Physician/Intensivist @ClevelandClinic, CritCare, Tech Geek, Animal lover. Passionate about ML/AI, MedTech and Innovating NextGen Critical Care Solutions.
Cleveland, OH
Joined July 2008
- Tweets5K
- Following447
- Followers1.1K
- Likes1.1K
Pinned Tweet
Here is an App I created!
snapify.mdcrit.com/
Stop drowning in ICU data.
Start thriving with Snapify.
Get your patients' story, instantly clear.
Saves me a great deal of time! Check it out!
#CriticalCare #ICU #App #Medicine
#MedTwitter #FOAMed
Furquan Pathan, MD retweeted
the most important thing you can do is pray for the miraculous conversion of all those that believe our purpose on earth is simply a stepping stone on the path to inhuman intelligence and are unable to see the beatific vision we are a part of
Furquan Pathan, MD retweeted
I’m so honored to be featured by the Claude team! What a great surprise to get during clinic AngioView.com
Meta has quietly taken the AR/VR crown, not just the hardware, but the whole stack. Apple's Vision Pro had the polish. It never had a compelling use case. AR development for it was way behind.
This is exactly what I’ve been waiting for 🔥
Meta just announced its new VR Glasses. Basically an Apple Vision Pro-like experience, but MUCH smaller and lighter.
• Only 100g
• Micro-OLED displays
• Snapdragon Reality Elite
• Eye + hand tracking
• External compute/battery puck
• $1,299
This is the direction I want VR to go. Smaller, lighter, more comfortable, while keeping the full spatial computing experience.
Launching Spring 2027. I’m getting these 100% 🙌
Another very good demonstration of how AI can pick up on real micro data patterns of association we humans never contemplated. A chest X-ray should not be able to flag cervical or colon cancer a year early because those organs are not even in the picture, yet Oncoformer does it by treating the film not as a search for a hidden mass but as a faint image of whole-body stress - converting pixels and routine labs into the same numerical space so it can pick up micro-changes in heart outline, bone density, and lung texture that track systemic inflammation and metabolism (reflective of distant Cancer pathophysiology) rather than local anatomy.
The informativeness of a chest X-ray decoded by AI
In this multimodal AI model called Oncoformer of 3.67 million individuals of EHR plus chest X-ray, cancer was predicted at least 1 year before diagnosis: AUROC all cancers 0.869, cervical cancer 0.905, colon cancer 0.896 @CellCellPress open-access
cell.com/cell/fulltext/S0092…
Frankly do not trust any single entity, state, people to do this. For the sake of humanity this tech needs to be decentralized and distributed across societies. Else we would be heading towards dystopian future.
I’m skeptical that the current calls to “slow down” mean frontier labs intend to slow their internal progress.
The more plausible distinction may be between slowing deployment and slowing capability development. Those are very different things. Labs can genuinely worry about safety, misuse, and social disruption while still believing that unilaterally reducing their own research pace would be strategically irrational.
That creates a strange equilibrium: everyone may want more caution, but nobody wants to be the one who gives up the lead.
The China question makes this even more interesting. Chinese labs have repeatedly used open-weight, lower-cost models to compress the gap with closed frontier systems and put pressure on their economics. If Western labs become more conservative about releasing their strongest models, it is not obvious whether Chinese labs would follow, or see an opportunity to gain influence through openness.
I also would not assume that public benchmarks tell us exactly where the frontier really is. The most capable systems may increasingly be evaluated, used, and improved privately before the public sees them.
That, to me, is the more important question than whether AI development “slows”: If capability continues accelerating while public access becomes more restricted, who gets access to the best models and who gets to decide?
The future of AI may be shaped less by a race between models than by a race over access.
Furquan Pathan, MD retweeted
A wearable patch that monitors antibiotic levels (vancomycin) in real time just received an FDA nod!
How important can it become? Just ask any physician how hard it is to deal with finding the right dosage for antibiotics!
"Vancomycin is a widely used antibiotic for treating life-threatening infections like sepsis, but it’s notoriously difficult to dose correctly, with doctors typically relying on periodic blood draws to track it.
Nutromics’ patch contains microneedles, which are coated with synthetic DNA-based sensors that can identify the target they are sensing and bind with it by folding."
This new FDA Breakthrough designation could help the company reach commercial launch in the US next year. The company is now working to launch further studies with US hospitals this year.
Source: medcitynews.com/2026/09/nutr…
Furquan Pathan, MD retweeted
The 3 reasons why AI won't reduce clinician jobs:
—Jevon's paradox (more efficiency, more use)
—"Lump of labor" fallacy (the type of work changes)
—Tasks ≠ Skills
@NEJM @DhruvKhullar
nejm.org/doi/full/10.1056/NE…
AI is already changing how students learn, how faculty teach, and whether traditional assessments still measure actual understanding. However, AI can reduce the “cognitive friction” and social learning that help develop problem-solving, independence and judgment. At the same time, AI can improve learning, accelerate experimentation and become an important professional skill. A new report from MIT details its findings and puts some recommendations to redesign curricula and assessments for an AI-rich world. It requires clear course-specific rules on when AI can/cannot be used, emphasizes hands-on and social learning, teach students to use AI effectively, ethically and critically; preserve the residential/human learning experience; and continuously reassess policies as AI evolves. The aim - create an “AI-aware” education that uses its advantages without outsourcing the development of human thinking.
aiandeducation.mit.edu/repor…
Furquan Pathan, MD retweeted
Trauma arrest; does airway come before hemorrhage control?
