Gorkey Vemulapalli

👉🏼 The Digital Technology in Clinical Medicine: From Calculators to ChatGPT 🤓 Somarajan Anandan 👇🏻 https://lnkd.in/gtRmmq8z 🔍 Focus on data insights: - Utilization of digital technology in clinical medicine - Impact of digital information technology on healthcare delivery 💡 Main outcomes and implications: - Advancements in clinical medicine due to digital technology - Need for balance between technology, clinical expertise, and humanistic values 📚 Field significance: - Regulatory push for improvements through clinical trials - Projects like "World Health Organizations Guidelines for Digital Interventions" and "Food and Drug Administration's Digital Health Center of Excellence" highlighting efforts in digital health services 🗄️: [#digitaltechnology #clinicalmedicine #healthcaredelivery]

1

I caught up with Nikhil Krishnan of Out-Of-Pocket fame to get his take on the current state of the healthcare technology landscape. Up first: his current thoughts on what's exciting and where there are opportunities for AI-driven solutions. First half of our Q&A below 👇 — 🟩 Q: When it comes to AI, what area are you most excited about? Patient-facing solutions. The AI admin and back-office space has quickly become saturated, and incumbents will always have a distribution advantage here. Which is why I’m actually surprised there aren’t more players entering the patient-facing side. I think the traditional healthcare companies are too nervous to actually go into that patient-facing area. And it's also a regulatory gray area, so if you actually get traction, you can probably help shape how this stuff ultimately gets regulated. So right now, I think the greatest opportunities are for patient-facing AI tools. 🟩 Q: Where specifically do you see whitespace to develop patient-facing solutions? I think there’s still a lot of opportunity on the patient-facing automation side. A great example actually is automated lab results. There has been an issue for the last 3 years that lab results automatically get released to patients, and they see something's abnormal. The patient is trying to figure out what’s going on and what they've read. Part of this is explaining the result, which is something Elaborate is working on, and in some cases, actually, it's really useful for the patient to get that notice immediately with follow up steps, like scheduling a diagnostic. Another area is AI that can autonomously make workflow changes within a local system. Viz.ai is working on this automated workflow piece in the context of situations like stroke care coordination and hypertrophic cardiomyopathy. This requires not only automation in terms of being able to understand the context of multiple datasets, but also make the requisite workflow changes (e.g. escalating to the right specialist). These are particularly useful in time-sensitive areas. 🟩 Q: Any other AI whitespaces? Monitoring of AI models. It's one of the areas that's still unknown in terms of how do you give peace of mind to your buyer? It becomes really hard to actually evaluate and audit a lot of these AI algorithms without some form of data exhaust. I feel like one of the things that will happen is you'll see some form of certifying entity emerge that can give a gold standard to hospitals. 🟩 Q: What else are you thinking about as it pertains to AI? I do think that pricing is one of the more interesting components here. For example, in AI scribing, how you think about pricing changes a lot depending on if you’ve got a human in the loop or not. Because adding a human layer goes from near zero marginal cost to cost scaling linearly. One of the ways that companies will be able to differentiate themselves — especially from incumbents — is the pricing model.

67

4 Comments

AI in Healthcare is a hot topic. But what is the reality of it today vs a few years from now? This is what I wanted to dive into with our good friend Ankit Jain. This was a great conversation on AI and the Trough of Disillusionment. Trough of Disillusionment is a phase in the Gartner Hype Cycle that occurs when interest in a new technology wanes after initial hype and expectations are not met. Internet, cell phones, etc all went through this phase of failed ventures until the real use cases started to be come common place. It's hard to remember that the internet went through this. We had a wave in early 2000 and then it all broke when it wasn't being adopted as fast or just wasnt ready for real prime time. I think we are in this phase right now with AI being a hot topic and people can see the potential. Yet, full scale practical changing of my entire world or job isn't there yet (for most things). check it out... https://lnkd.in/gjg6w_m2

23

10 Comments

👉🏼 ChatGPT's Inconsistency in the Diagnosis of Alzheimer's Disease 🤓 ArunSundar MohanaSundaram 👇🏻 https://lnkd.in/gyhpr7mn 🔍 Focus on data insights: - ChatGPT's responses in diagnosing AD and MCI stages varied. - Reassessment of ChatGPT's accuracy and reproducibility in diagnosis. - Discussion on reasons for differing results and suggestions for future studies. 💡 Main outcomes and implications: - Potential limitations in using ChatGPT for AD and MCI diagnosis. - Need for further research to enhance ChatGPT's diagnostic reliability. - Importance of understanding AI tools' nuances in clinical applications. 📚 Field significance: - Enhancing AI tools for medical diagnosis. - Impact of AI inconsistencies on clinical decision-making. - Future directions for improving AI-assisted healthcare. 🗄️: [#AlzheimersDisease #AI #MedicalDiagnosis #ClinicalApplications]

