Healthcare AI Guy Weekly Newsletter | 7/25

Our Picks ✨

Highlights if you’ve only got 2 minutes…

Teladoc + Microsoft

Teladoc doubled down on its Microsoft partnership to integrate AI tech into its virtual care platform.

Specifically, the expanded partnership will integrate Microsoft’s AI offerings, including OpenAI’s ChatGPT and Nuance’s DAX Express, to automatically document patient encounters for final review and signoff by clinicians. The aim is to help alleviate administrative and documentation burdens for physicians.

Nuance’s DAX Express, an AI-powered clinical documentation application, will work with GPT4 through the Azure Open AI service to also report encounters to other clinicians, enabling improved care coordination, follow-up, and care beyond the immediate encounter.

This continues the wave of more prominent healthcare players starting to integrate AI technology into their workflows, such as Epic, Northwell Health, and others have recently done. As Microsoft continues its AI dominance, healthcare companies hope these tools can help reduce burnout and improve the care delivery experience. I’m hopeful that will be the case. (link)

Benchmarking AI in healthcare

In an effort to establish a reliable, trusted way to benchmark and evaluate medical AI models, MLCommons, an engineering consortium focused on building tools for AI industry metrics, has architected a new testing platform called MedPerf.

According to MLCommons, MedPerf can evaluate AI models on ‘diverse real-world medical data’ while protecting patient privacy. The goal here is to enhance the effectiveness of AI, reduce bias, and build public trust in AI's use in healthcare.

MedPerf was the result of a two-year collaboration including both industry and academia — over 20 companies and more than 20 academic institutions gave feedback (incl. Google, Amazon, Stanford, MIT, etc.). Moves like this are necessary measures to accelerate the adoption of AI in healthcare, given it’s a highly regulated industry with a lot more safeguards than most (e.g. HIPAA) — so thank you MLCommons! (link)

Seven AI companies agree to watermarking for AI safety in the US

The White House came together with seven AI companies to agree to a set of AI safety rules. The companies are Amazon, Google, Meta, Microsoft from big tech, and Open AI, Inflection, and Anthropic from the AI front.

The companies voluntarily pledged to implement watermarking for AI-generated content. The move aims to enhance safety and combat deep-fake images and audio. However, these new standards are only voluntary, so there won’t be any actual ramifications of ignoring them…

With the media attention and statements, it feels like the White House has really started getting the ball rolling on AI safety and regulation. As we covered before, regulation can actually serve as an accelerant in healthcare so these moves might help adoption as trust builds and safety/risk concerns start to ease. (link)

Below are some of the official statements:

Google: Advancing bold and responsible AI
Inflection: The precautionary principle
Open AI: Moving AI governance forward
White House: Voluntary commitments from leading AI companies

Miscellaneous 🔍

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

• Microsoft announces AI for enterprises — to be integrated with Microsoft Teams & Office 365 (link)

• Two big chatGPT updates — higher message limit & custom instructions (link)

• Meta just released an open-source ChatGPT rival, Llama 2 (link)

• ChatGPT for Android dropping later this week (link)

• Apple wants to enter AI frenzy with its own ‘Apple GPT’ (link)

• Study highlights bias exhibited by GPT-4 in diagnoses (link)

• Google is testing journalist tool, Genesis, which can write news stories (link)

• FTC thinks AI raises competition concerns (link)

• Research to merge human brain cells with AI secures Australian defense funding (link)

• How a Boston hospital is priming medical residents for an era of AI medicine (link)

• Study shows AI can expand scientific discoveries by predicting and generating hypotheses that humans might not consider (link)

• Opinion: Healthcare needs to be more involved in AI (link)

• Opinion: AI not ready to replace physicians (link)

• ChatGPT out-scores medical students on complex clinical care exam questions (link)

• The next frontier for LLMs is biology (link w/out paywall)

Deal Desk 💸 

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

📈 K Health: a digital primary care app based in Israel, just raised $58M in funding from existing investors, including Cedars-Sinai. K Health will use the funds to build out its suite of AI-powered medical services. (link)

📈 Neura Robotics: an AI commercial robotics startup based in Germany, raised $55M in funding. European investors Lingotto, Vsquared Ventures, Primepulse, and HV Capital led the round. Their upcoming MiPA service robot is aimed at offices, care facilities and homes. (link)

📈 Peterson Health Technology Institute: The Peterson Center on Healthcare, launched a new nonprofit with $50M, the Peterson Health Technology Institute (PHTI). PHTI will provide independent evaluations of innovative healthcare technologies, delivering rigorous, evidence-based assessments that will analyze the clinical benefits and economic impact of digital health solutions, as well as their effects on health equity, privacy, and security. (link)

📈 GenHealth AI: a medical data prediction startup for payors and providers based in Boston, raised $13M in seed funding. Craft Ventures and Obvious Ventures co-led the round. (link)

Tool Box 🧰

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

🔧 Cedars Sinai + K Health: Cedars-Sinai, one of the largest academic medical centers in California, just invested and partnered with K Health, a digital primary care startup, and will adopt their AI chatbot. The chatbot will help steer patients toward the right level of care and connect them with a human provider if needed. (link)

🔧 Northwell + Feinstein: Northwell Health, New York's largest healthcare provider, and its research arm, the Feinstein Institutes for Medical Research, plan to create a new AI-focused research center. The center will support the development of large-scale data models and AI to identify and address healthcare disparities and patient risk. (link)

🔧 DeepScribe: DeepScribe, an AI-powered medical scribe, has hundreds of physicians nationwide using their tool for patient visits. The startup turns these conversations into usable medical records via AI and has a team of ~200 people that manually improve/revise the records. (link w/out paywall)

AI Images of the Week 📸

Funny memes and pics from around the web…