Shehab Anwer, MD

MD Cardiology 🫀 Founder @TheAdimension

A wayfarer, across #heart, #art & #technology:

[Shehab Anwer, MD @ShehabAnwer]

Moving #AI in #CVImaging from opaque masks towards outputs that serve clinical, research or #AI workflows🫀Open Source = Open Invitation for Collaboration @ https://t.co/toXz7Ajubo

#EchoFirst

[Shehab Anwer, MD @ShehabAnwer]

Overview → https://t.co/3iYH7Hz8nO
Results → https://t.co/2YctJQ0Ajj → https://t.co/Bh1Ovb8IOU
Adapters → https://t.co/4VGnbMmcbW
ATRIA-EchoTrace → https://t.co/wkEoVahYa8
LV only. 200/200 parsed. Median 4.98 mm. Worst published.

[TheAdimension @TheAdimension]

Stop asking the model for a picture of the border. Ask it for the coordinates 🫀

In this video, founder Dr. Shehab Anwer (@ShehabAnwer) introduces ATRIA-EchoTrace:
MedGemma 1.5, QLoRA-adapted so the LV endocardial border comes back as a JSON polygon.

Drag a vertex > Metrics update > The corrected points re-enter training in the same format!

Built for the workflows people actually run: clinical review and AI engineering — not a heatmap you can only accept or bin.

Research use only. Not a medical device. Every contour is a proposal.

Would you rather receive a mask, a polygon, or a spreadsheet?

#EchoFirst #AIinHealthcare

[Shehab Anwer, MD @ShehabAnwer]

Opaque mask: model decides🫀Structured Polygon: you decide!

I recorded the workstation on a held-out CAMUS frame: best / median / worst of the 200 frames - are all public, including the 25 failures on our website!

Move a point. The metric moves. The correction is already in the corpus format.

Research use only. Not a device.

Would you ship the proposal, or only the revised trace?

[EHJ-IMP Editor-in-Chief @EHJIMPEiC]

🫀 Can serial ultrasound congestion scores predict outcomes better than a single assessment in acute heart failure?
A new prospective study introduces ΔVExPLUs, combining serial changes in Venous Excess Ultrasound (VExUS) and lung ultrasound (LUS) to improve risk stratification in patients hospitalized with acute decompensated heart failure (ADHF).
🔍 Key findings:
✅ 104 patients with ADHF underwent serial assessment of VExUS and lung ultrasound during hospitalization to evaluate changes in systemic and pulmonary congestion.
✅ The combined change (ΔVExPLUs) provided stronger prognostic information than either VExUS or lung ultrasound alone, highlighting the value of integrating venous and pulmonary congestion assessment.
✅ Patients with persistent or worsening ultrasound congestion experienced significantly worse clinical outcomes, emphasizing that successful decongestion should be assessed with imaging rather than symptoms alone.
✅ The study supports multiorgan ultrasound as a practical bedside tool to guide treatment response and identify patients at higher risk before discharge.
💡 Clinical takeaway: Heart failure congestion extends beyond the lungs. Integrating VExUS and lung ultrasound may provide a more complete picture of decongestion and improve risk stratification, paving the way for more personalized management of ADHF.
📖 Read the full article: https://t.co/eAUK9cZ2sa
#CardioX #HeartFailure #POCUS #LungUltrasound #VExUS #Echocardiography #CardiacImaging #AcuteHeartFailure #Cardiology #MedTwitter

[NEJM AI @NEJM_AI]

The promise of medical AI is not simply better performance — it is better care grounded in evidence that can be trusted. On AI Grand Rounds, @DrXiaoLiu discusses what rigorous evaluation looks like as technologies move from publications into practice. 🎧 https://t.co/cHzTla7ZOh

[NVIDIA Healthcare @NVIDIAHealth]

⚡️ Healthcare AI moves faster when founders can pair breakthrough research with accelerated computing and production-ready infrastructure.

Join us alongside Google Cloud, Google Research, and leading healthcare VCs for a five-part virtual series built for digital health and biotech founders.

Check it out 👇

[NVIDIA Healthcare @NVIDIAHealth]

Train medical robots where anatomy, physics, perception and policy meet.

NVIDIA Isaac for Healthcare introduces GPU-native medical physics simulation with:
🫀 Simulation-ready anatomical digital twins
⚙️ Device–tissue interaction modeling
🩻 X-ray, ultrasound and camera data paired with exact ground truth
⚡ Thousands of parallel simulation environments
🤖 Policy training and evaluation before a real robot moves

Explore the framework for surgical robotics, endoscopy, endovascular intervention and robotic ultrasound.

