Vincent Grek

In The New Yorker, Tad Friend captures the collision of #entrepreneurship, #science, and #philosophy in the quest to radically extend human #lifespan. Visionaries like Peter H. Diamandis are rallying investors and technologists around a bold promise: that exponential advances; AI, gene reprogramming, stem-cell therapies, and continuous biomarker tracking, could unlock biological immortality. Figures such as Bryan Johnson embody this future-forward enthusiasm with their extreme, highly public health regimens. Yet the audience need to be reminded that longevity is not about billionaires outpacing biology, it’s about rigorous, evidence-based precision geromedicine. We emphasize that while breakthroughs are coming, they must be tested in clinical trials, translated responsibly into practice, and made accessible to society at large. Our perspectives anchor the discussion in science and equity, highlighting that true progress means ensuring that added years are lived in #health. Whole article: https://lnkd.in/dMwmVeMw Eric Verdin, Nir Barzilai, Steve Horvath NUS Academy for Healthy Longevity

119

12 Kommentare

Nina Capital is pleased to formally announce our investment in Datamonk. Datamonk has successfully raised pre-seed funding to address the significant challenge of migrating decades of medical #imaging #data, by developing an agentic AI-powered platform that automates migrations. The funding round was co-led by Healthy.Capital and Nina Capital, with participation from angel investors including Jeroen Tas, former Chief Innovation & Strategy Officer at Philips, and Harm-Jan Wessels. The capital will support platform development, the advancement of AI capabilities, and team growth. *What's the need?* Hospitals maintain decades of imaging data in Picture Archiving and Communication Systems (#PACS). PACS vendors and healthcare providers face operational and financial pressure to migrate massive volumes of #radiology data—often hundreds of terabytes per hospital—from legacy on-premise PACS systems to modern #cloud-based infrastructures, yet today’s migration processes remain largely manual (consultant-led), expensive, and time-consuming, often taking a year or more. *What's the solution?* Datamonk’s solution leverages agentic AI to automate these migrations and concurrently improve data quality. The software agents detect and resolve metadata issues, standardize study naming, and validate data integrity while the migration runs. This allows hospitals to go live faster with data that is immediately ready for clinical workflows, research, and AI applications. At Nina's, we are excited about joining efforts towards a scalable way to standardize and migrate imaging data, creating the foundation for interoperability and future innovation. The Datamonk team was founded by Jaap Gielink (serial entrepreneur), Jai Bhatia (co-founder of Viewics Inc., acquired by Roche Diagnostics), and Matthew Condron (former CTO of a leading healthcare data migration consultancy). We are delighted to partner with Datamonk as they build the data layer for modern healthcare and solve this fundamental infrastructure challenge. Link: https://lnkd.in/dwdH7vew #wearenina #healthcare #HealthTech #MedicalImaging #AIinHealthcare at Tech.eu by Tamara Djurickovic

77

9 Kommentare

3 Healthcare AI papers I'm reviewing today 1. 📚AI in Medicine: Medical multimodal foundation models in clinical diagnosis and treatment: Applications, challenges, and future directions https://lnkd.in/eNV5CDSp Medical Multimodal Foundation Models (MMFMs) combine diverse data types (imaging, text, labs) to improve diagnosis, treatment planning, and precision medicine. Recent advances in large datasets and multimodal architectures (vision-only and vision-language) enable strong generalization across tasks like segmentation, classification, and clinical report generation. Key opportunities lie in holistic integration of multi-organ/multimodal data, but challenges remain in optimizing representations and scaling real-world clinical adoption. 2. 📚 BMJ Dig Health & AI - Optimising large language models for clinical information extraction: a benchmarking study in the context of ulcerative colitis research https://lnkd.in/erXhhX9k This study compared open-source and closed-source LLMs for extracting the Mayo Endoscopic Subscore from colonoscopy reports. It found that QLoRA fine-tuning improves open-source performance significantly, but GPT-4o with prompt engineering still outperforms them by 5–11% and is more cost-effective. Overall, GPT-4o is the most efficient option today, while QLoRA-optimized open-source models are viable fallbacks, though both leave room for improvement in instruction following. 3. 📚 JAMIA Open - Generative artificial intelligence for automated data extraction from unstructured medical text https://lnkd.in/ekfy8-VX A GenAI pipeline using an open-source LLM was developed to extract structured data from right heart catheterization notes with built-in guardrails and a retry mechanism. It achieved high performance (99% precision, 85% recall, 91.5% F1, 90% accuracy), with missed values as the main error and hallucinations extremely rare (<0.01%). The study shows LLM pipelines can reliably mine unstructured clinical data, improving research efficiency and clinical applications.

