregnancy or abdominal imaging to mind. But as medical practitioners know, ultrasound is much more than that, though the technology is often underused because of its cost and limited number of trained specialists.
With more user-friendly technology that leverages artificial intelligence (AI), caregivers across departments—from emergency care to cardiology to primary care—could use ultrasound to quickly diagnose and treat patients suffering from a plethora of conditions. In fact, they’d be able to use it wherever the point of care happens to be, whether that’s in a hospital, in a clinic, or at a patient’s home.
Exo, a medical imaging software and device company, is working to develop a point-of-care ultrasound (POCUS) ecosystem that includes a handheld ultrasound device, workflow software that connects with EHRs to improve patient interaction and reduce provider burnout, and AI and education to simplify the technology.
“If you think about it, the first association with ultrasound is radiology and maternal-fetal medicine, maybe even stress echocardiology,” says Sandeep Akkaraju, CEO and co-founder of Exo. “Up until this point, it’s been a tool that has been in the hands of experts and power users.” That expertise is a scarce resource, Akkaraju notes, in comparison to a growing need for ultrasound at the point of care.
“We’re getting to a stage where every other specialty wants to use this tool at the point of care as opposed to a scheduled exam,” he says. “They may want to rule something out on a patient quickly, or be dealing with a patient who has walked in with terrible stomach pain. Ultrasound is used to rule out things—like kidney stones, gallstones, UTIs, ectopic pregnancies—while triaging a patient.”
This call for quick imaging to deliver better care is expanding, Akkaraju says. It’s driven in no small part by the evolution of training for medical professionals. “Over the last 10 or 15 years, everyone graduating from medical school has trained on ultrasound,” he says. “So they want to know, ‘Where is my machine?’ Instead of referring someone to radiology to get an answer, it’s changing.”
The trend isn’t limited to doctors. Nurses are increasingly identifying that they want an ultrasound to do an IV line, Akkaraju notes. And EMTs might ask for one if a patient has a pneumothorax they need to address right away. It’s even impacting acute care in the home; for example, a dialysis patient may require a new catheter.
“The options have become countless. This is a tool to make quick decisions, to rule things out or in. That’s why we’re focused on this segment, because POCUS situations need very quick answers to very tough questions,” says Akkaraju. “Every doctor or caregiver could adopt this technology.”
A transformation in healthcare
Akkaraju points to a massive shift in healthcare, toward democratization of care and better equity and equality. “Pre-COVID, care meant getting to that central facility. Now we’re seeing massive decentralization,” says Akkaraju.
For Exo, developing new imaging tools is exciting but presents some tough problems to solve. “It’s a very complex process,” says Akkaraju. For change to happen, “it needs to be simple to get the images into the system, document them seamlessly, review them, and get all the way to billing and EHR.”
This means working closely with EHR vendors as well as institutions themselves to ensure an uninterrupted flow of information. “Think of the apps we use on a daily basis,” says Akkaraju. Providers and patients are used to a certain level of seamless, unobtrusive applications in their day-to-day lives. But in healthcare, it’s still common for tools to rely on older operating systems or technologies.
“We talk to doctors and the first thing that comes across is that we’re all used to the same great apps … with finesse and ease of use, but in their work life, these folks are using software that is designed with antiquated technology,” Akkaraju says.
How AI plays a part
“We acquired an AI company called Medo that is in the process of building an AI around ultrasound,” says Akkaraju. While AI won’t take the place of human interaction, it can help speed processes up.
Akkaraju offers an example. “If you look at the procedure for scanning the thyroid, it’s a 45-minute procedure from scan to analysis to data input. You’re looking at every angle of the thyroid.” With AI, it’s instead possible to analyze the thyroid in real time.
“What they’ve been able to do is take a procedure that’s usually 30 to 45 minutes and make it no more than five minutes per lobe—and you’re done and getting answers,” he says. “The power of AI is simplifying the workflow.”
AI can also help with perennially challenging care scenarios, like trying to get an ultrasound on a newborn. “A week-old baby is going to be squirmy. They won’t stay still,” Akkaraju says. “You literally have 10 or 15 seconds to scan a baby.”
When looking for something like hip dysplasia, the current standard of care is to lift the leg and check for indicators. “But it has a low sensitivity, and it’s an aging art,” he says. “Imagine being able to use a device and move that sensitivity to more than 90% using ultrasound, with 10–15 seconds around each hip. That’s how we can change the standard of care.” Moreover, such an advance would allow for organizations to place AI-enabled devices “into the hands of every caregiver.”
It isn’t just AI that helps speed up the processes, Akkaraju notes. Exo has worked with physicians to ensure that scans can be documented quickly and efficiently based on preferred user experiences.
“We do our jobs in micro-moments and interactions,” he says. “We can do so much in under 60 seconds, and that’s how we’re looking at it. How do you handle documentation so you’re not just sitting at a desk? We’re building these solutions so it’s in your pocket. You could document and review while waiting in line for coffee.”
Practicing at top of license
With the increased demand for ultrasound, Akkaraju notes, AI can simplify the workflow to make access to this technology more universal for providers. “We need to build an ecosystem around hardware, workflow solutions, education: the ‘app-ification’ of medical procedures and exams to look at things with a fresh lens and provide simple solutions,” says Akkaraju.
This also helps highly skilled practitioners work at the top of their license, streamlining workflows to let them focus where patients need them most.
The initial goal, Akkaraju says, has been to make the POCUS devices incredibly user friendly. “It’s got to be simple [and] lower the learning curve,” Akkaraju says. “Those are the kinds of tools we want to build and offer the world.”
Exo pays a lot of attention to onboarding, too, and starts with devices that everyday users are familiar with, such as smartphone cameras. “Right from the interface up, we design it to look and feel like a smartphone camera. It’s got to look and feel like something you’re already using,” says Akkaraju.
This technology also needs to focus on productivity: What are the key jobs a nurse or physician may need it for, and how can the technology provide a workflow that the caregiver can master with a reasonably short amount of training? Here, the AI can function as an educational tool, guiding first-time users of the technology.
“I see AI as training wheels. The AI is there to help you in a pinch, but once you’ve learned how to do an exam using the technology, the training wheels can come off,” says Akkaraju. “It’s there to teach people how to do complex things and get comfortable with them.”
Akkaraju says POCUS technology can be a difference-maker not just for rural care centers or areas with care deserts that need more access to ultrasound, but also for larger systems where patients would otherwise need to travel to a few flagship hospitals to visit radiology centers. “It’s evolving very rapidly, and has the legs to continue to evolve and blossom into other fields,” he says.
Matt Phillion is a freelance writer covering healthcare, cybersecurity, and more. He can be reached at matthew.phillion@gmail.com.