Rachael Cuevas-Ortiz, MBA

The future of medicine isn’t just coming. It’s already here. Recently at Google Zeitgeist, Danaher collaborators Jennifer Doudna and Eric Topol shared the stage and a powerful vision for what’s next in healthcare. From personalized gene editing to AI-powered empathy in clinical care, their insights are shaping the next wave of biomedical breakthroughs. These aren’t distant possibilities—they’re active areas of research and development across Danaher and our collaborators. Swipe through the carousel for key takeaways, and watch the conversation: https://bit.ly/4nXYLjE

178

Multimodal AI in healthcare is here. There is a great need for infrastructure build, integration and governance. We can help. As imaging, clinical notes, genomics, lab data, and wearable signals are brought together into unified AI systems, healthcare organizations are moving toward a new model of data-driven clinical decision-making. The opportunity is significant: from earlier detection and stronger risk prediction to more personalized care. But integrating and operationalizing multimodal AI requires more than new technology. It requires strong AI governance, data strategy, and regulatory alignment to ensure these systems are trustworthy, compliant, and scalable. At Kosten Digital, we help healthcare and life sciences organizations navigate the growing role of multimodal AI in healthcare through AI governance frameworks, data and AI strategy, and compliance support—helping teams move from experimentation to real-world implementation. 🔗 Read or listen to the blog: https://lnkd.in/dGAJSuNK #KostenDigital #MultimodalAI #HealthcareAI #AIGovernance #AICompliance #DigitalHealth #AI #healthcare Claudia Zylberberg Daniel Merkier Edgardo Resnik Mariano Tribuj

10

2 Comments

"A concern is that AI models trained on one demographic (urban patients) may lead to underperformance on a different population (rural patients). Efforts to produce equitable model performance are essential." – Greg Slabaugh, Professor of Computer Vision & AI, and Director of the Digital Environment Research Institute (DERI) at Queen Mary University of London. The movement to transform and democratise healthcare with AI holds incredible promise, especially for underserved communities. However, as new technologies emerge, it's crucial to ensure equitable access and performance. A recent SUCCESS Magazine article delves into how AI is shaping the future of healthcare, from addressing rural health declines to empowering frontline staff and aiding in rare disease diagnosis. It also highlights critical challenges, including data disparities and the risk of AI models underperforming on diverse populations. Learn more about this vital conversation and the movement to transform healthcare with AI: https://lnkd.in/eJ7T3EPZ #AI #HealthcareInnovation #HealthEquity #DigitalHealth #ArtificialIntelligence

18

A new NEJM AI paper from Liam McCoy and colleagues takes a new approach to a question we've seen asked a lot lately: can LLMs actually reason like clinicians? The team used script concordance testing to evaluate a wide swath of leading models, including GPT-4o and o3. Performance was impressive in places but still well below that of senior residents and attending physicians. In particular, the models showed systematic overconfidence: they jumped to extreme conclusions and rarely recognized when new data should not change a hypothesis (a behavior that will be familiar to anyone who has interacted with these tools). This is another powerful reminder that clinical reasoning is not just knowledge retrieval. It’s judgment under uncertainty: the ability to weigh evidence, tolerate ambiguity, and revise hypotheses carefully. For educators, that means our job hasn’t changed. LLMs can be a remarkable tool for teaching and reflection, but they cannot replace the human processes that make great clinicians. AI can help us practice those skills, and ideally even surface our blind spots, but it can’t yet replace the messy, human process of figuring things out. If we want clinicians who are safe and adaptable, we still have to teach reasoning, not just automate it. Take a read here: https://lnkd.in/eJqk2P5a

1

Is AI good for healthcare—or just good for hype? In this episode of Pondering AI, political scientist Michael Strange joins Kimberly Nevala to unpack the complexity of global health systems, the limits of #AI as a cure-all and why real innovation requires humility, nuance and interdisciplinary thinking. 🎧 Listen now: http://2.sas.com/60414eKVF #ArtificialIntelligence

