Sabitlenmiş Tweet
Excited to announce I've joined the WHO Fides Network! 🌍🏥 Committed to sharing reliable health info and combating misinformation. Stay tuned for evidence-based content to keep you informed and healthy. #WHOFides
English
Dr.Mathan K 🇮🇳
62.3K posts
Dr.Mathan K 🇮🇳
@kmathan
Epidemiologist
Tamil Nadu, India Katılım Ocak 2008
2.2K Takip Edilen3.3K Takipçiler
Dr.Mathan K 🇮🇳 retweetledi
BMIU @OffCMCVellore welcomes you to join for a Webinar on AI Governance where Prof.@ravi_iitm will be speaking with a special focus on #Health. Link: cmcv-in.zoom.us/meeting/regist… Meeting ID 831 0186 2993 Passcode 696027 #responsibleai #ai4health #aigovernance #safeai
English
Dr.Mathan K 🇮🇳 retweetledi
The Selected Works of Jawaharlal Nehru are now available online in an easy-to-read format. They shall be used by scholars, students, and history buffs, though it may be too much to hope that the Congress High Command avails of them.
Here is the link: nehruarchive.in/selectedworks
English
Most of the hospitals aren’t even aware of this impact via DPDPA, those who know are figuring out a way to fix that.
OpenMed@OpenMed_AI
Indian patient data processed through US cloud APIs violates DPDPA. Doesn't matter how good the API is. If data crosses borders, you're non-compliant. On-premise isn't a preference. For 1.4 billion people, it's the law. How many hospitals still route PHI through cloud APIs?
English
Really nice to see someone flag this area @thekaransinghal , IMO Evaluation methodology in health AI is arguably the most underdiscussed bottleneck in the field right now.
A few thoughts from a clinical epidemiology lens that might push the discussion further:
The ecological validity argument is valid and strong. Answering MCQs clearly don't capture real world use. But the same argument applies in reverse, we need to know how untrained users interact with these systems, not just how models perform under ideal conversational conditions.
HealthBench is an interesting instrument, but the field would benefit enormously from knowing more about rater selection, blinding, inter-rater reliability, and conversation sourcing. Can we use that as a reproducible scientific tool, is the real question.
The 99% triage figure needs context — CIs, sample size for the emergency subset, and the operational definition of "emergency" would help the community interpret it properly.
And the biggest gap: we're still entirely in the realm of process measures. When we start measuring those hard outcomes like say …does this actually change morbidity, mortality, time-to-diagnosis?
The questions you pose at the end deserve a serious, multi-stakeholder answer. Looking forward to seeing this evolve.
English
Dr.Mathan K 🇮🇳 retweetledi
📢 Applications Open - A Short Course on Digital Health by team DHRUVA
DHRUVA (Digital Health for Responsive, User-centered, Viable Applications) Team invites applications for a 3-day in-person short course on Digital Health, scheduled 1–3 June 2026 at the Indian Institute of Science (IISc) Bengaluru campus.
✳️Digital health interventions fail when the right people aren't talking to each other.
Researchers produce evidence that never reaches implementers.
Founders build tools that funders don't understand.
Designers create products clinicians won't use.
The gap isn't technical — it's relational.
That's why this is designed specifically for the people who sit at every node of that network:
→ Researchers & epidemiologists
→ Startup founders & product builders
→ Public health administrators & program implementers
→ Venture capitalists & impact investors
→ NGOs & civil society organizations
→ Industry professionals & health system partners
→ UX/service designers in health
When these voices share the same table, digital health actually works.
#DigitalHealth #HealthTech #PublicHealth #GlobalHealth #HealthInnovation
English
Dr.Mathan K 🇮🇳 retweetledi
“Jasprit Bumrah is the Bradman of bowling, that man is a genius and we are blessed to live in his times”
TFW World Cup Daily out now in your feeds and YouTube.
English
Singularity is not far away!
Adam Carlson@admcrlsn
One of the greatest charts I have ever seen
English
How and when to use causal and associational language
✳️This article explores the critical importance of using causal language in medical and epidemiological research when the underlying study objective is to determine cause and effect.
⚠️ The authors argue that avoiding such terminology in favour of vague words like "association" creates ambiguity, making it difficult to define research questions or address potential biases.
💡 Ultimately, they suggest that being explicit about causal intentions encourages transparent reporting and more rigorous scientific thinking.
doi.org/10.1136/bmj-20…
English
News: OpenAI + OpenClaw.
Anthropic responds 👇🏻
Claude@claudeai
New in Claude Code: Remote Control. Kick off a task in your terminal and pick it up from your phone while you take a walk or join a meeting. Claude keeps running on your machine, and you can control the session from the Claude app or claude.ai/code
English
Wispr Flow android is here
wisprflow.ai/android
Tanay Kothari@tankots
We offered 5 people a Porsche 911 GT3 RS if they could get @WisprFlow to make a mistake It's the fastest and most accurate AI voice dictation app that's 3x more accurate than ChatGPT, Claude, or Siri. Today, we’re finally launching on Android. Download now: play.google.com/store/apps/det… As a part of the launch, we’re giving away 6 months of Wispr Flow Pro for free. Like, retweet and comment ‘Wispr Flow’ to get it. Enjoy. — Written with Wispr Flow
English
✳️ The 8 core statistical ideas from the past 50 years and their future applications.
❇️ The synergy between these ideas, big data, and modern computational power.
🌐 Most important statistical ideas of the past 50 years_Gelman | Tea Tasting Statisticians tea-tasting-statisticians.github.io/posts/Most-imp…
English
Trial emulation has turned a corner.
What next?
Counter factual simulations?!
Phylo@phylo_bio
Introducing agentic trial emulation: Biomni scaled clinical trial emulation using 12M patients EHR data at Mount Sinai. Researchers used Biomni to autonomously emulate multiple landmark anticoagulation RCTs end-to-end within their EHR—turning months of manual target trial analysis into scalable, agent-driven workflows that can extend across large numbers of trials and diseases. Running many emulations in parallel enabled something new: learning a stable institutional transport signal that quantifies how published trial effects systematically translate in local care. Read more: phylo.bio/blog/scaling-a… Exciting work with @BenGlicksberg @joshualampertmd Justin Kauffman et al
English
Hospitals cannot treat doctors like workmen in a factory: Madras High Court - thehindu.com/news/national/…
English
Dr.Mathan K 🇮🇳 retweetledi
Essentially a ‘human based world model’ layer exists in between the LLMs layer and the data layer.
Until we figure the ‘causal underpinnings’ of these world models, AGI will remain elusive.
Bo Wang@BoWang87
Professor Judea Pearl — the pioneer who invented causal reasoning in AI — says scaling won't save us. "Mathematical limitations that are not crossable by scaling up." The brutal truth: LLMs aren’t learning how the world works. They are learning how we describe the world. This resonates with most biologists: Drug discovery is hitting the same wall. We have mountains of genomic data, but most AI models just find patterns in published papers — not in the raw biology itself. They're learning what scientists think causes disease, not what actually does. Pearl's causal revolution? That's how we move from "this gene correlates with cancer" to "this gene causes cancer" — and finally design drugs that work. Until then, we're building very expensive parrots.
English
Dr.Mathan K 🇮🇳 retweetledi