Mohamed Hagahmed, MD (@hagahmedmd) from ResusX:Rounds:
The correct sequence might flip what you memorized. What order do you go in? Comment your answer.👇
We have arrived early AGI period indeed!
GPT-6 Astra, trained on ~100K+ NVIDIA Grace Blackwell NVLink72. From ChatGPT to o1 to Astra in 4 years.
AGI has arrived. Congratulations @OpenAI team.
400K GPUs coming online next.
Furquan Pathan, MD retweeted
GPT-6 Astra, trained on ~100K+ NVIDIA Grace Blackwell NVLink72. From ChatGPT to o1 to Astra in 4 years.
AGI has arrived. Congratulations @OpenAI team.
400K GPUs coming online next.
Watching GPT6 Astra rollout along with early testers posting the kind of work it is capable of, indeed really feels like early AGI has arrived! And I can’t tell if I should be celebrating or bracing.
Furquan Pathan, MD retweeted
Nueva revisión de hiperkalemia aguda (BMJ 2026). Lo que cambió, y que probablemente sigas haciendo:
🔴 Dextrosa al 10%, no al 50%. La glucosada hipertónica puede EMPEORAR la hiperkalemia.
🔴 Insulina 5 U (o 0.1 U/kg, máximo 10 U). Misma caída de potasio que con 10 U, con casi la mitad de hipoglucemias (OR 0.55).
🔴 1 g de gluconato de calcio no sirve. 97% necesitó tres dosis de 1 g y una sola dosis no funcionó en NINGUNO. Dosis: 1-3 g.
🔴 El calcio no "estabiliza la membrana". Restaura la conducción por canales de calcio. Traducción práctica: sirve cuando hay QRS ancho, no en la onda T picuda aislada.
🔴 Bicarbonato: no se recomienda de rutina.
🔴 La onda T picuda aparece en menos del 25% de los pacientes con K de 5.5-7.0. Un ECG normal NO descarta hiperkalemia.
🔴 Reevalúa el potasio a los 60 min y otra vez a las 2-4 horas. El rebote es frecuente
El tema ya está re-escrito y actualizado en Memodi y Accio. Completo en el canal (t.me/memodiapparticulos).
Furquan Pathan, MD retweeted
By next year, using a computer will be optional. Work will radically change.
I’m genuinely impressed with HeyClicky, the screen-drawing AI buddy app is executed so well. @FarzaTV What if you took it hardware? A KVM-style device that intercepts keyboard, video, and mouse so the AI works on every OS without software restrictions.
At the turn of the 20th century, anesthesia besides antibiotics has a large contribution in transforming medicine forever. It enabled complex operations, safer childbirth, transplants & critical care.
It has indeed spared billions of unimaginable suffering and restored dignity to healing. I would name it as one of medicine’s most profound humanitarian gifts.
A family physician in his sixties had his first colonoscopy with zero medication. No propofol, no Versed, no Demerol. He rated his own pain out loud, in real time, and here is the full map.
Insertion: a 2 or 3. Discomfort.
First flexure: spikes to 8, 9, 10. Tears in his eyes. Moaning.
Transverse: eases to 6 or 7.
Second flexure: 10 out of 10 again. Almost unbearable.
Each peak lasts about 30 seconds. The whole procedure runs 10 to 15 minutes. Then it is over.
He did it awake and watched the entire thing on the monitor. His blood pressure hit 150 over 90 in the waiting room from anticipation alone, and he meditated it back down. His reasons were not about safety. No needles, no IV, and he could drive himself home. He ran 31 Los Angeles marathons and took nothing after a hernia repair. He wanted the point of reference, to know exactly what he sends his patients into.
That last reason is the interesting one. Most of us recommend procedures we have never had. We describe the discomfort in general terms because general terms are all we have. He decided that was not good enough. Now he can tell a patient the exact shape of the hard part: two spikes, roughly 30 seconds each, at the two flexures, with real relief in between. That is not bravado. That is a clinician buying himself a level of specificity he could not get any other way.
And notice what the fear did to him anyway. He knows the science better than his patients ever will, and his pressure still climbed before anything started. Knowing the mechanism does not exempt you from the body's response to it. Worth remembering the next time any of us tells an anxious patient there is nothing to worry about.
Then the turn most people miss. He still calls the colonoscopy the gold standard. He still says get screened if your doctor recommends it. And he says most people should choose propofol or conscious sedation. The doctor who felt every second of his own thinks you should probably sleep through yours. He never once confuses his personal dare with his medical advice, and that discipline is the whole point.
Mike Neskovic, MD, 33 years in family medicine, on why he did it raw and why he does not recommend you copy him.
Listen to the full conversation on The Podcast by KevinMD. Link in the replies.
If you have had a colonoscopy, what made you choose sedation or skip it? #ThePodcastbyKevinMD