9

Experts delve into the methodology, practical adoption strategies for community oncology practices, and the broader implications of AI's expanding role in shaping future oncology workflows. Read the full interview with Christine Vecchio and Eirini Schlosser here: https://hubs.ly/Q03vcnG00 | Cleveland Clinic Dyania Health

2

Estimating treatment effects in real-world healthcare is rarely straightforward. Differences in treatment availability and patient populations across hospitals, combined with strict privacy constraints, make causal inference particularly challenging. In Propensity Weighted Federated Learning for Treatment Effect Estimation in Distributed Imbalanced Environments, Alejandro Almodóvar, Juan Parras and Santiago Zazo from Universidad Politécnica de Madrid propose PW FedAvg, a federated learning approach designed to address these issues. By integrating a propensity weighting scheme with the TEDVAE model, their method accounts for treatment assignment imbalance across distributed data nodes. The results show clear gains in accuracy compared to traditional FedAvg and other FL-based models, offering a promising path forward for fairer and more precise treatment effect estimation in decentralised clinical settings. 👉 https://lnkd.in/e77acB8n #SYNTHEMA #CausalInference #FederatedLearning #HealthcareAI #TreatmentEffects

6

As AI evolves from insight generation to real-world impact, how will it reshape the future of care delivery? In Redefining the Boundaries of Medicine, John Halamka, M.D., M.S. and Paul Cerrato examine how advances in AI and data science are driving a shift toward more precise, personalized, and equitable healthcare — while underscoring the need to balance innovation with human-centered values and thoughtful governance. Read more: https://lnkd.in/gg3kY_vW #AI #HealthcareInnovation #DigitalHealth #HealthEquity #HealthIT #MayoClinicPlatform

25

1 Comment

🧠 AI isn’t just a tool in ophthalmology, it’s a responsibility we carry with every patient. In health tech and ophthalmology, artificial intelligence can do more than speed processes, it can improve outcomes, reduce disparities, and empower clinical teams. But its impact depends on thoughtful implementation. Here’s how to leverage AI effectively: 1️⃣ Focus on real clinical needs: Don’t adopt AI for the sake of novelty. Identify where it can meaningfully improve patient care or workflow efficiency. 2️⃣ Ensure transparency and validation: Algorithms must be explainable, validated, and continuously monitored to maintain trust. 3️⃣ Measure outcomes and iterate: AI solutions are only as good as the results they deliver. Collect data, analyze performance, and refine constantly. 💡 When implemented responsibly, AI enhances human judgment rather than replacing it, creating safer, more equitable care for patients and stronger decision-making for providers. 📚 Source: American Academy of Ophthalmology – AI in Ophthalmology #HealthTech #Ophthalmology #AIinHealthcare #DigitalHealth #ClinicalInnovation #PatientCare #MedicalTechnology #VisionCare #ArtificialIntelligence #FutureOfHealthcare #MedicalAI #HealthcareInnovation #PatientSafety #HealthEquity #PrecisionMedicine #EvidenceBasedMedicine #InnovationInMedicine https://lnkd.in/geC4Csmi

5

Real-world evidence (#RWE) generates the signals that change clinical practice. But most teams are navigating AI adoption, causal methods, and regulatory expectations all at once without infrastructure built for it. #BlackSwanCausalLabs exists to close that gap. Strategy consulting. Bespoke tooling. Causal inference expertise. Training. Regulatory support. All of it practitioner-led, at the frontier of AI and RWE science. BlackSwanCausalLabs.com #RealWorldEvidence #CausalInference #Pharmacoepidemiology #AIinHealthcare

1

As healthcare professionals navigate the AI landscape to improve processes and patient outcomes, we’re all asking the same question: How do we do this right? In a recent podcast, JCHI's Christopher Longhurst, UC San Diego Health joins fellow healthcare AI experts James Hoffman, St. Jude Children's Research Hospital & Paul Sharek, Seattle Children's to explore that question as well as what’s working, what’s not, and what might come in the future.   They discuss how AI is helping to reduce clinician burnout, improve patient safety, and identify hidden trends in quality data, but also highlight the risks of overpromising, the urgent need for rigorous evaluation, and the ethical responsibility to address health equity. Sneak peak from Dr. Longhurst ⤵️ "We're all excited about the opportunities to use these new (AI) tools to impact burnout. I think there’s also some more direct connections to quality improvement and patient safety that we see overtime. For example, as we roll out these AI scribe tools that may save clinicians a lot of time, they may also be able to provide things like diagnostic decision support, which is sort of the Holy Grail for health informatics." Listen ⤵️ https://lnkd.in/g54GMbV3 Hosts ➡️ Pedro Delgado & Amar Shaw