81

1 Kommentar

🚨 Delighted to share our latest publication in Mayo Clinic Proceedings: Digital Health on “Deep Learning Applications in Clinical Cancer Detection: A Review of Implementation Challenges and Solutions” https://lnkd.in/gj2jPZtq In this comprehensive review, my co-authors and I analyzed over 1,400 studies from 2018–2024, spanning imaging, genomics, pathology, and multimodal diagnostics. We highlight: ✅ The tremendous promise of deep learning (DL) in improving early cancer detection ✅ Persistent challenges in clinical adoption—including interpretability, bias, and regulatory hurdles ✅ Cutting-edge solutions, such as federated learning, explainable AI, synthetic data generation, and real-world trials ✅ A call for interdisciplinary collaboration to ensure AI tools are ethically, equitably, and effectively deployed in diverse healthcare settings I have always supported the integration of AI into oncology especially when I served as MD/CEO of the National Cancer Centre Singapore from 2017-2023 and even took a course with the MIT Sloan School of Management on of AI in Healthcare back in 2020. So when I had two junior college students Isaiah Yao and Dong Min approach me for an elective posting I set them a task with some directions on getting this paper done. They surprised me with excellent work done on this and this paper is a culmination of many months of work together on this. Deep learning is not a distant promise—it’s already transforming diagnostics. The question now is: How do we scale it safely and fairly for all? Grateful to my co-authors and to REMEDIS for supporting this work. Also to Yinxia Chao of Duke-NUS Medical School for referring the students to me. #AIinHealthcare #CancerDetection #DeepLearning #MayoClinic #DigitalHealth #OncologyInnovation #ExplainableAI #NCCS #REMEDIS #artificialintelligence Mayo Clinic SingHealth

82

6 Kommentare

Synthetic data is emerging as an important training tool in medical AI. At Stanford Medicine, our team developed RoentGen, a model that generates realistic chest X-rays from clinical text. This approach can help train AI systems in areas where real data are scarce so models can be built and tested even when real-world examples are limited. https://lnkd.in/gRUJmX6q

116

6 Kommentare

What’s the quickest way to integrate digital health applications into the healthcare system and increase the adoption of digital health solutions?   The simplest answer: monetary compensations, also known as financial reimbursement. In 2019 Germany became the first country in the world to integrate digital health applications (DiGAs) into its statutory healthcare system, with the adoption of the Digital Healthcare Act (DVG). Currently there are 44 applications with permanent and 14 with temporary/provisional approval listed in the DiGA directory. Another 11 applications (DiGAs) had been delisted, mainly because no evidence of a positive healthcare effect could be provided. In 2024, the average price per application prescription was €541, with prices varying significantly, ranging from €119 to €2077 per use. Use was often define as 90 days prescription.   Researchers from the Center for Evidence-Based Healthcare, Faculty of Medicine, Universitätsklinikum Carl Gustav Carus Dresden, University Hospital Dresden, Germany (Khira Sippli; Stefanie Deckert; Jochen Schmitt; Dr. Madlen Scheibe) recently conducted a systematic review of the evidence of positive healthcare effects. Here are their main findings: 👉🏼 The researchers reviewed 23 approval studies for 21 DiGAs  👉🏼 For 21 of the approval studies, the study design was a parallel, two-arm randomized controlled trial 👉🏼 The intervention duration ranged from 1.5 to 12 months, with a mean duration of 3.8 months. 👉🏼 Sample sizes varied widely, with the number of participants ranging from 56 to 1245, with a mean of N = 321 (intervention and control groups combined).  👉🏼 For all 21 studies, there was a significant between group effect at post-assessment. 👉🏼 All outcomes, apart from one, were measured by patient-reported outcome measures (PROMs). 👉🏼 All reported significant medium to large effects, but risk of bias was high, raising concerns about the robustness of the evidence supporting efficacy This systematic review highlights that the digital health application approval process exhibit potential for methodological improvement and should be closely monitored.

52

4 Kommentare