The count of medical devices that are both equipped with AI and approved by the FDA is approaching 1,000. No time like the present to canvass experts for their opinions on which tools are the best of the bunch (which numbered 882 as of mid-May).

Medscape had the notion and commissioned the legwork. The outlet started the informal competition with a simple qualification for entry: Finalists had to credibly claim to leverage AI for improving care while reducing costs.

Next, it whittled the field to “commercially available tools that were mentioned by more than one expert and/or have earned recognition from newly established organizations striving to standardize AI assessment processes,” such as the Coalition for Health AI and the Health AI Partnership.

From there independent reporter Sarah Amandolare asked a number of AI experts at AI-deploying hospitals to name the products making the biggest impact at their respective institutions.

Here are the top five as described in her resulting article, which Medscape published July 22.

1. LumineticsCore

LumineticsCore from Digital Diagnostics speeds diagnosis of diabetic retinopathy and macular edema right at the point of care, resulting in patients following through on care referrals at three times the rate of conventionally diagnosed patients, Medscape reports. More:

‘What makes the tool even more impactful, says Jennifer Goldsack, founder and CEO of the nonprofit Digital Medical Society, is that the company assumes full liability, so providers won’t be held liable if the diagnosis is wrong, and there’s a billing code enabling private insurance reimbursement.’

2. Abridge

This product, an ambient listening tool for clinical notes, originated at the University of Pittsburgh Medical Center. Designed to work with Epic as well as other electronic health record systems, Abridge is also in use at Emory Healthcare, Yale New Haven Health, Sutter Health, Christus Health and the University of Chicago, Medscape notes.

‘Unique to Abridge is the 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.’

3. Woebot

A conversational mental health app providing on-demand cognitive behavioral therapy, Woebot uses machine learning to understand patients’ messages before delivering pre-written responses created by clinicians.

‘So far, the app—a two-time MedTech Breakthrough award winner—is proving to be popular and convenient. More than 80% [of patients] like using it, a typical interaction lasts just seven minutes, and 77% of interactions occur when providers are off-duty.’

4. VBrain

Vysioneer’s VBrain speeds up brain-tumor contouring by about 30%, on average, and improves contouring accuracy by 12% compared with manual contouring, Medscape reports. Its FDA-approved deep learning algorithm detects the three most common types of brain tumors—metastasis, meningioma and acoustic neuroma. Amandolare quotes Nigam Shah, PhD, chief data scientist for Stanford Health Care:

‘It does 80% to 90% of the [contouring] job automatically. It doesn’t create a materially different outcome, but instead of a human painstakingly drawing something on the image for an hour, they can get it done in 15 minutes.’

5. GI Genius

This Medtronic offering is an AI-powered “backup camera” that helps identify colon lesions during colonoscopy. Its software creates a findings report immediately after endoscopy.

‘In a 2020 study, the tool increased the adenoma (pre-cancerous polyp) detection rate by more than 14%. In a separate study, it analyzed polyps 82% faster than the endoscopist.’

Shah of Stanford underscores that an FDA approval does not assure a product will deliver value by improving care while containing costs. He points to LumineticsCore as an example of a tool that has shown “a clear increase in the quality of care.”

Read the rest at Medscape.

Nabla’s Integration with Epic: Everything You Need to Know - Nabla's latest blogpost covers their integration with Epic, and explains how clinicians can leverage Nabla to streamline clinical documentation on the leading EHR in the industry.

Buzzworthy developments of the past few days.

• CVS should probably stop making job applicants face video lie detectors armed with emotion AI. This week the company reached a tentative settlement in a pending class-action suit. The primary plaintiff alleged in 2023 that he wasn’t hired partly because, unbeknownst to him during his interview, his facial expressions would be analyzed by Affectiva software. This version of emotional-intelligence AI is attractive to recruiters, who can tap it to assign job candidates’ “employability scores,” according to the complaint. In coverage of the development, HR Dive notes that the settlement terms are fuzzy. Left unstated is how much CVS will pay out, if anything, and whether the company has agreed to stop using emotion AI—stealthily, at least—to size up job candidates.
   
