CERIS’ Post

Transformative impact of Generative AI in healthcare, emphasizing its potential in diagnostics and patient care, while acknowledging challenges like data privacy and bias. Insights from Peter Lee on how Generative AI is shaping the future of Healthcare from after-visits patient summary to diagnostic procedures to the personalization of treatment plans.

Magnetic Ventures portfolio company Paradigm recently worked with OpenAI to evaluate #GPT4 to reduce the administrative burden of screening patients for clinical trials. The results exceeded accuracy of industry best practices while dramatically decreasing the time to read, extract, and interpret medical record data. By incorporating #LLMs into Paradigm’s platform, the team can screen patients faster, more accurately, and at a dramatically lower cost. I'm excited to see what is next for this team. Paradigm is changing the status quo for how we conduct clinical trials. Their mission is to create a more equitable model where any patient in any geographic location can have access to clinical trials as a routine care option. This is just another example of the impact technology is having across the healthcare ecosystem. Congratulations to Paradigm and OpenAI team. #healthequity #clinicaltrials #deeptech #convergence #AI #healthcareinnovation https://lnkd.in/gxAkmmGP

About 10% of #healthcare expenditure in the U.S. will be going to medical #AI. It will be a huge waste if we do not know how to use it correctly. In this interview with Fast Company, I discussed some of our findings and work on increasing the public impact of #AI. 👇👇👇

Google just launched Med-Gemini, an advanced AI fine-tuned for medical tasks. Its performance surpasses previous models, including GPT-4, across various medical benchmarks, exhibiting an average superiority of 44.5% over GPT-4 on seven multimodal benchmarks. Med-Gemini excels in tasks such as medical summarization, generating doctor referrals, and simplifying medical documents. Notably, it outperforms human experts in handling complex text-based medical tasks. This development represents a significant stride in AI for healthcare, hinting at potential enhancements in medical diagnostics and patient care. Read more about the work here>Paper: https://lnkd.in/e5eW6Sis 

This is a concise 10-minute "30,000 foot" discussion of AI in Healthcare. However, I think the concepts are applicable to many industries. Bill Siwicki (Healthcare IT News) interviews John Halamka, M.D., M.S.. John has had a storied career in healthcare technology leadership and successfully brought IT visions into reality at the many organizations fortunate to have had his leadership. Key points:   ➣ Highlights of "clinical decision support systems" and how far we've come from the archaic (by today's standards) static rules engines to the potentially game changing capabilities of probabilistic generative AI. ➣ The inherent issues with a system that provides different answers/solutions the more you question it; and what needs to be done in today's version of generative AI to protect ourselves. ➣ Outline of what the Mayo Clinic is doing to create a "private" platform that queries structured and unstructured Mayo data (deidentified of course); and how it's sped up academically oriented technology projects from 18 months to a few weeks. ➣ Summary of anticipated regulatory steps to be taken in the coming months and the importance of a public/private oversight partnership.   The quote of the interview, "MDs and nurses who use AI, will replace MDs and nurses that don't." https://lnkd.in/eCG7S63h

Using AI for public impact: Insights from Dr. Soroush Saghafian Read more about the five key takeaways from an AI pioneer about its potential impact in healthcare! 1. UNDERSTAND AND MITIGATE ALGORITHM AVERSION IN HEALTHCARE 2. BRIDGE AI SOLUTIONS AND PUBLIC POLICY 3. PROMOTE EQUITY THROUGH ETHICAL AI PRACTICES 4. OVERCOME REGULATORY AND IMPLEMENTATION CHALLENGES 5. THE FUTURE OF AI IN HEALTHCARE: PREDICTIVE AND PROACTIVE MODELS https://lnkd.in/ehn8xeWb

As models like this begin to roll out in a big way, I always have in the back of my mind both the ONC final rule (45 CFR Parts 170/171) and the White House AI EO, as it relates to foundation models. It will also be interesting to see how much generative AI models (LLM/MLM/SLM), with these and similar use cases, begin to flirt with the edges of medical device classification. Nothing in this article indicates a medical device use case for this early testing of MedLM ... yet.

Exciting AI breakthrough in healthcare from Google & DeepMind! 🔥 Their new Med-Gemini model achieved an impressive 91.1% accuracy on the MedQA benchmark for medical question-answering, surpassing previous best results. It also outperformed GPT-4's average by 44.5% on multimodal tasks and even exceeded human experts in some cases! Med-Gemini leverages innovative techniques like self-training, uncertainty-guided search, custom encoders, and multimodal fine-tuning to handle diverse medical data. This could revolutionize healthcare with accurate diagnoses and informed clinical decisions. An exciting step for AI in medicine! What are your thoughts on the potential impact of Med-Gemini? Let's discuss!

Artificial intelligence (AI) has exploded in popularity, making the predictions of its impact on healthcare much more real. We've all seen glimpses with user-friendly tools, but AI's true potential in diagnostics, drug development, and patient care is vast and can be overwhelming. This new book cuts through the noise with 100 focused questions and clear answers, guiding you through the essentials of AI in healthcare. https://lnkd.in/e7ZbmHWk

When one interviewer asked Red House CEO Jeff Hillam "What’s one trend that excites you?" he immediately discussed the role of AI in healthcare and RCM today. "The use of artificial intelligence and machine learning in the medical billing industry is becoming more common, providing operational benefits for those in the revenue cycle management field," Jeff explains. He also notes that in the healthcare industry as a whole, companies like Red House are working to create data sets that can be used by artificial intelligence to assist clinicians in diagnosing, treating, and curing illnesses. It is fascinating to witness the integration of AI in the healthcare industry, especially on the data side! Read Jeff's full interview here: https://lnkd.in/eHGQ7Hnd.