Pediatric and Adult Neurosurgeon @gwsmhs @childrensnatl, founder @surgicaldsc, computer vision / machine learning to make surgery safer.
Washington, DC
- Tweets 2,546
- Following 622
- Followers 2,480
- Likes 3,419
Pinned Tweet
New Job in Surgical Data Science: please RT/share!
We're hiring a PhD computer vision / machine learning staff scientist in my NIH-funded lab @ChildrensNatl
Unique access to clinicians, clinical data, research infrastructure.
Send CV cover letter to email in the image below.
2
8
39
5,989
Daniel Donoho, MD retweeted
What if AI could actually understand what happens during surgery?
Not simply recognize an image - but identify surgical actions, track instruments, assess technical skill, recognize critical steps, and potentially predict whether a procedure will succeed.
That is the emerging world of surgical data science.
On this episode of @ProgressPotent1, I’m joined by Dr. Daniel Donoho, MD - @ddonoho - Neurosurgeon at @childrensnatl and @gwsmhs, Founder of the @surgicaldsc and Co-Founder of the Foundation for Digital Neurosurgery.
Dr. Donoho's work sits at a fascinating intersection of neurosurgery, computer vision, machine learning and surgical education.
We discuss:
🧠 How AI can decode what surgeons are doing from video;
🤖 Whether AI can objectively measure surgical skill;
🩸 How algorithms can predict hemorrhage outcomes from surgical video;
🎥 Why millions of hours of surgical footage could represent an enormous untapped dataset;
🔬 Creating a standardized “language” for surgical actions and gestures;
🏥 Whether AI could provide continuous feedback inside the operating room;
🌎 How surgical data could democratize access to expertise around the world;
🧠 And where all of this may ultimately lead as brain-computer interfaces and neurotechnology begin generating entirely new forms of human neurological data;
The big question:
Are we moving from recording surgery to actually teaching machines to understand it?
And if the brain itself becomes a data-generating interface…
Who owns that data?
#Neurosurgery #SurgicalAI #ArtificialIntelligence #SurgicalDataScience #DigitalSurgery #Neurotechnology #BCI #MedicalAI #FutureOfMedicine #ProgressPotentialAndPossibilities @ampragency
2
4
181
We've updated the most complete surgical intelligence leaderboard of current Vision Language Models (VLMs), benchmarking them on 10 datasets:
eval.surgicalvideo.io
All VLMs are significantly below the specialized model ceiling
More in our paper: arxiv.org/abs/2603.27341
2
5
1,425
something non-obvious about our surgery benchmark: it doesn't look like the math benchmarks. not all specialized skills (surgery vs math) are created equal.
you'd think that they'd be similar given both surgery and math require a lot of specialized knowledge possessed by a limited group of specially-trained people.
in the math benchmarks, Astra/Fable moggs and then everything else is just foundation models in descending order of released data/size. fine-tuned mini-models are non-players on frontiermath, etc. whereas here fine-tuned mini-models (and not-so-small but still fine-tuned models like LemonFM) somehow get to the top. these benchmarks were created by other independent teams so it's not like we're tipping the scales in our favor somehow (i promise we arent) besides, of course, fine-tuning which is the point, anyways.
the reason is that the entirety of math is "publishable" (the proofs, steps, problem) whereas here surgery is basically a surgeon acting on instinct/experience that is not entirely write-down-able from their med school/residency/fellowship training. in fact, you can go to two different hospitals and the procedure for the *same surgery* will look entirely different up to some high level goals and steps. plus, you have to be careful with medical data because of HIPAA etc which leads to overheads and limitations due to anonymizations etc.
so not all specialized skills are created equal. the ones that can be more or less written down in their entirety (math) will have lots of data and will be internalized by GPT/Claude pretty quickly. the ones that are based on "experience training" (like surgery) aren't so amenable to the current models. and, no, you can't just make the llm read a medical textbook and assume it generalizes (we tried that in the paper).
We've updated the most complete surgical intelligence leaderboard of current Vision Language Models (VLMs), benchmarking them on 10 datasets:
eval.surgicalvideo.io
All VLMs are significantly below the specialized model ceiling
More in our paper: arxiv.org/abs/2603.27341
2
5
1,225
Daniel Donoho, MD retweeted
🌟 SDSC founder @ddonoho recently joined @FRANCE24 to talk about our work with surgeons in Tanzania and the role AI can play in expanding surgical capacity.
One of Dr. Donoho's points gets right to the heart of why we’re there: Tanzania already has exceptional surgeons doing incredibly complex work, so how do we take the excellence that exists in a few hands and spread it as broadly as possible?