1

...an interesting interview with Micky Tripathi - National Coordinator for Health Information Technology ... "The agency head discussed ONC’s accomplishments over the past two decades, improving documentation burden among clinicians and artificial intelligence opportunities." Micky has been a stellar ONC Head, and the healthcare interoperability progress being made is really fantastic. My favorite quote from the article is, "The basic mantra is that we want to make sure that these technologies are used *for* patients, and not *on* patients (emphasis added)." #onc #interoperability #tefca #healthcareai #tripathi #digitalhealth

1

The Medical Revolution continues. And kudos to Google DeepMind for allowing the free sharing and use of AlphaFold 3 Server. Allowing global researchers to utilized the inmense AI powered biotech builder. AlphaFold 3, with its advanced capabilities to predict protein structures, can provide significant benefits to humanity in various ways: 1. Drug Discovery and Development: By accurately predicting the structure of proteins and their complexes, AlphaFold 3 can help scientists understand how drugs interact with proteins at a molecular level. This understanding can lead to more efficient drug design and development, potentially reducing the time and cost associated with bringing new medications to market. 2. Understanding Diseases: Many diseases, including cancers and genetic disorders, are linked to proteins functioning improperly. AlphaFold 3 can help researchers understand the structural basis of these malfunctions, which is crucial for developing targeted therapies and interventions. 3. Biotechnology Advancements: In biotechnology, enzymes and other proteins are often used to catalyze reactions, create biofuels, and treat wastewater. AlphaFold 3 can assist in designing and optimizing these proteins for specific industrial applications, enhancing efficiency and sustainability. 4. Agricultural Improvements: Protein engineering can be used to develop crops that are more resistant to pests, diseases, and environmental stresses. AlphaFold 3's insights into protein structures can facilitate the creation of such crops, potentially increasing food security globally. 5. Basic Scientific Research: Understanding protein structures is fundamental to molecular biology. AlphaFold 3 can accelerate basic research by providing insights into the life at the molecular level, helping scientists uncover new biological pathways and mechanisms. By unlocking detailed knowledge of protein structures and their interactions, AlphaFold 3 can drive innovation across many fields, ultimately contributing to health, environmental sustainability, and various industrial processes. #ai Dominic Ligot https://lnkd.in/g_TFN6D6

8

On Tuesday 2 July Ankur Kalra and I are hosting a webinar on 3 key questions to evaluate AI investment opportunities in healthcare: 1. What part of your investment's AI lifecycle is defensible?  2. How does generative AI influence your investment strategy?  3. How should you assess Leadership Capability in the AI team? We'll also be talking about what breakthroughs in AI seem imminent in 2024 - what should you look out for. Ankur who is based in the US is the founder and CEO of Hop Labs, a machine learning consultancy focusing on reproducible ML research and scalable ML systems in healthcare, life sciences, and other industries. Prior to Hop Labs, Ankur was a senior technologist at Amazon, and was involved in the launch of Amazon Prime and Amazon Instant Video. Our virtual event is for: ➡️ Investors with an interest in AI in healthcare ➡️ Innovative companies deploying AI and seeking investment in 2024 ➡️ Advisers working with investors and companies seeking funding If you'd like to join us pls register using the link in the comments.

9

4 Comments

Big Tech is transforming healthcare, and the latest report from CB Insights breaks it down. Amazon, Google, Microsoft, and NVIDIA are not just dabbling in healthcare; they're revolutionizing it. Here is my take on some forensic profiling 😉 (based on the full report). 💡 Amazon is focusing on primary and specialized care, leveraging virtual platforms and partnerships to build a comprehensive healthcare ecosystem. Their goal? Making healthcare more accessible and efficient. 💡 Google is the data powerhouse, turning health data into gold for biotech innovations. With 90% of big tech activity in genetic therapeutics, Google is paving the way for groundbreaking advancements. 💡 Microsoft is the AI enabler, equipping healthcare organizations with cutting-edge tools to enhance clinical research, drug R&D, and care delivery. Their acquisition of Nuance Communications is a game-changer in integrating AI into healthcare. 💡 Nvidia stands at the forefront of smart hospitals and AI-driven drug discovery. Their powerful hardware and innovative AI solutions are set to redefine medical imaging and surgical applications. These tech giants are not just making waves—they're creating tsunamis in the healthcare industry. The future of healthcare is here, driven by relentless innovation and strategic investments. 👉 What are your thoughts on tech’s role in healthcare? 👉 How do you see these advancements impacting patient care and medical research? Let’s dive into this conversation! Share your insights and let’s explore the future of healthcare together. #HealthcareInnovation #TechInHealthcare #FutureOfMedicine

