These five AI tools for healthcare may be the best right now

These five AI tools for healthcare may be the best right now

The possibilities are as overwhelming as the promise.

Nearly 900 AI healthcare tools have been approved by the U.S. Food and Drug Administration (FDA), with experts encouraging their use to support patient care amid doctor shortages and hospital closures.

photo of Nigam Shah
Nigam Shah, PhD

But just because a tool is FDA-approved doesn’t mean it provides value, says Nigam Shah, PhD, chief data scientist at Stanford Health Care.

You want AI that improves care while lowering costs. “Are we going to make things better or worse?” said Shah, whose team has created AI review processes for health systems. “Nobody is regulating that.”

To help you navigate the wide range of AI options available, we asked Shah and other experts which AI tools are having the greatest impact in their hospitals.

To narrow down the list, we prioritized commercially available tools that have been cited by more than one expert and/or have received recognition from newly formed organizations aiming to standardize AI assessment processes, such as the Coalition for Health AI and the Health AI Partnership.

The result is a top five list of the best AI tools for healthcare currently available (until something better comes along).

1. LumineticsKern

What it is: AI that diagnoses diabetic retinopathy
For whom: Point-of-care providers, such as optometrists, ophthalmologists, general practitioners or endocrinology clinics

Nearly half of people diagnosed with diabetes also have diabetic retinopathy, the leading cause of blindness in the United States. About half of them don’t know it.

An eye exam can detect the disease, but patients may have to wait a week or more for results. The delay discourages follow-up care.

The FDA-approved LumineticsCore accelerates the diagnosis of diabetic retinopathy and macular edema and encourages more patients to visit an ophthalmologist.

After a retinal camera captures images, the system uses an AI-based algorithm and a separate diagnostic algorithm to detect and analyze biomarkers. It provides a diagnosis at the same appointment, so patients can leave with a referral to an ophthalmologist.

photo by Luminetics Core
LumineticsKern

Stanford researchers found that these patients are three times more likely to follow through on referrals than clinician-diagnosed patients. LumineticsCore also recently passed Stanford’s assessment, demonstrating “a clear improvement in the quality of care,” Shah said.

According to Jennifer Goldsack, founder and CEO of digital medicine nonprofit DiMe, the tool’s impact is even greater because the company assumes full liability, meaning healthcare providers won’t be held liable if the diagnosis is incorrect. There’s also a billing code that allows private insurance to cover the cost.

2. Shorten

What it is: An ambient listening aid for clinical notes
For whom: Doctors in any specialization who want to spend less time behind the computer

photo by Bruce Darrow
Bruce Darrow, PhD

No keyboards. No typing required. That’s how some doctors see the “hospital of the future,” according to Bruce Darrow, PhD, interim chief digital and information officer at Mount Sinai Health System. Doctors and nurses will do the talking, and the information will go where it’s needed.

That future is already taking shape with the rise of AI-based ambient listening tools like Abridge, which originated at the University of Pittsburgh Medical Center in Pittsburgh.

Doctors often collect documentation after they see patients, said Christian Carmody, MBA/MIS, chief technology officer at UPMC. “This helps them do that more efficiently, so they can spend more time caring for patients.”

photo by Chris Carmody
Christian Carmody, MBA/MIS

According to Abridge’s calculator, a hospital system with 200 physicians can expect to save approximately 85,000 hours per year (425 hours per physician) and save $1.7 million per year in physician turnover (due to burnout) by using Abridge.

“Almost all of the physicians have told us that it saves us time every day that used to be spent updating administrative notes,” Carmody said.

With patient consent, the tool records interactions and transcribes them in real time. An AI algorithm collects key pieces of information to create clinical notes, which clinicians review before adding them to patients’ medical records.

photo by Abridge Epic Devices
Shorten

Unique to Abridge is its use of explainable AI: when you highlight parts of the AI-generated note, the tool reveals where in the source transcript the information came from, making it verifiable. It also provides recordings and transcripts to patients.

