"RAxSS: A Novel Approach for Medical Time Series Classification"

a new study combined electrocardiograms (ECG) with vectorcardiography (VCG) — a 3D view of the heart’s electrical signals — to detect thickened heart walls, or left ventricular hypertrophy (LVH), more accurately. using a transparent AI model, researchers created the “Marcos” system, which spotted LVH with about 75% accuracy, better than traditional ECG rules that often miss early cases. this approach could help doctors diagnose heart strain earlier using standard ECG data, without extra tests — a step toward smarter, more accessible heart care. Source: https://lnkd.in/eWvJ_JJV Published Date: October 16, 2025 👉 Comment BIOHACK if you want more science and health news like this! #BiohackYourself #HealthNews #ScienceNews #ResearchUpdates #Biohack Disclaimer: This content is for educational and entertainment purposes only and is not a substitute for medical advice. Always consult a healthcare professional. Full disclaimer: https://lnkd.in/eJE9Rsty 🧠 We explore all angles — ancient wisdom, modern science, and everything in between. No allegiance to Big Pharma or Big Natural. 🔍 We cite studies, but encourage you to read them, question funding, and review the methods. Stay curious. 📚 Not all journals are equal. Peer-reviewed ≠ perfect. Check the source, think critically, and decide for yourself. ⚠️ One study isn’t the full story. Science evolves. We’re here to inform, not to tell you what to believe.

Protocol accepted at JMIR Research Protocols We are testing a simple idea with real impact for busy clinicians: Can an AI generated podcast help residents learn from research articles as effectively as reading In our study, resident physicians will engage with two cardiology papers that differ in complexity • one as a short podcast summary generated by AI • one by traditional reading We will compare comprehension and also look at motivation and cognitive load. The design lets each participant both listen and read so we can make a fair comparison. Why this matters: If AI audio holds up, it could offer a practical way to keep up with the literature on the move. If it does not, we will learn where reading remains essential and how to design better audio summaries. Data collection starts soon, so stay tuned. Authors: Matthias Stadler, Constanze Catharina Richters, Martin R. Fischer, and Fabian Hutmacher Preprint: https://lnkd.in/d2PRwRaq #MedicalEducation #AIinEducation #ResidencyTraining #ResearchProtocols

Researchers have developed BONE-Net, a hybrid AI model that detects osteoporosis from knee X-rays with high accuracy (86%) and specificity (94%). By combining DenseNet169 and Vision Transformer (ViT) architectures with a custom Attention Module, BONE-Net captures both local and global bone features for early and precise diagnosis. This innovation could support faster, low-cost osteoporosis screening and help clinicians intervene before fractures occur. Source: https://lnkd.in/eHRnaPeE Published Date: October 15, 2025 👉 Comment BIOHACK if you want more science and health news like this! #BiohackYourself #HealthNews #ScienceNews #ResearchUpdates #Biohack Disclaimer: This content is for educational and entertainment purposes only and is not a substitute for medical advice. Always consult a healthcare professional. Full disclaimer: https://lnkd.in/eJE9Rsty 🧠 We explore all angles — ancient wisdom, modern science, and everything in between. No allegiance to Big Pharma or Big Natural. 🔍 We cite studies, but encourage you to read them, question funding, and review the methods. Stay curious. 📚 Not all journals are equal. Peer-reviewed ≠ perfect. Check the source, think critically, and decide for yourself. ⚠️ One study isn’t the full story. Science evolves. We’re here to inform, not to tell you what to believe.

#AINews 🌍 Leveraging LLMs to Revolutionize Stroke Identification and Healthcare Automation UT Southwestern Medical Center's groundbreaking study reveals that large language models

New hybrid EDL + 3D-CNN model uses EEG to detect early Alzheimer’s — achieving ~99% accuracy on public datasets. It ensembles five 3D CNNs across frequency bands (delta → gamma), learns from raw signals (no hand features) and beats many prior EEG/MRI approaches. Promising, non-invasive screening — but needs bigger, diverse cohorts and explainable AI for clinical adoption. Source: https://lnkd.in/e6Y2UE3Y Published Date: October 12, 2025 👉 Comment BIOHACK if you want more science and health news like this! #BiohackYourself #HealthNews #ScienceNews #ResearchUpdates #Biohack Disclaimer: This content is for educational and entertainment purposes only and is not a substitute for medical advice. Always consult a healthcare professional. Full disclaimer: https://lnkd.in/eJE9Rsty 🧠 We explore all angles — ancient wisdom, modern science, and everything in between. No allegiance to Big Pharma or Big Natural. 🔍 We cite studies, but encourage you to read them, question funding, and review the methods. Stay curious. 📚 Not all journals are equal. Peer-reviewed ≠ perfect. Check the source, think critically, and decide for yourself. ⚠️ One study isn’t the full story. Science evolves. We’re here to inform, not to tell you what to believe.

