🧠 🪄 Neurologist, Neurointensivist. Professor, Educator, & Researcher. Digital Health/Space/AI-ML advocate. NSICU & STR-X founder. Author. Tweets mine. 🚫 DM

US
The 18th Annual Stroke & Cerebrovascular Disease Review is coming to The Ritz-Carlton Amelia Island this October 14-17. This year’s curriculum map: 🧠 Large vessel occlusion dynamics — from MeVO/M2 to basilar 💊 ICH reversal timing — when the 4-hour clock matters 🔬 CREST-2 keynote — the investigators, in person @Demaerschalk @ThomasBrott ⚡ Status epilepticus /stroke mimics— beyond first-line, Rapid EEG 🫀 Occult AF detection & management 🏥 Prehospital stroke systems of care 🔄 Minimally invasive ICH surgery — ARCH trial update 🧬 Cerebral amyloid angiopathy — emerging therapeutics 🏥 Stroke Economics and Social Determinants of Health (SDOH) Jointly accredited: CME / Pharmacy CE / Nursing CE. Built for the whole team — neurologists, neurosurgeons, APPs, nurses, pharmacists. Registration details below. 👇🏽 Save the dates: October 14-17, 2026. ce.mayo.edu/stroke2026 @tamrar1 @RabihTawkMD @wchrisfox @ThienHuynh15 #StrokeAmeliaIsland #MayoCPD #Stroke #Neurology #CME #NeuroCriticalCare
2
1
15
860
👏 @Google @STRXLabs @gisellecoelhomd Expand healthcare triage at scale
Good to see MedGemma, a unified, open, transparent foundation model by @Google published @NatureMedicine today, with diverse medical image-language capabilities (Figure), can run on a smartphone, weights @huggingface nature.com/articles/s41591-0…
146
🧠 Less than 3 weeks to the 18th Annual Stroke & Cerebrovascular Disease Review at Amelia Island, Oct 14–17 | Live + Livestream Lots of stroke and Cerebrovascular topics and trials to review ! A few highlights: → Dr George Harston: Can AI Improve Health Equity in Stroke Care at Global Scale? → ARCH Trial update @berejah → MIND Trial results @babaksjahromi → Hands-on skills fair In person or livestream. Register: ce.mayo.edu/neurology-and-ne…
2
4
238
W David Freeman, MD, FRSM, FNCS, FAAN retweeted
Today we have set out how we’re building AI to accelerate science and improve people’s lives. Just some examples in the last week or so: - Mapped all 9B possible single letter genetic changes across the human genome with AlphaGenome Atlas and made it openly available to researchers. - Billions of decisions depend on weather predictions so we introduced WeatherNext 3, our most accurate and capable global weather AI model to date. - We published AI & Economy ATLAS, a comprehensive open-access look at how people are using AI globally. - AI has enabled extraordinary advances in language translation. Today our services are available in nearly 300 languages, spoken by 7B people We’re focusing our efforts on four key areas: health, natural disaster and weather resilience, learning, and economic opportunity. blog.google/innovation-and-a…
440
691
6,313
804,700
W David Freeman, MD, FRSM, FNCS, FAAN retweeted
Mayo’s @gisellecoelhomd and @freemanwd are exploring how volumetric capture and mixed reality can create immersive surgical simulations and expand training opportunities for clinicians around the world. Read more: education.mayo.edu/schulze-a… pubmed.ncbi.nlm.nih.gov/4239…
1
5
9
791
Nimodipine in the catheter flush cuts periprocedural vasospasm by more than half
 German Stroke Registry–ET, 25 centers, n = 16,583 — the largest look yet at a habit that varies center to center. Nimodipine-containing flush used in 7,548 (45.5%). •Vasospasm: IPTW-adjusted 1.8% vs 4.8%; OR 0.38 (95% CI 0.30–0.47), P<0.001 •90-day mRS 0–1: OR 1.23 (1.13–1.34); mRS 0–2: OR 1.16 (1.08–1.25), both P<0.001 Caveat: registry, not randomized — flush composition correlates with center culture, and IPTW can't fix unmeasured operator skill. But the cost of adopting it is low. Reference: 📄 J NeuroInterventional Surg, 2026 Sep 11 · DOI: 10.1136/jnis-2026-025972 · PMID 42728086 pubmed.ncbi.nlm.nih.gov/4272…
