“AI is not meant to replace clinicians.”
I’m listening to Angela Adams, registered nurse and CEO of AI radiology follow-up management platform Inflo Health. Sharing why the company is a solution for clinicians, patients, and the health care industry as a whole, she says that artificial intelligence technology in health care is directed toward repairing chaos and harm. “It should replace all of the broken parts of health systems that we cannot continue to throw people at.”
It started when Adams, then a critical care nurse at Duke University Medical Center in North Carolina, received a text from a colleague and friend who had gone to the ER with severe abdominal pain. Adams’ friend was rushed into emergency surgery for acute appendicitis. While she was there, a radiologist caught a significant breast lesion, suspicious for malignancy, that required immediate follow-up. The radiologist documented it, but the finding vanished into the system.
“There was no communication to her primary care doctor,” Adams recalls. “And so she went 10 months [until] delayed diagnosis and treatment.”
A subsequent PET scan revealed metastatic breast cancer that had spread to her brain. Adams’ friend died a year and a half later in 2020, the same year Adams co-founded Inflo Health with CTO Nate Sutton.
Adams, whose background spans critical care nursing and health care AI leadership — long before the post-pandemic AI surge — is using artificial intelligence to improve preventive care and patient follow-up in radiology. It’s built around Inflo Health’s mission, “never miss a follow-up.”
Adam says that if Inflo Health had been in place, her friend would have received a text message stating that she had a follow-up due to the radiology findings, and that her doctor would have been notified as well.
How Inflo Health uses AI
When radiologists spot suspicious findings on scans ordered for something entirely different, those discoveries often get lost in the system’s chaos.
According to a 2015-17 study from the University of Washington and Lahey Hospital and Medical Center, approximately 50% of follow-up radiology recommendations aren’t adhered to (excluding mammograms), resulting in delayed diagnoses, legal risks and increased health care costs. A recent study found that missed follow-ups resulted in $3 million in annual health care costs.
Historically, radiology departments and hospital leadership have held conflicting views on delegation regarding patient follow-up communication. Adams says there’s often a breakdown of communication in the health care system, but there’s also a translation gap, like when a radiologist’s expertise doesn’t always get clearly translated to the doctor who ordered the study. Sometimes, findings can be misinterpreted or overlooked.
Modern imaging technology (often enhanced with AI) has gotten remarkably good at detecting unrelated abnormalities, which Adams describes as “incidentalomas” — findings that weren’t the original reason for the exam, like the breast lesion discovered during her friend’s appendicitis CT scan.
“We’re seeing a 40% increase in [imaging] detection alone,” Adams says. More findings mean more follow-ups that need coordination, overwhelming an already strained system.
While things have changed dramatically in health care, some aspects are dangerously behind, says Adams. In the past, a radiologist in a hospital could call the primary care physician and escalate patient details in the event of an emergency. Phone calls have now been replaced by automated workflows; yet, the technology is not necessarily benefiting patients receiving care, as potentially important information may be getting lost.
Inflo Health uses natural language processing and large language models to ensure that radiology follow-up appointments and recommendations are never missed.
First, the Inflo platform automatically scans imaging reports, such as X-rays, CT scans, MRIs and ultrasounds, to identify and extract relevant data and key points. While numbers regarding its accuracy aren’t available, studies have been conducted on whether this helps or hurts patients, depending on the doctor.
These recommendations are prioritized by urgent or high-risk situations, allowing care teams to identify which cases require attention first. This approach reduces manual tracking, which is where most follow-up appointments often slip through the cracks.
Inflo Health also integrates with existing workflow systems to monitor follow-ups in real-time, and tasks are escalated through text messages and provider platform notifications, providing visibility into staff efficiency.
Adams maintains a strict human-in-the-loop approach.
“AI isn’t replacing radiologists. It’s empowering them to deliver more reliable patient care,” Adams tells me.
According to company data, the automation handles 60% to 70% of follow-up cases from start to finish — the straightforward scenarios where patients respond to messages and complete their appointments. The remaining cases are escalated to human care coordinators, like those involving complex situations with multiple findings or oncology patients navigating treatments.
Patients and radiologists have greater visibility into the process, which can ultimately save lives, according to Adams.
“We’ve got them at the very tip top of the [workflow] pyramid, and AI automation is handling the bulk of it so that they can really focus their time, their knowledge, their energy on those really complex cases,” Adams says.
The effect of AI on health care
Certain types of AI were implemented into the health system as early as the 1960s, Adams tells me, including Apache (Acute Physiology and Chronic Health Evaluation) scores to predict mortality, as well as methods to predict 10-year cardiac risk — tests that are now embedded in clinical care.
But the traditionalist mentality of the health care system is holding humans back, and inevitably hurting them, she says.
Adams says that, when it comes to adopting technology, health systems are a decade behind other industries. Throwing more humans at the problem isn’t working, in her opinion. “At the end of the day, AI — and its underpinnings — are just math,” she says.
According to the philosophy behind Inflo Health, when technology is implemented to support humans rather than replace them, the outcomes benefit not only the clinicians and their teams’ communication streams but also the broader health care system.
The impact appears to be measurable: Working with Inflo Health, East Alabama Medical Center boosted follow-ups by 74%, according to the American College of Radiology. Additionally, Inflo Health reports that 125,000 total lives have been impacted to date.
This data supports something Adams emphasizes: “Technology’s highest calling is to give humans back the two most important things in life that you cannot buy, which are health and time.”
Don’t miss any of our unbiased tech content and lab-based reviews. Add CNET as a preferred Google source.
