Healthcare AI Guy Weekly | 6/4

OpenAI’s growing data empire and healthcare foreshadowing, Atropos Health series B bridges evidence gaps, Google AI search errors highlight potential healthcare fears, and more

Welcome back folks —

We have some exciting news…. the first-ever in-person Healthcare AI Guy event!

For those in the Bay Area: JOIN US for an evening of Lightning Talks focused on Healthcare and Generative AI with fellow founders, engineers, researchers, product managers, clinicians, executives, operators, and investors 🧠. I am co-organizing with my friends at HealthTech Hang and AI Product Creators. If you are interested in sponsoring or speaking let me know!

Dr. Graham Walker, who leads AI/Innovation @ Kaiser Permanente Medical Group will be one of our keynote speakers and we’ll be announcing more soon…!

Click here to register by EoD Wed, June 19 for super-early bird tickets!

Now, let’s get straight to it:

  • OpenAI’s growing data empire and healthcare foreshadowing

  • Atropos Health series B bridges evidence gaps

  • Google AI search errors highlight potential healthcare fears

  • 8 new tools/partnerships, 8 funding updates & link-worthy content

Our Picks

Highlights if you’ve only got 2 minutes…

1/

OpenAI’s growing data empire and healthcare foreshadowing

As we predicted for 2024, IP battles and licensing deals in the AI space will begin to heat up. One of the leaders in the race, OpenAI, is pushing on the gas to grow its data empire and announced a series of new data partnerships recently:

  • News Corp: OpenAI signed a $250M deal over 5 years paying News Corp for access to WSJ, NY Post, The Sun etc content (link)

  • Sanofi: Sanofi‘s partnership with OpenAI will give OpenAI access to proprietary data to develop AI models for its biopharma models. (link)

  • Vox Media: The Vox partnership gives ChatGPT training access to content from brands like Vox, The Verge, New York Magazine, Eater, and more. (link)

  • The Atlantic: The Atlantic’s content will be more discoverable within OpenAI products. (link)

Despite pushback on AI firms and their use of training data, media companies and other organizations with proprietary data are finding few viable alternatives other than conceding. AI models run on data, and only a few entities will win the foundational AI model race (i.e. OpenAI), so companies like News Corp, Vox Media, and The Atlantic, along with Sanofi in drug development, see this as the best path forward and a way to monetize their data. Similar partnership deals will become increasingly important in healthcare for those trying to build the best healthcare-specific LLMs — enabling companies with top-tier engineers, models, and computing power to combine with the best proprietary datasets available.

2/

Atropos Health series B bridges evidence gaps

Healthcare doesn’t have any shortage of data, but it’s been a massive challenge converting that data into actionable evidence. That is where Atropos Health comes in. Atropos transforms real-world data into real-world evidence by quickly producing retroactive observational studies using an evidence platform of 200M+ patient records, as well as a partner network packed with big names like Mayo Clinic and Clarify. Atropos Health now raised $33 million in a Series B funding round to enhance its automation of personalized real-world evidence using generative AI. The funding will support:

  • Expansion of the Atropos Evidence Network, focusing on oncology and other specialty areas.

  • Development of value-based care decision-making tools.

  • Increased investment in the ChatRWD™ application, an LLM-independent framework built to eliminate hallucination risk while slashing the time needed to produce reports using a chat-based interface.

Atropos is building a future where every clinician and every scientist can make decisions with the best possible evidence available and we are here for it! (link)

3/

Google AI search errors highlight potential healthcare fears

Google's recent unveiling of its new AI capabilities for search, a move to protect its search dominance, has sparked controversy due to a series of errors and untruths. The AI feature, known as AI Overview, has been providing incorrect answers, such as suggesting glue as a pizza ingredient and recommending the consumption of rocks for nutrients… Google stated that the majority of AI Overview queries result in "high-quality information" and that the shared examples are uncommon but when 2+ billion people use Google for reliable information this is obviously a problem.

This is a great example of what NOT to do in healthcare. We need to make sure our healthcare tech is cutting edge, but not cutting corners. Just this week the nation’s largest nurses union went on strike to demand that AI tools used in healthcare be proven safe and equitable before deployment. AI will have growing pains but if we want these AI systems to augment clinical work and gain public trust the systems need to be ready for primetime. (link)

Tools & Partnerships 🔧

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

TOOLS

  • Epic unveils AI Trust and Assurance Suite: Epic unveiled an open-source AI validation tool called the AI Trust and Assurance Suite that will be free and available through the public coding website GitHub where health systems can add it to their EHR systems. Providers can use the tool to validate AI models that integrate with EHR systems, including models developed by Epic and other organizations. (link)

  • AI-driven robotic system to perform a head transplant: Neuroscience startup BrainBridge just introduced a new concept for an AI-driven robotic system capable of performing head transplants — aiming to help patients with terminal illnesses and paralysis. The transplant tech would graft your head — your memories, your abilities — onto a donor body. The startup wants to try its mechanized system on a real-life person within eight years. (link)

