Dave Clifford

This is not an endorsement--just research at this point. However, #testdata for #healthcare is always an issue. It was great timing to meet Brad Genereaux from NVIDIA just a few days before they released an open source #synthetictestgeneration. https://lnkd.in/evkZrUEs There are also companies like Atropos Health that are collecting large amounts of data for #healthcare.

16

6 Comments

I’ve been given advice that the fastest way to monetize HER Heard without charging the end user (woman) is trackers and selling data. Instead, I’d like to build something more trustworthy. I don’t want to be in a headline like this. If you are not paying, then you are the product (you/your data are being sold). There are newer ways to deindentify data or even to aggregate things women search for that can be of interest to various stakeholders. In theory that could be okay to sell, but in a post-Roe world with laws encouraging bounty hunting, I am vetting the choices ahead. I am all for moving fast, without breaking things.

6

The overwhelming nature of healthcare largely can be attributed to the size of the market. It’s unfathomably large–$4 trillion in the US alone–and made up of countless submarkets. The American healthcare market could support dozens of FAANG (Facebook, Amazon, Apple, Netflix, Google)-scale companies, but today only one exists (UnitedHealth Group). Healthcare at its core is (1) a data, operations, and logistics problem, and (2) a consumer experience and engagement problem. Both are areas where the tech world excels. When it comes to data, operations, and logistics, the current state of healthcare leaves endless opportunity for improvement. $765 billion is wasted annually, with the main driver being human administrative overhead. Instead, some of the biggest problems in healthcare are going to be solved by technology. Healthcare provides the best opportunity out there to use the hottest tools in tech–especially AI–to displace huge incumbents. https://lnkd.in/gdUrrNEZ

7

3 Comments

This article is an interesting take on the age old dichotomy of governance versus control, now as seen in the world of AI. What we have historically done in other domains is accept that direct control isn't going to work, and instead shifted to guidance, education, delegation, and some amount of post-fact monitoring. Though I agree that in the AI space deviation from control policies can be even harder to detect, I do believe the root challenge to be one we have seen before and expect that the same compensating mechanisms can be put to good use. https://lnkd.in/g7ferE8g

29

2 Comments

Another week of AI news cycle churn begets another evolving dynamic in Congress…Caroline Reppert and I delve into the potential healthcare implications and how the current landscape sets the scene for the rest of this Congress… From drug discovery and medical imaging to diagnostics and robotic surgery, AI's transformative potential in the biosciences and healthcare domains is undeniable. However, as with any disruptive technology, there are concerns and challenges that must be addressed to ensure responsible and equitable adoption. Recent developments showcase the rapid momentum behind AI in healthcare settings. The University of Wisconsin's collaboration with Microsoft and Epic demonstrates the use of large language models (LLMs) in drafting patient messages and streamlining communication while ensuring data privacy. Similarly, Nvidia's advancements in AI-powered drug development, medical imaging, diagnostics, and robotic surgery underscore the industry's commitment to leveraging AI's capabilities to transform healthcare practices. On the policy front, global leaders recognize the need to strike a delicate balance between fostering innovation and implementing appropriate guardrails. The recent G7 commitments and initiatives reflect this sentiment. Concerns surrounding equitable access, bias mitigation, and the establishment of clear frameworks must be addressed. In the US, both the Senate Health, Education, Labor, and Pensions (HELP) Committee and the House Energy and Commerce (E&C) Committee are planning hearings and roundtables to educate members on the nuances of regulating AI in the right way and at the right time. While substantial progress towards comprehensive AI legislation may not materialize in the current Congress, the ongoing discourse and initiatives serve as critical markers for the future. The most immediate avenue for tangible progress is likely to be through the appropriations process, where funding allocations and directives can shape the trajectory of AI's integration into various sectors, including healthcare. The current Congress may lay the groundwork for future advancements, but the most tangible progress is expected to occur through targeted funding allocations and directives within the appropriations process. As the #AI revolution unfolds in #healthcare, stakeholders must work collaboratively to harness its potential while mitigating potential risks and unintended consequences. Monument Advocacy

16

Start your week with a quick but insightful analysis of the state of investments in #digitalhealth by Michael A. Greeley at Flare Capital Partners. "The opportunities to “transform the business of healthcare” (Flare Capital’s tag line) remain obvious and acute, notwithstanding that the level of investment activity has receded from the $30 billion highwater mark in 2021 to a more manageable $10 – $12 billion annual pace. Healthcare technology now represents approximately 8% of all venture capital investments. Just as important is the number of companies funded which is trending back to 400 – 500 companies per year, implying that round sizes have decreased dramatically to ~$20 million, according to data from Rock Health." Read the full article. Julien de Salaberry John Nosta Gil Bashe John Ruffolo Michael Wasserman Suzanne Grant, CD Donald Jones Rizwan Kheraj Alexei Doudkine Laetitia MacDougall Levi Shapiro