40

10 Comments

To follow up on Dede Ainbinder recent post: If we use the current EHR data to build a AI clinical decision support tool it is unlikely to capture our clinical decision making process. EHR's are inherently designed to maximize billing and minimize risk of litigation. A clinical decision support tool based on this data will likely do the same. There is a large sunken cost fallacy here where we continue to hang on to garbage data. Just because the data gathering and storage process was capital and time intensive doesnt mean the data is valuable. To truly capture the clinical decision process we need to capture data which sits outside billing and litigation. Its hard to imagine a future where this data isnt used for billing and cannot be called upon during litigation. #EHR #healthdata #clinicaldecision

17

8 Comments

While research in AI has resulted in exciting new algorithms, there’s a lot that needs to be done to harness it to its fullest potential in drug discovery. Here are four considerations that should be taken into account to successfully implement AI in an organization. https://lnkd.in/gycRG6P9

7

Big healthcare AI news yesterday as Tempus AI received the first-ever CPT code for an algorithm-only analysis of previously-sequenced, laboratory developed test transcriptomic data. TLDR: AI molecular phenotyping has a new business model. The PurISTSM test uses RNA profiling to identify molecular subtypes in patients with advanced pancreatic ductal adenocarcinoma, which may inform first-line therapy selection and management. While I'm skeptical that payer balance sheets are going to solve the "who pays for healthcare AI" problem, the reimbursement pipes for clinical AI are maturing. I am surprised to see more progress for usage-based vs. performance-based models given the many challenges with FFS economics. It will be interesting to see how payers assess the "medical necessity" and therefore reimbursability of clinical AI as these models come to market. https://t.co/eLe18SmihV

421

9 Comments

What are important benchmarks you use with LLMs with #AIinhealthcare? Can #LLM excel in medical encoding? Yes! But what is your measurement of success? That you found a billable code or is it the RIGHT billable code? Does it help you ALSO find the RIGHT non-clinical code because of Regulatory pressures like #USCORE and #USCDI? Or for Research, Adverse Events, Observational outcomes, etc...? Questions you should be asking yourself in measurements of success: - Did you benchmark manual review vs LLM review? Was the encoded encounter the same code that left the organization on the bill? Was there a human in the loop to confirm its accuracy? - Was the encoded encounter changed from point of care to #CDI? Did CDI recommend a change after LLM? - Is the LLM recommending other code sets for the encounter because encoding is only a portion of the clinician's need? Are they the right codes or purely a crosswalk? - Did the LLM provide you an international version of your encoding requirement or did they give you a US version (AKA training Hallucination)? Here at IMO Health, we leverage terminology services as an expert system to couple IMO Clinical AI to train LLMs. This greatly improves the accuracy and support of medical encoding and supporting the overall F-Score of what truly occurred during the encounter. If you’re curious about how technology can positively impact medical coding accuracy, this read is for you. https://bit.ly/3VBZFqn  #DataQuality #NLP #ClinicalAI #ClinicalTerminology

10

1 Comment

Is this NEJM AI article a crushing indictment of LLMs' capabilities with regard to clinical documentation and medical coding? Are we still years away from this technology being practically deployable in patient care and hospital operations? At first, the headline and subsequent reporting about this paper would suggest as much, but looking into the methods of this paper demonstrates a subtler message. To start, I applaud the authors' efforts to cut through some of the hype around LLMs and show that they can't perform complex clinical reasoning right "out of the box" (at least to a degree that's ready for production). Taking a deeper dive, however, it's very important to note that their approach does not include any elements of fine-tuning or the use of RAG (Retrieval-Augmented Generation), which would be essential to getting reasonable results for a task as rule-based and outlier-heavy as medical coding. If anything, the study demonstrates to me that LLMs match concepts quite accurately in many cases, and refinement with already-existing tools is likely to improve performance dramatically. Capabilities to add content for RAG weren't present in OpenAI's ChatGPT at the time of this paper, but were added shortly thereafter, as I recall. Would love to see this study repeated to include the corpus of coding rules as a vector database, as I suspect even in a few short months, the results of this study would be dramatically different. One takeaway is that AI is moving faster than the traditional research enterprise can easily adjust to, and we'll need new paradigms to examine these models and ensure that patients benefit from these technologies as soon as they are safe for deployment. However, it's easy to criticize and speculate, so perhaps the most important takeaway is that I'm delighted we have great researchers like the team in this article starting to tackle these important practical issues. Would love thoughts from others in my network who are far more expert than me! #LLM #AI #CDI #HIM #MedicalCoding