  • Clover Health marketing AI software to providers, payers: The AI tool will be available to all of Clover’s Medicare Advantage payers and providers to minimize healthcare costs. According to Clover, members who used the AI tool were 37% “more likely to have had their blood pressure controlled, 21%” were “more likely to have a colorectal screening and 12%” were “more likely to have had a breast cancer screening.” (link)

  • AI brain implant language breakthrough: Researchers at UC San Francisco just developed a brain implant that utilizes AI to help a stroke survivor communicate in both Spanish and English, switching between languages seamlessly via brain activity. (link)

  • AI to help assess patient acuity in emergency departments: A large language model could help assess patient acuity in emergency departments, according to a study published in JAMA. The LLM, when given pairs of patient histories extracted from emergency department documentation, correctly picked the higher acuity patient in 89% of scenarios – comparable to a subset that used a physician reviewer. (link)

PARTNERSHIPS

  • Banner Health + Qventus: Banner Health, based in Arizona, partnered with Qventus to implement AI technology for optimizing OR scheduling and operations. The pilot at four facilities resulted in significant improvements, adding 2.1 cases per OR per month and releasing 359 hours of OR block time monthly. Banner plans to expand the AI solution enterprise-wide. (link)

  • Notable + Beacon: Notable signed a new partnership with Beacon Health System to streamline the process of self-scheduling procedures and tests by using AI to scan its EHR for any pending orders, automatically reaching out to those patients via personalized messages, then enabling easy scheduling for any needed appointments. (link)

  • Google + Highmark Health/Allegheny Health: Pittsburgh-based Allegheny Health Network rolled out an LLM, named Sidekick, to all of AHN’s 22,000 employees, as well as all 44,000 employees employed by its parent company Highmark Health. Highmark and AHN developed the tool with Google. (link)

Deal Desk 💸 

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

FUNDING

  • LabGenius, a London, U.K.-based machine learning platform designed for the discovery of new antibodies, raised $44.5M in Series B funding. M Ventures led the round and was joined by Octopus Ventures, LG Corp, and existing investors Atomico, Kindred Capital, Lux Capital, and Obvious Ventures. (link)

  • Atropos Health, a Palo Alto, CA-based pioneer in translating real-world clinical data into high-quality personalized real-world evidence for care, raised $33M in Series B funding. Valtruis led with participation from new strategic investors including Cencora Ventures, McKesson Ventures, and Merck GHI Fund, along with existing investors Breyer Capital, Emerson Collective, and Presidio Ventures. (link)

  • Valar Labs, an AI-powered cancer care prediction tool has closed a $22 million Series A round led by DCVC and Andreessen Horowitz, with Pear VC participating. (link)

  • Plenful, San Francisco-based pharmacy automation startup Plenful completed a $17M Series A funding round. Q Ventures led, with participation from Bessemer Venture Partners, Mitchell Rales, Susa Ventures, and existing investors. (link)

  • ExpressionEdits, a Cambridge, U.K.-based biotechnology company using AI and other technology to optimize protein expression, raised $13M in seed funding. Octopus Ventures and redalpine led the round and were joined by BlueYard Ventures, Wilbe Capital, Acequia Capital, Amino Collective, and Hawktail. (link)

  • HoneyNaps, an AI-powered sleep disorder analysis startup, added $11.6M in a Series B round. Investors were Korea Industrial Bank and Hi Investment Partners. (link)

  • Videra Health, an AI-driven mental health assessment platform closed a $5.6M Seed II round led by Peterson Ventures. (link)

IPO

  • Tempus AI, an AI precision-medicine startup filed for its long-awaited IPO, and will be looking to raise $100M. Since 2015, Tempus has reportedly built the world’s largest library of clinical and molecular data, as well as an operating system that allows providers and life sciences orgs to surface insights from the data. The company amassed $1.3B in funding before heading to the public markets, and will now be looking to grow past its roots in oncology to bring AI-enabled diagnostics and personalized medicine to areas like cardiology and radiology. (link)

market snapshot as of 6/4/24

Other Relevant News 🔍

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

  • AI cancer screening not yet covered by medicare, private insurers (link)

  • Musk’s Neuralink aims to enroll three patients to evaluate its brain implant (link)

  • Specific AI approach enhances suicide risk detection in behavioral health (link)

  • Epic exec says AI still a long way from speeding up prior auths (link)

  • Report: health systems implementing AI should have strict oversight (link)

  • Europe’s landmark AI Act passed (link)

  • Explaining AI like I’m 5 (link)

Visuals of the Week 📸

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

healthcare lagging (on brand) but trending in the right direction!

How early were people working out and wearing bikinis!? This Roman era mosaic is from the Villa Romana del Casale, built around 320 AD in Sicily.

That’s it for this week friends! Back to reading — I’ll see you next week.

Stay classy,

— Healthcare AI Guy (aka @HealthcareAIGuy)

PS. I write this newsletter for you. So if you have any suggestions or questions, feel free to reply to this email and let me know