8
581
🚨 IV Loberamisal for Acute Ischemic Stroke Neuroprotection within 48hrs 🧠 The Phase 3 LAIS Trial was just published, evaluating intravenous loberamisal for patients with acute ischemic stroke treated within 48 hours of symptom onset. Key findings at 90 days: ✅ 69.7% of the loberamisal group achieved full functional recovery (mRS 0-1) compared to 56.3% on placebo. ✅ The treatment was safe, with no increase in serious adverse events or mortality. This novel drug dually targets the PSD-95 pathway and the α2-GABAA receptor, offering a promising neuroprotective strategy to complement standard stroke care. Reference: Li S, Feng B, He D, et al. Loberamisal for Acute Ischemic Stroke: The LAIS Randomized Clinical Trial. JAMA. Published online September 10, 2026. doi:10.1001/jama.2026.16557. doi.org/10.1001/jama.2026.16…
2
10
35
2,054
💬 Editorial Insight: A Tantalizing Prospect in Stroke Care 🧠 Dr. Jeffrey Saver’s JAMA editorial contextualizes the LAIS trial results, noting that while reperfusion therapies are reaching their limits, neuroprotective strategies may finally be yielding results. Key perspectives from the editorial: 👉 Dual-action potential: Loberamisal combines neuroprotective and neuroreparative mechanisms across prereperfusion, postreperfusion, and recovery phases. 👉 Practical window: The 48-hour treatment window accommodates the large proportion of stroke patients who arrive too late for reperfusion therapy. 👉 Scientific caution: The trial defies historical paradigms with a treatment effect size comparable to early thrombolysis, calling for thoughtful evaluation. 👉 Call for replication: Because extraordinary claims require extraordinary evidence, independent multinational trials and participant-level data sharing are critical next steps. Reference: Saver JL. Neuroprotective Therapy for Acute Ischemic Stroke: A Tantalizing Prospect. JAMA. Published online September 10, 2026. doi:10.1001/jama.2026.17400 doi.org/10.1001/jama.2026.17…
1
1
129
🧠 Is automated CSF exchange poised to replace standard passive drainage in neurocritical care? A meta-analysis synthesizing data from 834 patients across 7 studies compares Active Cerebrospinal Fluid (ACE) exchange—including neuroendoscopic lavage and IRRAflow systems—against conventional External Ventricular Drainage (EVD) in intraventricular hemorrhage (IVH) and subarachnoid hemorrhage (SAH). Key Findings: 🩸 2.7x Higher Hematoma Clearance: ACE significantly accelerated blood clot resolution compared to EVD (OR 2.72 [95% CI, 1.22–6.08], p=0.015). 💪 66% Better Functional Recovery: Patients treated with ACE achieved higher odds of favorable neurological outcomes (mRS score <= 3) (OR 1.66 [95% CI, 1.12–2.48], p=0.012). 🛡️ 78% Lower Infection Rate: Continuous active irrigation and drainage markedly decreased post-procedural infections (OR 0.22 [95% CI, 0.12–0.42], p<0.001). 🔌 78% Lower Shunt Dependency: Rapid blood evacuation significantly reduced chronic hydrocephalus and permanent shunt placement (OR 0.22 [95% CI, 0.10–0.50], p<0.001). ⚖️ Comparable Safety Profile: No statistically significant differences were observed between ACE and EVD regarding mortality (p=0.716), rebleeding risk (p=0.257), or length of ICU stay (p=0.447). Subgroup Highlights: • Neuroendoscopic lavage demonstrated the highest efficacy for hematoma clearance (OR 3.70, p=0.010). • IRRAflow systems provided superior infection reduction (OR 0.24, p<0.001) [6]. • Younger patients (<60 years) derived the greatest benefits in clot clearance and functional recovery. 