82

18 Comments

One key reason why healthcare technologies like AI, telemedicine, and software haven't structurally lowered costs or broadened access for the entire system after decades of experimentation (excluding pure financial engineering of payer contracts) is the limited reimagining of care delivery itself, with asynchronous telemedicine as a notable exception. In the same way that digitizing physical paper failed to exploit the capabilities of a new computing medium for word processing, hosting physician-patient conversations over video does not transform the actual delivery of care. In telemedicine, many of the inputs (e.g., clinician time) and therefore cost structures of in-person care persist, which is why we often see fee-for-service telemedicine companies struggle to achieve economies of scale (i.e., growing revenues non-linearly while growing costs linearly). Until we deconstruct the components of a medical appointment – from intake and triage, to chart review, history gathering, physical examination, diagnosis, treatment, and follow-up – we will continue to be underwhelmed by the disconnect between the transformative potential of these technologies and real-world patient/system impact. I’m optimistic that the emergence of reliable and performant medical AI agents will urge us to perform this deconstruction and redesign the fundamental building blocks of a “medical appointment.”

215

39 Comments

Fascinating insights shared by NEJM AI on the limitations of #LLMs in medical coding! The study by Ali Soroush, MD, MS, and colleagues reveals critical benchmarks in LLMs' ability to map medical codes accurately. Key findings highlight the error-proneness of these models, with #GPT-4 showing the best but still limited performance. The in-depth analysis provided in the study underscores the complexities of medical code querying and the need for extensive training and perhaps model adjustments to ensure precision and reliability. The real takeaway? There's a significant journey ahead in harnessing AI effectively in healthcare administrative tasks. That is why there is a need for medical-specific models trained on medical knowledge and not just random stuff on the internet. 🔗 https://lnkd.in/g2XtCYPC #AIinMedicine #HealthTech #ArtificialIntelligence #MedicalCoding HOPPR

16

2 Comments

Today the most important article in chatgpt and "A text-based conversational agent for asthma support: Mixed-methods feasibility study" - Demonstrates the effectiveness of chatbots in providing asthma support through features like risk assessment and control tracking. - Reports a statistically significant 8% improvement in asthma control post-use of the chatbot. - Highlights the need for improved conversational depth and personalization to meet rising expectations for chatbot fluency. - Emphasizes the importance of balancing conversational capability with accuracy and safety in future health-focused chatbots. #chatgpt #asthma #chatbot #healthcare

Faith in AI Solutionism:qure.ai Fabulous discussions with Somesh Pathak about technology portfolios of qure.ai for TB in Sriperumbudur. #worldtbday2024 #remotedeploymentai #aiforremotedeployment #sdoh #cdoh

8

1 Comment

MCG Health #interoperability experts, Rajesh Godavarthi and Daniel Cawood, have written a new article on the scope and key elements of the Centers for Medicare & Medicaid Services' Interoperability and Prior Authorization Final Rule (CMS-0057-F). This includes (1) an analysis of the implementation/impact of prior auth APIs, (2) a practical guide for #automation processes, and (3) recommended timelines for payers to prepare for the adoption of the new technologies. Read more: https://bit.ly/4e48f8n

32

1 Comment

👉🏼 Generative artificial intelligence in healthcare: A scoping review on benefits, challenges and applications 🤓 Khadijeh Moulaei 👇🏻 https://lnkd.in/e9N9Y2CK 🔍 Focus on data insights: - GAI models like ChatGPT, Google Bard, and Microsoft Bing AI were frequently employed in healthcare. - Various applications of GAI in healthcare were identified, including offering insights on health conditions and disease diagnosis. - Benefits and challenges of GAI in healthcare were explored, highlighting the importance of data accuracy and information reliability. 💡 Main outcomes and implications: - Providing rapid access to information and valuable insights. - Improving prediction and diagnosis accuracy. - Addressing challenges related to generating inaccurate content and ensuring data reliability. 📚 Field significance: - GAI has the potential to revolutionize healthcare by enhancing diagnosis, treatment, and care through advanced data insights. - Understanding the benefits and challenges of GAI is crucial for maximizing its potential in healthcare settings. 🗄️: [#artificialintelligence #healthcare #datainsights #applications #challenges]

2