🖥️ Surgical video is part of the answer. Through our work with colleagues in Tanzania and Children's National Hospital, we’re looking at how AI can help surgeons learn from existing expertise, including in particularly challenging neurosurgical procedures.
🌍 Dr. Donoho also touches on something that matters enormously as surgical AI develops: local data matters. A model built on generalized surgical knowledge still needs to be tested with local hospitals, surgeons, procedures, and data to know whether it actually works in that setting.
Our work in Tanzania grew out of years of relationships, including Dan’s friendship with Dr. Hamisi Shabani and their efforts to bring together the country’s neurosurgical community. Those relationships ultimately helped lead to SDSC’s first program in Tanzania at Muhimbili Orthopaedic Institute.
🩺 The technology may be new. The goal is much more fundamental: help more surgeons learn from great surgeons, so more patients can access great surgical care.
📺 Watch the France 24 segment (timestamp at 8:55): france24.com/en/kenya-cracks…
#SurgicalAI #GlobalSurgery #Neurosurgery #GlobalHealthEquity
1
3
93
We benchmarked GPT-6 Astra on the @SurgicalDSC @CAAI_Booth Surgical AI Leaderboard
- astra is the new generalist leader, but falls behind tiny specialist models (1000x smaller)
Full Leaderboard: skblv.github.io/neurosurgery…
Paper: arxiv.org/abs/2603.27341
HT @kskblv @EricFithian
We benchmarked Claude Fable 5.1 and Gemini 3.8 Flash on surgical intelligence:
- The performance of the two new releases is spiky (e.g. good on tools, bad on VQA)
- Frontier LLMs still underperform tiny specialized models
Check out the full LLM leaderboard linked below
6
17
4,684
Daniel Donoho, MD retweeted
Work hard, play hard 💪 We host a department dinner during each visiting speaker's time on campus.
Excited to learn more about the research endeavors of Dr. Dan Donoho, a data driven surgeon scientist working hard to provide access across boarders!
#neurosurgery #globalhealth
1
15
1,920
Daniel Donoho, MD retweeted
Neurosurgery Grand Rounds (@NeurosurgeryOSU) today with Dan Donoho, M.D., (@ddonoho) of Children’s National Hospital (@ChildrensNatl) and George Washington University (@GWtweets).
He showed us how AI and computer vision are changing neurosurgery: teaching machines to understand what’s happening in the OR, flagging the moments that drive outcomes, and turning surgical video into a training tool for residents.
The goal isn’t smarter software. It’s safer operations and better surgeons.
@OSUWexMed @OhioStateMed
#AI #neurosurgery #surgery #training
7
20
1,404
Daniel Donoho, MD retweeted
1/ 🧠 Recognizing what's in a surgical video is one thing. Understanding what's happening is another.
👩🏽🏫 Before AI can understand surgery, we first have to teach it the language.
When SDSC trains a surgical tool detection model, thousands of human-verified annotations help teach the model what it's seeing. A scalpel needs to be a scalpel. A curette needs to be a curette. Actions, anatomy, phases, and eventually more complex surgical concepts all require clearly defined ground truth.
📢 We’ve all heard the adage "Garbage in, garbage out."
A mislabeled example can become a mistake the model learns. That's why human oversight isn't something surgical AI eventually graduates from. Humans define the ontology, verify the data, and ultimately determine whether a model's output has meaning in the real world.
Getting the labels right is only the foundation; just because a model recognizes a surgical instrument in a single frame does not mean it understands an operation.
🩺 Surgery unfolds over time. Actions combine into maneuvers. Maneuvers form tasks. Tasks become procedural steps. Together, they reveal the structure of an operation.
🤖 Teaching AI to represent those relationships is part of the transition SDSC describes as Digital Surgery 1.0 → Digital Surgery 2.0. Digital Surgery 1.0 gave us cameras, robotic systems, tracking tools, and other technologies capable of digitizing and sensing the operating room. Digital Surgery 2.0 asks what happens when we can begin to understand, measure, and learn from the surgical processes those technologies capture.
For #NeurosurgeryAwarenessMonth, that's the future we're working toward at SDSC: not AI that simply recognizes what's visible in neurosurgical video, but models that can help us understand how surgery unfolds and ultimately turn that understanding into objective, clinically meaningful surgical metrics. And it starts with something decidedly human: teaching the model what matters.
1
1
3
130
Physician, benchmark thyself!
[solve for the equilibrium]
Two years ago I took two hundred pairs of simulated patients and did something boring: I decided, one pair at a time, which of two patients would be seen first. I asked state-of-the-art LLMs (@google @openai @AnthropicAI) to do the same with the same patients. I just repeated the experiment.