• AI has an image problem in healthcare. Among physicians, for example, a majority worry the technology will come between them and their patients. That segment probably overlaps with the 40% who believe healthcare AI is more hype than substance. But cheer up. Images are eminently shapable. All it takes is some technological know-how and a little savvy with PR and marketing. So suggests the chief marketing officer of software supplier Athenahealth. “Just as important as building and evolving the technology,” writes the CMO, Stacy Simpson, “is our ability to market AI’s benefits to physicians and patients alike, to ensure that it’s leveraged to help reclaim what’s at the heart of exceptional care: a meaningful patient-physician relationship.” In a piece published July 23 by Fast Company, Simpson lists four simple steps to get there from here.
   
• When healthcare historians look back at our time, they’ll make much of AI’s role in spurring change. They might even call the before & after “seismic.” Or maybe a practical “tsunamAI” of change. And they’ll judge the end result a net positive—as long as today’s healthcare leaders succeed in harnessing its power. The predictor of the scenario is Lee Shapiro, cofounder and managing partner at 7wire Ventures and a member of the Forbes Business Council. Suggesting AI will reboot healthcare across clinical, research and administrative areas, Shapiro urges healthcare leaders to balance rewards with risk, partner up with suppliers and include end-users in model development.
   
• Lee Shapiro’s healthcare AI bullishness will get no argument from Saeed Hassanpour, PhD. A biomedical computer scientist, Hassanpour is the inaugural director of the newly created Center for Precision Health and Artificial Intelligence at Dartmouth Health in New Hampshire. “AI can revolutionize patient care by making it more predictive, preventive and personalized,” he tells Dartmouth’s news operation. “The future of AI in healthcare is incredibly promising.”
   
• But let’s not get carried away here. In healthcare, novel technologies like AI face more regulations, cybersecurity concerns and financial constraints than many if not most other industries. And in a recent McKinsey & Co. survey, more than 75% of health system executives said AI adoption is a priority while admitting they lack the resources to make it happen. The hurdles are noted by Alexis Kayser, healthcare editor at Newsweek. Upon interviewing a handful of physicians with these facts in mind, she found emotions decidedly mixed. “Like any new technology, we found that some docs really liked it,” academic cardiologist Thomas Maddox, MD, told Kayser. “Other docs said it wasn’t their thing. They pretty quickly said, ‘We aren’t gonna use this,’ and they moved away from it.” Read the rest.
   
• Seconding the note of cautious optimism is Sowmya Viswanathan, MD, MBA. Paradoxically, emerging technologies like AI can actually thwart digital transformation, warns the chief physician executive at BayCare in Florida. She reminds Healthtech Analytics that AI is still new—and still expensive. “You put ‘AI’ on the project, and the cost goes up tenfold,” Viswanathan says. What’s more, she adds, “If it’s going to add to the burden of provision of care, it will be a failure.” She’s not completely bearish on the technology—just realistic bordering on skeptical. Hear her out in an article filed by Xtelligent reporter Shania Kennedy.
   
• Microsoft is working with 2 academic health systems on generative AI for medical imaging. Together with Mass General Brigham in Boston and UW Health in Wisconsin, the Big Tech biggie is looking to refine multimodal AI foundation models for the benefit of patients, clinicians and administrators alike. (In order: shorter wait times for test results, burnout relief and bottom-line improvement.) Microsoft says the collaborations will facilitate research and innovation projects aimed at “delivering a wide array of high-value medical imaging copilot applications.”
   
• Mistakes, mishaps and failures. AI stumbles into its fair share of all three. The U.K.-based website Tech.co won’t let it forget a single incident. Among the greatest misses of 2024 so far: “New York City chatbot advises small businesses to break the law,” “Horrifying Willy Wonka experience captures the world’s attention” and “Air Canada defeated in court after chatbot lies about policies.” Catch up with these and other amusing and mostly harmless AI stumbles here.
   
• Recent research in the news:
   
  • Mayo Clinic: Researchers explore potential for AI to predict patients’ pain management support needs after surgery
     
  • MIT: AI model identifies certain breast tumor stages likely to progress to invasive cancer
     
  • NIH: Findings shed light on risks and benefits of integrating AI into medical decision-making
     
• Funding news of note:
   
  • Pearl raises $58M to help dentists make better diagnoses using AI
     
• From AIin.Healthcare’s news partners:
   
  • Cardiovascular Business: FDA clears heart rhythm AI that turns smartphones into medical devices