8

My colleague and friend Elie Gottlieb, PhD recently spoke with The AI Journal about the science behind our work at Sleep.ai and what it takes to build credible intelligence in health. The interview dives deep into topics like polysomnography, “forecast-to-action” modeling, and device-aware systems that can interpret data from different sensors and environments. My biggest takeaway is that the true measure of intelligence is interpretation. AI achieves impact, particularly in health, only when it has the human-expert guardrails to understand the origins, limitations, and meaning of the data used in these models. And sleep data only becomes meaningful when interpreted through context. Our models learn how multiple wearables and apps measure sleep and reconcile those signals into a coherent picture of how people rest and recover. Through that synthesis, sleep becomes an organizing layer across health, informing nutrition, fitness, mental health, women’s health, and chronic care programs where recovery and regulation shape outcomes. For anyone working to connect data, design, and decision-making in health, this is an insightful read. Link in comment. #SleepAI #SleepTech

44

26 Comments

At AACR 2026 Annual Meeting last week, Cellworks presented three studies demonstrating how mechanistic AI can help bring greater precision, access, and explainability to oncology: Huge thanks to our collaborators, especially Charu Aggarwal, MD, MPH, FASCO and Tejas Patil for their partnership in advancing this important work. What we shared: - Cellworks Study Supports Liquid Biopsy for Predicting Chemo-Immunotherapy Benefit in Advanced NSCLC. New findings show potential for liquid biopsy to expand access to personalized treatment prediction in advanced NSCLC. - Cellworks Expands Evidence for Predicting Chemo-Immunotherapy Benefit in NSCLC with New Early-Stage Disease Study. New study supports the predictive validly of Cellworks’ mechanistic AI platform in early-stage NSCLC. - Cellworks Presents Novel Mechanistic Approach to Predicting Drug–Drug Interaction Risk in Oncology at AACR 2026. Study suggests biologically grounded predictions could reduce safety-related failures in combination therapies in drug development and clinical care settings. If you didn’t catch us at AACR, we’d be happy to connect. #AACR2026 #PersonalizedOncology #NSCLC #PrecisionDrugDevelopment #Cellworks

21

📊 Data is exploding, but insights aren’t keeping pace—AI can change that. CytoReason is using AI-powered disease models to help pharma leaders turn complex biological data into actionable insights. On today’s episode of the NVIDIA AI Podcast, co-founder Shai Shen-Orr shares how they’re bridging the data-to-insight gap. 🎧 Listen to the full episode: https://nvda.ws/42vPrLi

123

4 Comments

STARGATE: A $500 Billion Leap Into the AI Future — But What About Humanity? With a $500 billion investment, Stargate is set to build the world’s most powerful AI data infrastructure designed to fuel the next wave of intelligence, automation, and planetary-scale innovation. But as we race ahead, we must ask: Are we building a future that serves humanity, or one that leaves too many behind? That’s where MED21 comes in. While Stargate lays the AI backbone, MED21 is pioneering how that intelligence can deliver equitable, affordable, and accessible healthcare to millions, especially those left out of today’s profit-driven system. Imagine AI not just powering financial markets or surveillance, but early diagnosis for a child in a rural village, or virtual care for a working mother without insurance. This is the future we must build: . Power + Purpose. . Technology + Compassion. . Innovation + Inclusion. MED21 – Where AI Meets Human-Centered Medicine. https://med21.us

17

4 Comments

Excited to share Medley, our vision for the future of human-AI collaboration in medicine. We're building a new paradigm that leverages AI diversity to tackle complex diagnostic challenges. See how it works!

1