2
337
Fable 5 and GPT-5.6-Sol vs finetuned YOLO on surgical benchmarks. Takeaway here is that the tiny finetuned YOLOv2 (with ~26M params) still beats the SOTA out-off-the-shelf models. Specialized training is still needed for specialized tasked.
Credit: @kskblv
We benchmarked Fable on surgical tool detection so you don’t have to. At almost 2 cents/image, Claude Fable 5 still loses to a compact image classifier. Will frontier VLMs ever catch up in surgical perception? Read our paper to learn more: arxiv.org/abs/2603.27341
1
3
4
885
Congratulations to the R Project on receiving the $1M Rousseeuw Prize for Statistics 2026! @RProgramming @RConsortium @Rbloggers
We now take for granted computational science requires a strong ecosystem of code sharing. However, before uv, conda, and pip, it was R that laid the foundations for the dissemination of statistical/computational methodologies not only to statisticians but also to biologists, social scientists, and other disciplines driven by data analysis. We now rest under the shade of large trees, but it is gratifying to see the Rousseeuw Prize go to the R Project which was one of the first plant the seeds and till the soil --- and kept it fertiles for almost three decades.
prweb.com/releases/2026-rous…
5
12
715
For Surgical AI, Claude Mythos/Fable 5 isn't all that scary. Loses to a YOLO. See updated benchmarks from @kskblv 🏥🤖
Real alpha is in specialized pipelines and specialized data (arxiv.org/abs/2603.27341) not bigger foundation models.
Collab with @SurgicalDSC & @CAAI_Booth
We benchmarked Fable on surgical tool detection so you don’t have to. At almost 2 cents/image, Claude Fable 5 still loses to a compact image classifier. Will frontier VLMs ever catch up in surgical perception? Read our paper to learn more: arxiv.org/abs/2603.27341
1
4
5
1,582
I was humbled to be invited to give a lecture for and start building together with you.
if we commit to unlocking data to serve our patients and global community, the next twenty years of our field will be more powerful than we can imagine.
This is an intense work of partnership - with industry, with hospitals, with patients and advocates like @HydroAssoc , and with our global surgical community.
As we reflect on an incredible month following the AANS Annual Scientific Meeting 2026, we simply want to say thank you. 🧠✨ Thank you to every member, student, resident, physician, speaker, sponsor, volunteer, partner, and supporter who helped make this year’s events hosted by the American Society of Black Neurosurgeons so impactful.
The momentum from AANS 2026 continues — and we’re just getting started. 💫
#ASBNOrg #ASBNEvents #ASBNConnect #AANS2026 #Neurosurgery
1
2
15
956
Daniel Donoho, MD retweeted
The future of neurosurgery is here—and the conversation was powerful. 🧠✨ During #AANS2026, ASBN and AANS brought together an incredible panel of experts for The Translational Future of AI in Neurosurgery, where over 50 attendees gathered to explore how artificial intelligence is transforming the field of medicine and neurosurgery.
Moderated by Dr. Sonia Eden @soncapone and Dr. Nnenna Mbabuike @NSMbabuike_MD, the discussion featured insights from:
▪️ Dr. Joseph T. Francis
▪️ Dr. Dan Orringer
▪️ Dr. Richard Byrne
▪️ Dr. Daniel Donoho @ddonoho
▪️ Dr. Samuel Browd
Panelists shared perspectives on how AI is helping close educational and healthcare access gaps by expanding the ability to share knowledge and resources globally, improving surgical precision, enhancing decision-making, and shaping the future of patient outcomes. The conversation also highlighted important ethical considerations, including the risks AI expansion may pose to low-income countries where access to technological infrastructure and resources remains limited.
Thank you to everyone who joined us for this impactful discussion and to our partners at AANS for creating space for innovation, collaboration, and forward-thinking conversations in neurosurgery.
#ASBNOrg #AANS2026 #Neurosurgery #ArtificialIntelligence #AIInMedicine #FutureOfMedicine #HealthEquity #ASBNConnects #ASBNEvents
@seattlechildrens @gwsmhs @childrensnational @mayoclinic @nyulangone @semmesmurpheyclinic
1
4
13
1,192
Daniel Donoho, MD retweeted
@ddonoho discussing a potential framework for improving trainee skill assessment in neurosurgery. @GWSMHS @ChildrensNatl @SNS_Neurosurg
ALT Professional man presenting on a stage behind a podium
1
1
226
Daniel Donoho, MD retweeted
📢 AI Stewardship Principle #1: Verification is Everything
If AI is the intern, then verification is your job.