👇 https://t.co/7VfbeCfKGA

[Shane Legg @ShaneLegg]

My journey to develop AGI spans 25 yrs, including 10+ yrs thinking about technical & societal perspectives at Google DeepMind.
AGI is on the horizon - we need deeper understanding of its implications. To help, we've created the DeepMind Institute. https://t.co/dpc2y4IGT1

[TheAdimension @TheAdimension]

@ShehabAnwer If you want the worst frame (18.59 mm), it is in the gallery on purpose 💪🏻💪🏽💪🏿

[Morgan Cheatham, MD @morgancheatham]

in the first chapter of healthcare AI, the battleground was access to longitudinal clinical data.

in this next phase, it will be access to biological specimens: tissue, blood, synovial fluid, and other samples containing 1,000× more information, linked to clinical context and longitudinal outcomes.

the winners will be those who can access, assay, and interpret these specimens at the lowest cost and highest resolution.

[Shehab Anwer, MD @ShehabAnwer]

@morgancheatham Excellent insight, @morgancheatham: Longitudinal data spans to the seconds it takes the cardiac tissue to work - I'd love to hear your thoughts about:
https://t.co/XfdzOtMvtF

[Shehab Anwer, MD @ShehabAnwer]

Opaque mask: model decides🫀Structured Polygon: you decide!

I recorded the workstation on a held-out CAMUS frame: best / median / worst of the 200 frames - are all public, including the 25 failures on our website!

Move a point. The metric moves. The correction is already in the corpus format.

Research use only. Not a device.

Would you ship the proposal, or only the revised trace?

[Shehab Anwer, MD @ShehabAnwer]

@pridgenmd Amazing work, @pridgenmd ! What if a foundation model is tuned to output coordinates for actionable segmentation? Check this out :)
https://t.co/XfdzOtMvtF

[Shehab Anwer, MD @ShehabAnwer]

Opaque mask: model decides🫀Structured Polygon: you decide!

I recorded the workstation on a held-out CAMUS frame: best / median / worst of the 200 frames - are all public, including the 25 failures on our website!

Move a point. The metric moves. The correction is already in the corpus format.

Research use only. Not a device.

Would you ship the proposal, or only the revised trace?

[Shehab Anwer, MD @ShehabAnwer]

Overview → https://t.co/3iYH7Hz8nO
Results → https://t.co/2YctJQ0Ajj → https://t.co/Bh1Ovb8IOU
Adapters → https://t.co/4VGnbMmcbW
ATRIA-EchoTrace → https://t.co/wkEoVahYa8
LV only. 200/200 parsed. Median 4.98 mm. Worst published.

[Shehab Anwer, MD @ShehabAnwer]

Opaque mask: model decides🫀Structured Polygon: you decide!

I recorded the workstation on a held-out CAMUS frame: best / median / worst of the 200 frames - are all public, including the 25 failures on our website!

Move a point. The metric moves. The correction is already in the corpus format.

Research use only. Not a device.

Would you ship the proposal, or only the revised trace?

[TheAdimension @TheAdimension]

Stop asking the model for a picture of the border. Ask it for the coordinates 🫀

In this video, founder Dr. Shehab Anwer (@ShehabAnwer) introduces ATRIA-EchoTrace:
MedGemma 1.5, QLoRA-adapted so the LV endocardial border comes back as a JSON polygon.

Drag a vertex > Metrics update > The corrected points re-enter training in the same format!

Built for the workflows people actually run: clinical review and AI engineering — not a heatmap you can only accept or bin.

Research use only. Not a medical device. Every contour is a proposal.

Would you rather receive a mask, a polygon, or a spreadsheet?

#EchoFirst #AIinHealthcare

[Shehab Anwer, MD @ShehabAnwer]

The number I care about in this clip is not Dice!

It is that a disagreement becomes the next training row with no conversion step!

If you open the workstation, drag one vertex and watch the JSON + millimetres update. That is the whole thesis that bridges human ingenuity & machine intelligence!

[TheAdimension @TheAdimension]

Results → https://t.co/ygbcJ0LIvX

Code → https://t.co/eTFRkIJqT9

Adapters → https://t.co/xxDHqZbwlY

Technical note → https://t.co/9ZUXKirbBB

CAMUS held-out: 200/200 valid polygons - median point-to-curve 4.98 mm

Built on our foundational framework, DEITY principles: Data, Ethics, Informatics, Technology, & You!

[Fei-Fei Li @drfeifei]

Incredible opportunity for robotic learning researchers/engineers to join @theworldlabs! ❤️‍🔥

[Google for Health @GoogleForHealth]

Published in Nature Medicine: @Google researchers and collaborators share lessons from prospective studies of AMIE, highlighting why real-world clinical evidence is critical to evaluating conversational AI in health.

Read more: https://t.co/vAKMGHioXM

[Shehab Anwer, MD @ShehabAnwer]

A cardiac imager cannot argue with a mask!

ATRIA-EchoTrace by The Adimension finetunes MedGemma 1.5 so an echo frame returns editable polygons, translating black box into interoperability!

Drag a vertex. Estimate Parameters. Integrate to your workflow: Still research work!

Built on open weight model & datasets. Open invitation to collaborate.

#EchoTrace #MedGemma