Um die Verwirrung komplett zu machen ist glaube ich der Ostersonntag der nach dem ersten Frühlingsvollmond. Dadurch kann es Konstellationen geben an denen in einem Jahr der Karfreitag vor und im anderen nach dem ersten Frühlingsvollmond liegt 🧐☝️😇
Posts by Stephan Ellmann
With everything happening in the US right now, how safe are scientific conferences for foreign participants in the US? I think this is something we have to discuss as a scientific community. I am hesitating to register for US meetings due to the unpredictable situation. Others who feel the same?
Diese Überraschung, dass #Merz es nicht kann. Weder politisch, noch menschlich oder strategisch. Hält sich in Grenzen bei mir.
Alternativer Kanzler (Wüst oder Günther). Rücktritt Merz und Söder. Fester Ministerposten für Habeck trotz Opposition.
Und eine davon auch noch schwarz.
Deutsche gehen heute in alte, verfallene, kaputtgesparte Schulen und Turnhallen und wählen CDU, keine Pointe
Wahnsinn! Herzlichen Glückwunsch und höchsten Respekt!
Bin auch mit im Club 👍😷 Auf ein sicheres 2025!
Schlechte Google Rezension wegen Wartezeit in der Praxis trotz sehr schneller Terminvergabe.
Termine für #GKV Patient:innen sind Mangelware? Bei uns gab's einen MRT-Termin in weniger als 24h. Der Dank? Eine 2-Sterne-Bewertung wegen 2h Wartezeit. #Fairness beginnt bei gegenseitigem #Verständnis. #Gesundheitssystem #Kassenpatienten #PKV #Terminvergabe #KV #Budgetierung
On this day, a very special baby was born who would grow up to revolutionise the world, uniting heavens and Earth in one. Even today, his disciples continue to work to better understand the universe he illuminated.
His name was Isaac Newton! 🧪 ⚛️
PS: Merry Christmas! 🎄
#Medicine #ArtificialIntelligence #Radiology #DigitalHealth #MedicalAI #Healthcare #ResearchMethodology #PeerReview #AcademicPublishing
What's your take on maintaining scientific rigor in AI healthcare publications? How can we better evaluate AI capabilities in medicine? 📚 Source: BMJ 2024;387:e081948
capabilities ✅Focus on practical, validated assessment methods for AI in healthcare
𝐌𝐨𝐯𝐢𝐧𝐠 𝐅𝐨𝐫𝐰𝐚𝐫𝐝: We need: ✅More 𝚛̲𝚒̲𝚐̲𝚘̲𝚛̲𝚘̲𝚞̲𝚜̲ ̲𝚙̲𝚎̲𝚎̲𝚛̲ ̲𝚛̲𝚎̲𝚟̲𝚒̲𝚎̲𝚠̲ for AI-related medical research ✅Clear distinction between human cognition and AI processing
limitations. The anthropomorphization of AI systems through human cognitive testing frameworks doesn't advance our understanding of these tools.
𝐖𝐡𝐲 𝐓𝐡𝐢𝐬 𝐌𝐚𝐭𝐭𝐞𝐫𝐬: This publication raises concerns about peer review standards in AI healthcare research. The conclusions could mislead healthcare professionals about AI capabilities and
🔍 Results simply confirm the expected: newer models perform better across ALL tasks ⚠️ Drawing parallels between human cognitive decline and model versions is scientifically unsound
Newer versions are complete replacements, not aged iterations
𝐂𝐫𝐢𝐭𝐢𝐜𝐚𝐥 𝐈𝐬𝐬𝐮𝐞𝐬: 🧩 Fundamentally flawed premise: Applying human cognitive tests to AI systems lacks scientific validity 📊 The "aging" comparison misrepresents how AI models evolve.
⁉️Authors suggest AI shows "cognitive aging" similar to humans⁉️
🔹Only ChatGPT-4o achieved a "normal" cognitive score (26/30)🔹All models showed difficulties with visuospatial tasks🔹"Older" models performed worse than newer versions
The authors administered the Montreal Cognitive Assessment (MoCA) to leading AI models (ChatGPT-4, Claude, Gemini), reporting that:
𝐒𝐭𝐮𝐝𝐲 𝐎𝐯𝐞𝐫𝐯𝐢𝐞𝐰:
1735023156058.jpeg
🤖 Recent Hashtag#BMJ Study Claims "Cognitive Decline" in AI Models - Let's Examine This Critically #AIinMedicine #ResearchQuality
Wissenschaft braucht Reichweite. Daher werbe ich fürs folgen meiner Universität @fau.de
Bitte unterstützt das by reposting.
Wobei mir das mit dem Spanferkel schon gefällt 😂
Scientists, academics, researchers: We’re excited to share that @altmetric.com is now tracking mentions of your research on Bluesky! 🧪
We are now in the "everything" era of multimodal medical AI.
Great to see this inspiring talk by @erictopol.bsky.social at @rsnasky.bsky.social 🥳
" #Radiology is the leader & pacesetter [for medical AI]...We are all learning from you."