UBC KGH Interventional Radiology #IRad
Kelowna, British Columbia
Joined March 2017
- Tweets1.8K
- Following594
- Followers769
- Likes12.5K
Nevin de Korompay retweeted
The first guidelines on preventive care in cirrhosis
At the #ACG2026 Postgraduate Course, Dr. @ebtapper will discuss the brand new ACG Clinical Guideline on Preventive Care for Patients with Cirrhosis! Will you be there?
📕Read the guideline: bit.ly/acg-preventive-cirrho…
🗣️Register for the PG Course: bit.ly/acg-2026-register
Nevin de Korompay retweeted
🧵OK, I'm doing a talk next week to a group that won't have any idea why I'm there, so I'm going to repost some of my older 🧵
This one is on the 'figure-of eight' and also the 'pursestring' sutures.
We'll go over what they're for and some technical pointers and pitfalls.
Nevin de Korompay retweeted
.@GBarnesMD updates that LMWH NOT IV UnFx heparin is now the class I recommended Rx for initial rx of acute #PE , even for pts undergoing intervention #VSS2026 . Spread the word!
Nevin de Korompay retweeted
Jobs that are strong "bundles" of tasks are less susceptible to AI replacement.
"Reading scans is a task. Radiology is a job. A radiologist does not only classify images. She triages cases, communicates with other physicians, trains residents, takes responsibility for difficult calls, and signs a diagnosis that others are willing to act on. What the market buys is not a classification exercise but a bundled service."
cepr.org/system/files/public…
Nevin de Korompay retweeted
Jensen Huang does highlight an important distinction between a task and a purpose. If your job is defined by a purpose (helping doctors diagnose disease using imaging) rather than simply a task (studying and describing scans), you will be able to leverage AI rather than be replaced by it.
I like Jensen Huang’s optimism about the future impact of AI on jobs but he keeps peddling this notion, this week on the Ezra Klein show, that “Every single radiology application has AI in it.” Radiologists are actually overworked today because AI has NOT made substantial productivity gains broadly.
nytimes.com/2026/09/23/opini…
Nevin de Korompay retweeted
In my opinion the problem in radiology is often not the uncertainty itself but the failure to clearly lay out the solution in the impression. If there is an evidence-based pathway to greater certainty through time or another imaging modality or ultimately tissue diagnosis that pathway should be clearly stated using a sensible scientific approach. The problem is the sloppy impression that simply says “likely this but cannot exclude that” and stops there. Uncertainty is unavoidable. The key is providing a rational path to resolve it.
Nevin de Korompay retweeted
As more radiologist work goes tele- and work-from-home, what does this mean for interpretation quality and downstream patient care?
A fascinating single-center study from South Korea just published in @AcadRadiol suggests that traditional onsite interpretations may be better: "Teleradiologists issued more follow-up examination recommendations than on-site radiologists for outpatient abdominopelvic CT examinations, with greater inter-radiologist variability." academicradiology.org/articl…
Nevin de Korompay retweeted
The MEMBRANE trial randomized patient with chronic SDH (with or without surgical evacuation) to MMA symbolization versus standard care.
The rate of recurrent SDH >10 mm or rescue surgery was basically half that in the MMA group compared to SOC
Middle Meningeal Artery Embolization With n-Butyl Cyanoacrylate in Chronic Subdural Hematoma
jamanetwork.com/journals/jam…
Nevin de Korompay retweeted
#throwbackthursday
Pregnancy and Uterine Artery Embolisation: Myth Busted
link.springer.com/article/10…
Nevin de Korompay retweeted
My beef with the term "chronic thrombus": it's misleading, and gives the wrong idea.
Posthrombotic changes should be better thought of as *fibrosis*: scar tissue occluding the veins! (much tougher than some soft clot).
Look at that intima: scar.
This impacts treatment!
Nevin de Korompay retweeted
First in the Human VECTOR PROCEDURE by Dr. Christopher Bruce during #SACIS2026 – at Riyadh, Kingdom of Saudi Arabia
Nevin de Korompay retweeted
Replace with performative whole body MRI
A call to abandon performative physical examinations. Join me?
Physical Examination for Asymptomatic Adults | New England Journal of Medicine nejm.org/doi/full/10.1056/NE…
Nevin de Korompay retweeted
#tb to 2024! 🔙
Safety and Efficacy of Percutaneous Morton Neuroma Cryoneurolysis Under Ultrasound Guidance
link.springer.com/article/10…
Nevin de Korompay retweeted
Superficial Temporal Artery Access with a 5 French Sheath for Coronary Angiography in a Patient With Leriche Syndrome, Subclavian Artery Occlusions, Multiple Acute Arterial Thromboses, and Heart Failure journals.sagepub.com/doi/ful…
When CSF disappears into the venous system, how do you find the fistula? 💧🧠
A comprehensive JNIS review of CSF–venous fistulas takes us from pathophysiology to localization and treatment:
🔎 Decubitus DSM & CTM
🩸 Percutaneous therapies
🧭 Transvenous embolization
🧠 Surgery & post-treatment care
A practical roadmap for a rapidly evolving field.
📖 Read the full review: bit.ly/4fV6QCP
#JNIS #CSFLeak #Neurointervention #SIH
Nevin de Korompay retweeted
Hyperdense capsule sign in chronic subdural hematoma is 1) a predictor of recurrence/progression under usual care (surgery or not), and 2) a predictor of benefit from adjunctive middle meningeal artery embolization.
pubs.rsna.org/doi/10.1148/ra…
Learning at #ENRS2026
Nevin de Korompay retweeted
Excellent work! Look at those ultrasound pics.
Case 163:with @BhaveshPopat7 @shubhamvanshi07 @DrAntariksh @DrVaibhavThakur
Thoracic Duct Embolization using direct percut retrograde TD cannulation at entry into IJV-LSCV with USG+fluoro guidance
YouTube:youtu.be/YOoL6vPvuGQ?si=_Y11…
#irad #medx @SIRRFS @SIR_ECS @SIRspecialists
Radial or femoral for stroke thrombectomy?
In >6,300 patients, TRA and TFA achieved similar recanalization and functional outcomes.
But the trade-off matters:
⬇️ Access-site complications with TRA: 3.4% vs 7.5%
⬆️ Access failure: 6.99% vs 2.18%
🔓 bit.ly/4z4Mdgr