10

Exciting News for State & Local Health Agencies: $255M CDC Data Modernization Program. As a leading mobile technology provider for state government health and human services, we understand the importance of quality data and data interoperability. This enhanced funding supports state, local, and tribal governments focus on real-time, seamless data sharing among agencies. The emphasis on robust and secure data exchanges promotes streamlined service delivery across government programs, improving outcomes for the most vulnerable populations. #hhsit #DataModernization, #GovTech, #SystemsofEngagement https://lnkd.in/g3ibTN38

26

The AI in Health Conference is coming up on September 9-12. Themes for this year include #FoundationModels in health, applications of #LLMs in health, AI in #Neuroscience and #Neurotechnology, AI for the healthy industry, #DigitalTwins in healthcare, ad patient engagement and #Equity in #HealthAI.   Submit your abstract to disseminate the results of research! #AIHealthConference #AI #ArtificialIntelligence #DataScience #Health #KenKennedyInstitute

50

1 Comment

News Report Finds Racial Bias in Radiology Reports A recent Yale study found that when asked to summarize test results to be understood by nonprofessionals, the Language Model (LLM) produced results at lower reading levels if told the patient was black vs white. This news report has implications well beyond health, it impacts education and other areas. The findings suggest that if our data is biased, then AI's will be biased. Check out the full report here: https://lnkd.in/edSN_iTa #health #AI #education #racialbias #diversityandinclusion

6

I submit that Healthcare and Education are two sectors that could profoundly benefit from AI. This is a positive development. "Providence, Microsoft, and the University of Washington have developed an open-weight, AI-powered pathology model called Prov-GigaPath. Prov-GigaPath, according to a May 22 news release from Providence, is designed to improve patient care and transform cancer diagnostics by capturing global patterns across whole pathology slides. The model, trained on 1.3 billion pathology image tiles from 171,189 digital whole slides provided by Renton, Wash.-based Providence, is one of the first foundational models for digital pathology that has been pretrained using real-world data. Additionally, unlike prior models that analyze small portions of slides, Prov-GigaPath captures both local and global patterns across entire slides using the GigaPath vision transformer architecture for pre-training, according to the release. It was also found to outperform existing pathology models." #ai #genai Paul W. Kevin Rank, MBA Molly Zimmer https://lnkd.in/gu2MMkKq

1

If you're looking for an informative session to wrap up your week, consider registering for this Technology Community of Practice session that will talk about driving better outcomes, overcoming data obstacles, and applying trauma-informed principles to your data practices. See you there!

1

Great presentation by Dan Buckland, MD for those looking to design Health AI/ML models. "...if we have been poorly treating a demographic or population, that will show up in the clinical data. We might not have race, gender or zip code as an identifier in the original clinical data." shared Dr Buckland. Good insights to support the need of Asher Informatics PBC AI/ML independent equity performance audits, AI-Assess. Through a funded federal research contract, we've designed a tool in which we can provide independent real-world data subgroup population testing on clinical and signal AI/ML models and products. Right now, the MVP is designed to support #oncologyAI FDA 510K submissions and enable the FDA access to the testing of an algorithms performance and therefore better evaluate an AI/ML submission. We've partnered with Gradient Health to provide data that includes race, gender, age and we're utilizing Rhino Health's federated platform to be able to pull testing data from multiple sources. If you're an oncology AI developer, we're currently conducting feedback interviews and would love to connect.

9

1 Comment

Come join me tonight for an update on the Furture of AI in healthcare. We will be discussing how technology will be implemented in the healthcare field and how emerging tech innovations will advance patient outcomes. Register now. #hoodai #healthinformatics #ai #machinelearning #healthcare https://lnkd.in/d_erz_Mf

19

Hot off the press: a new article from the Lancet: Digital Health published last week explores how a team at Brigham and Women's Hospital tested using Chat-GBT (a large language model, or LLM) to assess how LLM's might be used to respond to patient portal messages, and how this might impact subjective efficiency, clinical recommendations, and potential harms. TL;DR: using ChatGBT reduced physician workload, improved consistency across physician responses, and enhanced educational value of responses. However, there are ethical concerns with LLM's biasing physician clinical decision making, and that "physicians might use LLMs' assessments instead of LLM responses to facilitate the communication of their own assessments" So your MyChart message might be answered by a robot from now on, and that robot might be the one making clinical recommendations for you... Thoughts? 🤖 💬 (as an aside, I was invited by LinkedIn to "rewrite this post with AI". I did not. 😉 ) https://lnkd.in/dVN9gwnU  

