The National Science Foundation (NSF) is setting up seven new institutes for studying foundational AI. Two of the initiatives have healthcare as a prime focus. Key facts to know about the development:

NSF is allocating $140 million to get the multipronged expansion up and running. Foundational AI can train on mass jumbles of raw data to help solve a broad range of problems. The spend is part of a federal effort to “advance a cohesive approach to AI-related opportunities and risks,” according to the agency.

Traditionally, most NSF interests lie outside of medicine. Its sister agency the NIH handles those. The NSF has an annual operating budget a bit below $10 billion while NIH’s yearly allowance sits at around $45 billion (pending a bump to more than $51 billion proposed by the Biden Administration).

The two nascent institutes extending the NSF’s reach into healthcare are the AI Institute for Societal Decision Making (AI-SDM) and the AI Institute for Artificial and Natural Intelligence (ARNI).

1. AI-SDM will innovate human-centric AI to help manage public health emergencies as well as general disaster responses. The initiative is led by Carnegie Mellon University.
2. ARNI is tasked with using AI to build understanding of the human brain. The work is spearheaded by Columbia University and will corral researchers from neuroscience, cognitive science and AI.

These and the other five new AI institutes will study the technology’s emerging risks and harms. They’ll also seek to diversify the AI workforce in the U.S. The NSF says the $140 million allotment brings the agency’s total investment in AI institute to almost half a billion dollars, adding that much of the outlay is shared with other funding sources.

White House Office of Science and Technology Policy Director Arati Prabhakar:

“These strategic federal investments will advance American AI infrastructure and innovation, so that AI can help tackle some of the biggest challenges we face, from climate change to health. Importantly, the growing network of National AI Research Institutes will promote responsible innovation that safeguards people’s safety and rights.”  

Full announcement here.

Buzzworthy developments of the past few days.

• AI helps head off heart attack over a lifetime. Researchers have designed a deep learning model that can predict a patient’s risk of myocardial infarction and other cardiac events as particular dangers develop over time. What’s more, the model presents its predictions in colorful graphs that are easily understandable to most patients. The work was carried out at Cedars Sinai Medical Center and is running in NPJ Digital Medicine. Journal study here, institutional news here.
   
• Billing and collecting begets buying and selling. Generative AI Solutions Corp. of British Columbia, also known as “GenAI,” has bought Global AI Billing Corp. of Delaware City, Del. GenAI says its interest in Global AI Billing hinged in no small part on the latter’s 10% ownership stake in Remitz, Inc., a medical billing vendor whose proprietary AI software has helped it collect more than $100 million from more than 1 million patients and/or their payers. Read the announcement.  
   
• Pathology partnership nudges field toward AI adoption, digital transformation. Digital pathology vendor Paige is working with pathology AI supplier Visiopharm to offer pathologists a one-platform pathway into AI-aided cancer diagnoses. At the same time, the combo offering will help the field of pathology delve more deeply into digitization. Announcement posted here.
   
• AI is ‘uniquely poised’ to play matchmaker between vendors and providers. So suggests Optimize Rx Corp., which recently surveyed more than 120 physicians. Key finding: Specialists as well as PCPs are ever on the hunt for information about products and services relevant to their individual practices. However, most “still experience gaps in the amount, type, clinical focus and timing of brand information they need to make fully informed patient care and prescribing decisions.” Full survey results and overview here.
   
• Ready for proteins invented by AI? Non-natural proteins that just might fool Mother Nature are in the offing at the University of Toronto. There researchers have developed a system that does the job with generative AI. The initial hope is to help hasten the pace of drug development. “All our proteins appear to be biophysically real,” explains Prof. Philip Kim. “They fold into configurations that enable them to carry out specific functions within cells.”
   
• AI may assist doctors wishing for more treatment options. The digital platform supplier myTomorrows has launched a beta version of AI-based software that finds clinical trials into which they might steer hard-to-help patients. The capability could help frustrated patients—and by extension their stymied doctors—try investigational meds and other therapies before they’re cleared by the FDA. The product is called TrialSearchAI. Learn more.
   
• Providers mitigate ‘demand-capacity mismatches.’ In a business-news item posted May 8, legacy media dinosaur CBS News takes a look at three U.S. healthcare systems at which physicians are giving ChatGPT a go. What all three have in common is a desire to communicate with patients without involving doctors. “Patient messages in-and-of themselves aren’t a burden—it’s more of a demand-capacity mismatch,” gastroenterologist Patricia Garcia, MD, tells the network’s MoneyWatch operation. “Care teams don’t have the capacity to address the volume of patient messages they receive in a timely way.” Read the whole thing.
   
• Aspirational healthcare AI startup gets $4M infusion. Autonomize AI (Austin, Texas) has locked down $4 million in seed funds to help the company refine and commercialize its healthcare AI platform. Autonomize says algorithms running on the platform facilitate data-driven decisionmaking by clinicians, researchers and other end-users—regardless of technical prowess. Investment leader Skip Fleshman of Asset Management Ventures says there’s “a lot of talk about AI in healthcare,” but “extracting insights buried within unstructured medical data in various silos is essential for fostering breakthroughs in patient care and overall health outcomes.”
   
• AI writes well more than a dozen academic papers in well less than half a year. OK, it had a little help from a physician-scholar. Or vice-versa. “I’m a researcher and I publish articles on a regular basis,” radiologist Som Biswas tells the Daily Beast. “Those two things linked up in my brain: If ChatGPT can be used to write stories and jokes, why not use it for research or publication for serious articles?” Full story here, summary here.

• Digital Magazine: This is Enterprise Imaging - In this digital magazine we talk about how moving from multiple PACS to a single enterprise imaging system is busting siloes and deepening integration; challenges in radiology imaging and how radiologists are getting more done—better and faster—by using enterprise imaging; skyrocketing image volume and an increased need for collaboration across multiple and geographically diverse sites has made image management far more complex and why cloud is a solution to this; our latest addition to Sectra Enterprise Imaging portfolio—ophthalmology and why it is a game-changer for ophthalmologists.

• Beyond the impression: How AI-driven clinical intelligence transforms the radiology experience - In this session, Nuance CMIO Sheela Agarwal, MD, and Senior Product Manager Luanne D’Antoni explore innovations in radiology report creation and the role of automated impression generation.

• AI quality assurance models saving lives and millions in avoided med-mal - Unrecognized imaging findings are an unfortunate, but undeniable, part of radiology. New advancements in artificial intelligence (AI) and machine learning offer a critical safety net that is improving care and saving lives — as well as avoiding millions of dollars in potential medical malpractice costs.