Healthcare AI Guy Weekly Newsletter | 8/8

Our Picks ✨

Highlights if you’ve only got 2 minutes…

Where will AI have the biggest impact? Healthcare

a16z just released a new thinkpiece arguing that just as emerging markets went straight from using cash to mobile payments (“leapfrogging” credit cards altogether), healthcare is going to move straight from fax machines to AI, “leapfrogging” traditional vertical software.

With AI, healthtech companies no longer need to fight the uphill battle of training people on software. Instead, they can sell AI that acts like a person and takes more and more work off healthcare professionals’ plates. The pitch is simple: “Don’t want more software? Cool, we can give you AI ‘people’ who are cheap, fast, cheerful and empathic.”

This revolution will start with non-clinical use cases, followed by clinical ones. AI will take on the work of call centers, scheduling, prior authorization, medical coding, revenue cycle management, and fighting medical bills.

a16z believes the two biggest problems in healthcare are (1) access and (2) cost, and both can be solved with AI.

When it comes to access, people won’t have to wait months to get quality care. As AI improves, every single person will have a world-class AI doctor in their pocket, spanning all medical specialties. Diagnoses will come months earlier, allowing expedited, AI-guided interventions by doctors who focus their time in more impactful ways.

When it comes to cost, AI will radically decrease spend by shifting care from all-human services to AI-augmented services. Note this is not true in industries that are already software driven—AI is more expensive than enterprise SaaS. But it’s way less expensive than human services, and healthcare is a $4 trillion industry that is mainly human services. AI will help us achieve a future where every person can afford world-class medical care, and medical debt is no longer the number one cause of bankruptcy.

Overall, it’s always worthwhile reading takes from industry thought leaders on where they think the puck is going. One thing is for certain—healthcare is ripe for disruption. (link)

AI tool detects breast cancer in mammograms as effectively as radiologists 

An AI platform can detect breast cancer in mammograms as effectively as experienced radiologists, according to a new study.

Published in the Lancet Oncology journal, the study was performed by Lund University in Sweden on ~80,000 women and “showed that AI readings of mammograms actually detected 20% more cases of breast cancer than the ‘standard’ reading by two radiologists.” The AI assessments were verified by one or two radiologists, depending on the patient’s risk profile.

According to this study, AI could reduce radiologists’ workload by about half, freeing them up for more advanced diagnostic work. Although providers are increasingly deploying AI in medical settings, there should still be caution over how algorithms are trained and validated, as well as the potential for bias and overdiagnosis. BUT the signs are hopeful. (link)

McKinsey’s ‘The State of AI in 2023’

The latest annual McKinsey Global Survey detailed a lot of interesting findings about the impact of generative AI and its promising impact on business transformation. Some quick hitters:

• 1/3 of survey respondents use AI tools in one or more business functions

• 25% of C-suite executives admit they use gen AI products for work

• 40% of respondents said their companies will invest more in AI

• 3/4 expect AI to disrupt their industry's competition within 3 years

• AI tools are mostly used in marketing, sales, product and service development, and service operations

• Main risks include accuracy, cybersecurity, and IP infringement

• ~4/5 AI adopters expect more than 20% of their workforces to be reskilled, while 8% foresee a 20% workforce reduction

• Companies use generative AI to enhance human tasks, not replace them.
  Companies are mostly employing generative AI to boost top-line growth and productivity

• Over 2/3 expect their companies to increase AI investment in the next three years

  (link)

Microsoft, Duke Health announce five-year partnership focused on AI

The barrage of health systems partnering with big tech to roll out AI continues. Microsoft and Duke Health launched a five-year collaboration to support AI applications in medicine, with a focus on building out infrastructure such as cloud-based data platforms.

Both are members of the “Coalition for Health AI” which was formed late last year to develop guidelines for the responsible use of AI in medicine. The partnership allows Duke Health to responsibly expand its use of AI while Microsoft gets to learn how academic medical centers evaluate algorithms.

Microsoft’s Azure OpenAI Service will also support Duke’s experimentation with LLMs. Duke is first interested in the ‘low-hanging fruit’: administrative applications of LLMs like optimizing operating room schedules and lining up documents for providers to sign off on. Duke will later “shift to more complex and delicate tasks like summarization, distilling research or medical cases while ensuring that the AI isn’t spitting out inaccurate results. (link)

Miscellaneous 🔍

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

• AI helps quadriplegic move his arms again (link)

• Mayo Clinic, Kaiser to team up with big tech on healthcare AI (link)

• Stanford student makes a pair of glasses that can transcribe speech in real time for deaf people (link)

• AI may take on doctors’ roles sooner rather than later (link)

• How Kaiser is using AI to meet patients' needs faster (link)

• Sutter Health piloting Epic & Microsoft’s AI tools (link)

• AI is improving in providing empathetic care (link)

• ChatGPT gets some new features, including multi-document chat (link)

• Goldman Sachs: AI investments could surge to $200 billion globally by 2025 (link)

• OpenAI filed a trademark for GTP-5 (link)

• Meta releases open-source model for sound generation: Audiocraft (link)

• Close to 2.5% of the US had their health data accessed by MOVEit hackers (link)

• Anthropic AI had their hackathon last weekend — Dr. Claude was the 3rd place winner (tweet)

Deal Desk 💸 

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

📈 Kyoto University: Kyoto University’s VC arm, Miyako Capital, has established a fund expected to raise $141M to back startups involved in deep-tech biology and artificial intelligence. (link w/out paywall)

📈 TytoCare: a virtual home care company based in New York raised $49M in funding. The round was led by Insight Partners with MemorialCare, Healthcare of Ontario Pension Plan and Clal participating. The funds will be used to expand AI offerings for chronic care and diagnostics. TytoCare’s customers comprise 220 health systems/plans across the U.S., Europe, Asia, Latin America and the Middle East. (link)

📈 Likeminded: an AI-supported digital mental health startup based in Berlin raised $6.4M in seed funding. Holtzbrinck Digital and existing investor Heartcore Capital invested. (link)

📈 FeelBetter: an AI-based medication management startup raised $5.9M in series A funding. Firstime Ventures and Shoni Health Ventures led, with participation from Random Forest VC, The Group Ventures, and existing investor Triventures. (link)

📈 Basys.ai: a VBC enablement platform for health systems and payers based in Cambridge, MA raised $2.2M in funding. Nina Capital led the round. (link)

Tool Box 🧰

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

🔧 Northwell Health Feinstein Institute: Scientists from Northwell Health’s Feinstein Institute successfully used AI-driven brain implants to enable a quadriplegic man, Keith Thomas, to regain movement and sensation. Surgeons implanted microchips in his brain that interface with AI algorithms to re-establish the connection between his brain, body, and spinal cord. When Thomas thinks about moving, the chip signals electrodes on his spine and arm muscles, facilitating movement. Additionally, sensors on his fingers send touch data back to his brain. Thomas' arm strength has more than doubled in just four months post-implantation — Go Keith! (link)

🔧 Google Health: Google Health launched its advanced, lightweight, multimodal generative AI models for medical imaging. The unique technique, termed "ELIXR", involves "grafting" language-aligned vision encoders onto a set LLM. (link)

🔧 Microsoft + Duke Health: Microsoft and Duke Health launched a five-year collaboration to support AI applications in medicine, focusing on building out infrastructure such as cloud-based data platforms. (link)

AI Images of the Week 📸

Funny memes and pics from around the web…