Healthcare AI Guy Weekly Newsletter | 12/19

Our Picks ✨

Highlights if you’ve only got 2 minutes…

1/

New Google Health AI model

Google just launched ‘MedLM’, a suite of medically tuned models that combine the power of large AI models with medical knowledge, enabling the technology to help:

• answer medical questions

• summarize medical information

• generate insights from unstructured data

• and more…

MedLM was built on top of Med-Palm2 and comes in two offerings: a larger version optimal for complex workflows, and a medium version best used for “scaling across tasks.” Several healthcare companies have been using MedLM, including Augmedix, HCA Healthcare, and Accenture. This is a further step in asserting Google’s dominance and assisting clinicians and researchers in conducting intricate studies, summarizing clinician-patient interactions and more with AI. (link)

2/ 

Healthcare groups to deploy more AI in 2024

58% percent of healthcare executives said their organization will implement or acquire a generative AI solution in 2024, according to a recent KLAS report. KLAS gathered insights from 66 healthcare executives (not a huge sample size but useful insights nonetheless) on current status, upcoming initiatives, and existing hurdles in AI adoption. Six key things from the report:

1. Adoption varies: Smaller orgs lag, while larger ones, including payers, are already using AI tools from big players like Epic, Google, Nuance, and OpenAI.

2. Current landscape: Just 25% have embraced generative AI so far.

3. Key motivation: The primary drive is to boost efficiency.

4. Enthusiasm vs. uncertainty: Execs are excited about AI, but many are still figuring out its best application.

5. Defined goals: For those with a clear strategy, enhancing efficiency in documentation, patient communication, and workflow is top priority.

6. The hurdles: Accuracy, cost, and security are the main concerns.

While only 25% of healthcare groups are implementing AI, expect that to more than double next year as leaders see opportunities to automate clinical documentation and improve patient communication, among other uses. All eyes are on 2024 being a HUGE year for AI in healthcare. (link)

3/

HHS AI Transparency Rule

AI transparency is in the headlines again with HHS’ publication of a rule that sets transparency requirements for predictive AI algorithms used in healthcare. The rule, which requires more transparency, has the goal of helping providers understand algorithms’ potential risks.

Specifically, EHR vendors will have to disclose their training data sources, explain how their models were tested, and share a strategy for identifying and controlling risks. The rule also includes a range of new provisions directed toward data interoperability and information blocking. The regulations do not apply to AI tools health systems have crafted on their own, however, a practice that has become more common. (link)

Tool Box 🧰

Latest on business, consumer, and clinical healthcare AI tools and partnerships…

🔧 Practice Better AI charting assistant: Practice Better, an all-in-one practice management software platform, launched its take on an AI charting assistant. The intention of the AI tool is “to give practitioners back critical face time with patients.” The tool securely records, transcribes and summarizes client sessions in a practitioner’s charting workflow+ list. (link)

🔧 AI made from living human brain cells: Researchers at Indiana University have developed Brainoware, a biocomputing system using living brain cells for basic speech recognition. With the promise of lower energy consumption than silicon chips, this breakthrough poses implications for the future of AI efficiency. (link)

🤝 HCA Healthcare + Google + Augmedix: Nashville-based HCA Healthcare has been conducting pilots with Google’s healthcare AI tools and using Augmedix in four emergency department hospital sites over the past several months. Google’s Augmedix platform, powered by natural language processing and Google Cloud’s MedLM on Vertex AI, transforms clinician-patient interactions into medical notes. (link)

Deal Desk 💸 

Spotlight on latest capital raises, M&A and investments…

📈 Totus Medicines: a biotech that uses covalent libraries and AI tools for drug discovery and is based in Emeryville, CA raised $65M in Series B funding. DCVC led and was joined by North Pond Ventures, Camford Capital, and the Regents of the University of Minnesota. (link)

📈 Twin Health: an AI-powered metabolic care startup based in Mountain View, CA raised $50M in Series C funding. Temasek led and was joined by insiders Iconiq Growth, Sofina, Peak XV, and Helena. (link)

📈 Medefy Health: a mobile-first healthcare benefits navigation platform raised $10M in Series A funding. They will use the funds to bolster AI capabilities. Mercury Fund led with participation from Advantage Capital. (link)

📈 BetterYou: an AI-based healthy habits companion startup based in MN raised $6M in Series A funding. Capital Midwest led with additional funding from Mairs & Power, Great North Ventures, North Coast Ventures, and NelNet. (link)

Other Relevant News 🔍

News, podcasts, blogs, tweets, resources, etc…

• AMA survey showing doctors are split on AI impact to them (link)

• Rock Health 2024 trends along the innovation maturity curve (link)

• Leaders at Epic say the company accelerated AI adoption in 2023 (link)

• Elevance Health global CIO optimistic about AI (link)

• Microsoft exec: physicians need to understand GPT-4 for safe use (link)

• Epic utilizing predictive models allowing clinicians to optimize OR time (link)

• Sanford Health developing AI in rural health to improve care (link)

• 8 things to know on healthcare spending as it reaches $4.5 trillion (link)

Visuals of the Week 📸

Funny memes, cool pics, and interesting data from around the web…