Stanford’s #OpenTSLM Breaks the “Modality Gap” in #MedicalAI Unlike vision-based workarounds, OpenTSLM-Flamingo explicitly encodes time series as a unique modality, achieving #SOTAperformance on clinical reasoning benchmarks and generating transparent, chain-of-thought rationales. Researchers at Stanford, ETH Zurich, Google Research, and Amazon introduced OpenTSLM, the first Time-Series Language Model natively designed to process continuous medical signals (e.g., ECG, EEG) within LLM frameworks.

Don’t miss the presentation by Dane Morey, PhD at the upcoming HFES - Human Factors and Ergonomics Society #ASPIRE25 conference! He’s presenting our work on investigating the impact of different data collection structures on evaluating nurses' understanding of an AI-infused patient data display. We found that asking people to respond freely produced meaningfully different responses than when asking them to respond with a set of predetermined options. Methods matter! 💡 Towards Scalable Research: Comparing Qualitative and Quantitative Data Collection Mechanisms 🗓️Thursday, October 16th | 5:10pm - 5:30pm CDT 📍 Grand Hall K If you are interested in the methodological challenges of scaling AI evaluation, join us for the discussion! #HFES #CognitiveSystemsEngineering #AIEvaluation #HumanFactors

𝐀𝐈 𝐟𝐨𝐫 𝐆𝐨𝐨𝐝 🔬 Have you ever heard about 𝐏𝐫𝐢𝐦𝐚𝐫𝐲 𝐂𝐢𝐥𝐢𝐚𝐫𝐲 𝐃𝐲𝐬𝐤𝐢𝐧𝐞𝐬𝐢𝐚 (PCD) before? How this rare, inherited condition causes respiratory distress early in life, leading to a lifetime of chronic infections? If not -and also if you have-, read on: with October being PCD Awareness Month, we are proud to highlight the story from the team at 𝐑𝐨𝐲𝐚𝐥 𝐁𝐫𝐨𝐦𝐩𝐭𝐨𝐧 𝐇𝐨𝐬𝐩𝐢𝐭𝐚𝐥 & 𝐆𝐮𝐲’𝐬 𝐚𝐧𝐝 𝐒𝐭 𝐓𝐡𝐨𝐦𝐚𝐬’ 𝐍𝐇𝐒 𝐅𝐨𝐮𝐧𝐝𝐚𝐭𝐢𝐨𝐧 𝐓𝐫𝐮𝐬𝐭 who are continuously looking for ways to automate key aspects of the complex, tedious diagnostic process to reduce wait times for patients in need of care. In the latest episode of the 𝐆𝐞𝐭𝐢™ 𝐯𝐢𝐝𝐞𝐨 𝐬𝐞𝐫𝐢𝐞𝐬, Guy Tamir speaks with Dr. Mathieu Bottier about how an AI model trained with Geti assists clinicians and reduces the time needed to assess electron microscopy images from 1–2 hours to just 1–2 𝘮𝘪𝘯𝘶𝘵𝘦𝘴 (!). This is what AI for Good looks like. 𝐖𝐚𝐭𝐜𝐡 𝐭𝐡𝐞 𝐟𝐮𝐥𝐥 𝐢𝐧𝐭𝐞𝐫𝐯𝐢𝐞𝐰 𝐡𝐞𝐫𝐞: https://lnkd.in/eJFec29X. Special thanks to Professor Claire Hogg, Mathieu Bottier, PhD, Tom Burgoyne - your work continues to inspire! Start building AI today with Geti: 🔗docs.geti.intel.com 🔗https://lnkd.in/gZxv7__6 Yannis Katramados Guy Tamir Dan C Rodriguez Dr. Oliver Hamilton Aleksey Burdakov #Intel #Geti #HealthcareAI #MedicalImaging #PCD #NHS #ComputerVision #NoCodeAI

I’ve heard this from several medical and veterinary researchers: “We want to share our knowledge. We’ve been taking and annotating pictures for years but we need an AI expert to help scale this out so clinicians can actually use it.” Geti enables researchers and specialists to share their expertise and build AI models without needing to be AI experts. It’s incredible to see this growing wave of AI empowerment and the openness across the scientific community to collaborate, share, and teach so others can care for more complex patients. Especially in rural or resource-limited areas, where specialists are hard to come by, AI tools created by experts in their field can empower generalists to serve patients they otherwise couldn’t. That’s powerful. Geti is part of Intel’s Open Edge Platform and is an open-source tool designed to make AI accessible: docs.geti.intel.com Happy creating and innovating.

Grateful to see BrainBit featured by Business Insider for our work on AI-driven, cross-species neurofeedback and long-term R&D. We’ll keep building the practical layer of neurotech: non-medical EEG hardware, software & tools, and a developer-first SDK and collaborating with neurofeedback practitioners, research teams and labs, partners. #Neurotech #EEG #BCI #Neurofeedback #AI #DigitalHealth #GlobalRecognitionAward2025