🕵🏻♀️ As Dr. @dhirajpangal shared in our recent Data Science Roundtable, every output from AI should be treated the same way you’d treat work from a junior trainee: Review it. Fact-check it. Hold it to high standards.
Whether it’s drafting text, generating code, or analyzing data, nothing should go out into the world unchecked. Because ultimately, the output reflects you.
🩺 In high-stakes fields like surgery, that standard IS NOT optional.
At SDSC, this principle is foundational. Building reliable surgical AI means more than generating outputs, it requires rigorous validation, continuous oversight, and alignment with clinical reality.
🌟 Speed is useful. Accuracy, precision and rigor are essential.
👀 Catch the full session here: surgicalvideo.io/blog/dr-dhi…
👩🏻💻 Watch previous DSRTS: tinyurl.com/sa7845s6
Our Data Science Roundtables (DSRTS) are one of the ways we stay anchored to real clinical needs. They bring together surgeons, researchers, and technologists to identify the questions that matter most, and to ensure the tools we build are driven by practice, not theory. As we plan upcoming sessions, we’d love your input on speakers and topics. If there’s a clinician, researcher, or innovator you’d like to hear from, please share your suggestions via this google form: tinyurl.com/hrdhsc6z
#SurgicalAI #AIinHealthcare #AIStewardship
1
1
193
Check this out next Tuesday 4/21
🚀 Next up at the SDSC Data Science Roundtable! Dr. Jacob Young on “Operating Room Video, Artificial Intelligence and Surgical Training”
Dr. Jacob Young is a neurosurgeon who cares for patients with brain tumors, including low- and high-grade gliomas, meningiomas and brain metastases. He specializes in brain mapping techniques that identify areas important for motor, language and sensory functions, enabling surgeons to minimize the impact on these areas during surgery to remove a tumor. His clinical research explores how to optimize functional outcomes and minimize complications after surgery for intrinsic brain tumors.
Dr. Young's translational research interests include first-in-human early-phase clinical trials of new immunotherapies and innovative ways to deliver drugs to treat cancerous brain tumors. In his lab, he focuses on understanding the microenvironment of glioma and how it contributes to intratumoral heterogeneity (differences between cancer cells within the same tumor), treatment resistance and cancer evolution at the time of progression. Working with numerous collaborators, his ultimate goal is to design treatment strategies that improve outcomes for patients with brain tumors. His glioblastoma research has been recognized at the institutional, regional, and national levels. During his postdoctoral training, he was awarded the Chan Zuckerberg Biohub Physician-Scientist fellowship, the Andrew J. Lockhart Focused Ultrasound and Immuno-Oncology fellowship, and the American Society of Clinical Oncology Young Investigator Award.
Dr. Young earned his medical degree from the University of Chicago Pritzker School of Medicine, where he was elected to the Alpha Omega Alpha Honor Medical Society. He completed a residency in neurosurgery and a postdoctoral fellowship in tumor immunology at UCSF, where he received an Exceptional Physician Award during his training. He was recognized with the Krevan's Award for excellence in patient care at the Zuckerberg San Francisco General Hospital and the Howard Naffziger Award for Outstanding Clinical Service from the department of neurological surgery.
🗓️ Tuesday, April 21, 2026
⏰ 2:00pm–3:00 pm Eastern
🔗 Register: tinyurl.com/4tuy57p7
#DataScienceRoundtable #AIinSurgery #SurgicalTraining
2
287
Daniel Donoho, MD retweeted
@FMoHealth A national partnership. A new model for surgical training. And a clear signal of where surgical AI is headed.
SDSC has partnered with Ethiopia’s @FMoHealth to integrate AI-driven surgical video analysis into OB/GYN training, working toward a future where every woman can access WHO-standard surgical care within 100 km of her community.
By combining global surgical video data with Ethiopia’s clinical leadership, this initiative aims to accelerate training, improve quality, and expand access to minimally invasive care, especially in underserved regions.
It’s also a powerful example of what responsible AI in healthcare can look like: government-led, locally governed, and built on real-world data.
📖 Read more: surgicalvideo.io/blog/a-nati…
Progress like this depends on data. If you’re a surgeon, researcher, or institution with surgical video, we’d love to collaborate. Contributing data helps train better models, improve global benchmarks, and ultimately expand access to safer surgery worldwide.
👉 Learn more and get involved: surgicalvideo.io/research
#SurgicalAI #GlobalHealth #WomensHealth #SurgicalEquity
2
3
152