15

2 Comments

Abt Global evaluated The Centers for Medicare & Medicaid Services Oncology Care Model's (OCM) effectiveness in reducing care costs while maintaining quality. To examine OCM's impact on #HealthEquity, we assessed four historically underserved populations: Black and Hispanic patients, those with dual eligibility for Medicare and Medicaid, and residents of resource-poor neighborhoods. Our findings showed that, before OCM, these groups had higher acute-care use and episode payments, but were less likely to receive timely chemotherapy after surgery, adhere to oral treatments, or receive hospice care. While OCM improved oral treatment adherence for all groups, care quality was not consistently improved and only the treatment of Hispanic patients showed reduced payment costs. Learn more: https://lnkd.in/ebAxbyMx #HealthDisparities #HealthResearch

17

I have an Idea to develop relevant GenAI application this week 😁 🏆 "Political Pulse App" 🏆 - Get details of what political leaders promised during election campaigns 🌀 - Get candid insights from over-the-top statements that leaders said in National media⚡ - Hear leaders feedback on voting & exit polls 🪐 - Compare their final thoughts tomorrow 🌠 - Analyze the emotions of hope, attitude, change and intelligence in the last 3 months 🌎 - Don't miss the acceptance or blame games as per final numbers 🫣 Publish results for all GenAI enthusiasts 😀 Give prizes as below ✔️ From 0 --> 1 🏆 ✔️ From 1 --> 0 🏆 ✔️ From 0 --> 0 🏆 ✔️ From 1 --> 1 🏆 #politics #india #elections #genai #justforfun #media

20

I just came across a fascinating article by Casey Sackett, Devin Harper, and Aaron Pavez from Webot in IEEE Spectrum Magazine about the exciting (and tricky!) world of large language models (LLMs) in mental health. In essence, they describe that current LLMs are challenging to control for applications for which these models are not trained and might not behave as wished. One example is that the data and objective the LLM are trained on is to provide answers and not help develop insights which could lead to a change in perception and behavior. An even more significant challenge is that uncontrolled use of LLMs could create serious risks for people seeking help in the healthcare domain. This has the effect that these models shy away from such situations and bail on responses. Nevertheless, the article is optimistic (as I am too) and outlines that the technology can be used and integrated into more traditional frameworks. Although (pre-trained) LLMs have many issues, there is still much potential left to improve them fundamentally. If you're curious about the future of AI in healthcare, this 11-minute read is worth your time! ⏱️ https://lnkd.in/exVqep2z

19

We agree! We are focusing on this use case at One Fact Foundation - if anyone wants to help (or needs help!) please reach out: hello@onefact.org or DM me. We have helped several health systems integrate my work on AI from grad school (https://lnkd.in/e-DACRwr) which is at 800+ citations now, and used this system in everything from: - improving recapture rates for the Centers for Medicare & Medicaid Services' hierarchical condition category codes. This can lead to increased reimbursement from government Medicare Advantage programs prevalent in Accountable Care Organizations or Ambulatory Surgical Centers - especially for private equity-backed care orgs :) - improving the prediction of diagnosis codes (usually the World Health Organization's international classification codes that rockstars like Tahilia J. Rebello, PhD help develop by coordinating thousands of clinicians worldwide) - reducing the burden on clinicians to take notes rather than spend time with patients. Often note-taking happens on paper, and in the United States, is bound by legal liability and cash incentives, or cash-related incentives like relative value units used to improve coding integrity in revenue cycle management programmes and increase case-mix index and reimbursement. Happy to chat about any and all of the above, especially for under-served (rural, remote, poor, low- and middle-income country tertiary care) systems. A dream would be to support the ASHA Aide programme in Ministry of Health and Family Welfare, Government of India or the Hukou system run by the Ministry of Health of the People's Republic of China. If you know anyone, feel free to cc me (jaan@onefact.org). Thanks to Theresa Osborne, Yubin Park, PhD, Christopher Whaley, @cynthia fisher (Power To The Patients and Patient Rights Advocate), Brown Institute for Media Innovation, the AIM-AHEAD Consortium, Michela Magas, Steven C. Hayes, Philip Resnik and all of our other funders and supporters. This would not be possible without your support! Excited to drive the cost of health care delivery as close to zero as possible.

9