Abridge is designed to work with Epic and other electronic health record systems and is also in use at Emory Healthcare, Yale New Haven Health, Sutter Health, Christus Health, and the University of Chicago.

3. Woebot

What it is: A therapy app supported by machine learning
For whom: General practitioners and mental health specialists

photo by Tarun Kapoor
Tarun Kapoor, MD

Mental health providers are struggling to keep up with demand, with some patients waiting weeks or months to see a specialist after requesting a referral, said Tarun Kapoor, MD, chief digital transformation officer at Virtua Health in Southern New Jersey.

Meet Woebot, a conversational mental health app that offers on-demand cognitive behavioral therapy (CBT). The app is only available to patients of partner healthcare organizations, which typically have more than 50 primary care physicians, according to Brad Gescheider, chief commercial officer at Woebot Health. The app has received breakthrough device designation from the FDA, which could help expedite (but not guarantee) FDA approval.

Woebot uses machine learning to understand patient messages and delivers pre-written (rather than self-generated) responses. Responses were created by the company’s team of clinicians and writers and are closely related to how clinicians actually respond during patient interactions.

“It doesn’t completely mimic human conversations with a therapist,” Kapoor said, but programming chatbots with generative AI can be too unpredictable.

photo of Mental health app Woebot
Woebot

Virtua Health prescribes Woebot to a limited group of patients with mild to moderate depression and anxiety, which CBT can help treat. Woebot is provided free of charge to Virtua Health patients. Hospitals and health systems pay a recurring monthly fee based on the number of primary care physicians they have.

So far, the app — a two-time MedTech Breakthrough Award winner — is proving popular and useful. More than 80% like using the app, a typical interaction lasts just 7 minutes, and 77% of interactions occur when providers are not working. One insurer, impressed with the findings, has asked for more data, which Kapoor hopes will lead to a reimbursement code.

“We need to show that people are using and benefiting from the tools we are recommending. We are not just increasing the cost of the system,” Kapoor said.

4. VBrain

What it is: An AI-powered tool for automatically contouring brain tumors
For whom: Radiotherapists

Before treating brain tumors, radiation therapists mark the precise location and mass of each lesion. By correctly identifying the “anatomical part” of each tumor, clinicians can deliver radiation more effectively, without damaging surrounding tissue, Shah said. But this manual process, known as tumor contouring, is extremely tedious and time-consuming.

“We literally take an image, place dots and connect them into an area,” Shah said.

Vysioneer’s VBrain speeds up contouring by an average of approximately 30% and improves contour accuracy by 12% compared to manual contouring. The FDA-cleared deep learning algorithm detects the three most common types of brain tumors: metastases, meningiomas, and acoustic neuromas.

Stanford researchers validated the tool in a study of 100 patients treated with stereotactic radiosurgery, which involves prior contouring.

“It does 80-90 percent of the task automatically,” Shah said. “It doesn’t create a significantly different outcome, but instead of a human being painstakingly drawing something on the image for an hour, they can do it in 15 minutes.”

5. GI Genius

What it is: AI that detects polyps during a colonoscopy
For whom: Gastroenterologists and endoscopists

Medtronic’s GI Genius is an AI-powered “backup camera” that’s FDA-cleared to find colon lesions during a colonoscopy. Once integrated with existing endoscopy equipment, the tool appears directly on the endoscopist’s monitor. It identifies suspicious polyps in real time and highlights them with green boxes.

GI Genius photo highlights colon lesions
GI genius

“It’s like a second set of eyes,” said Kapoor of Virtua Health. “The cost is not prohibitive, but look at the benefits.”

GI Genius uses deep learning algorithms that can analyze unstructured data sets, including images, and make predictions. In a 2020 study, the tool increased the detection rate (ADR) of adenomas (precancerous polyps) by more than 14%. The risk of colorectal cancer drops by 3% for every 1% increase in ADR. In a separate study, it analyzed polyps 82% faster than the endoscopist.

As a bonus, the software generates a report immediately after an endoscopy. “Doctors love that,” Kapoor said. “They don’t have to spend as much time on documentation.”