💡 Bottom Line: Active CSF exchange represents a promising advance over passive EVD, accelerating clot evacuation while reducing infection and permanent shunt dependency. What are your thoughts on adopting automated active irrigation in the neuro-ICU? Let us know your perspective below! 👇 Reference: Bahadori AR, Röblom A, Niemelä M, et al. Active Cerebrospinal Fluid Exchange Versus External Ventricular Drainage in IVH and SAH: A Meta-Analysis. Stroke Vasc Interv Neurol. 2026. ahajournals.org/doi/suppl/10… @berejah @babaksjahromi @RabihTawkMD
8
25
1,334
“Time is Brain” for SAH 🩸 too! Does the aSAH clock start at the door, or at the first headache? A 7-center cohort from China says the prehospital window is the outcome lever we keep undercounting. 726 adults with aneurysmal SAH. Rapid transfer (to an endovascular-capable hospital in 8 hours or less) vs delayed (after 8 hours): 🚑 386 made it in 8 hours (53%). 340 did not. 🧠 6-month mRS >2: 8% vs 23% (adjusted OR 5.3) 💔 12-month death: 4.9% vs 8.8% (adjusted OR 2.6) ⚠️ Delayed transfer also tracked with more herniation, hydrocephalus, ischemic stroke, and rebleeding Only about half of this network hit the 8-hour window. If your region still treats aSAH transfer as get-there-sometime-today, this is the number to take to the transfer committee. What cutoff does your system actually hit? Drop your comments below 👇🏽 Reference: Mi S, Xiao ZK, Wang B, Duan Y, Liu A. The impact of rapid transfer on the prognosis of aneurysmal subarachnoid hemorrhage: a multicenter cohort study. World Neurosurg. 2026. doi:10.1016/j.wneu.2026.125269 doi.org/10.1016/j.wneu.2026.…
7
29
3,840
🚨 Why does blood pressure spike during an ischemic stroke, and what is the optimal BP during thrombectomy? 🧠 When a large blood vessel is blocked, the nervous system triggers heightened sympathetic activation to preserve blood flow to the brain. Clinically, this protective spike is known as the acute hypertensive response, acting as a variant of Cushing's reflex. Since this elevated BP is a natural mechanism to push blood past the clot, should we maintain this high, individualized baseline during mechanical thrombectomy surgery? 🏥 The DETERMINE trial [1] tested this exact hypothesis by comparing two management strategies: 🔹 Individualized BP: Keeping mean arterial pressure within 10% of the patient's elevated, natural baseline. 🔹 Standard BP: Targeting a flat systolic BP of 140-180 mmHg. The Results: ❌ No clinical benefit. A favorable 90-day functional outcome was seen in 44.2% of the individualized group vs 48.8% of the standard group . ⚠️ Safety, including mortality and symptomatic brain bleeds, was virtually identical. Why did the physiologic approach fail? Because managing the body's acute hemodynamic instability is incredibly difficult. Patients in the individualized group only spent a median of 58.1% of the procedure in their precise target zone. The Takeaway: While supporting the brain's acute hypertensive response makes physiologic sense [2], trying to maintain those exact individualized targets during surgery is too complex right now. Standard BP targets remain the most practical strategy until we develop real-time, dynamic autoregulation monitoring. Read the full study and physiology review below: 👇🏽 1. Mäıer B, Gory B, Chabanne R, et al. Individualized vs Standard Blood Pressure Control During Thrombectomy for Ischemic Stroke: The DETERMINE Randomized Clinical Trial. Neurology. 2026;107:e218420. doi:10.1212/WNL.0000000000218420 doi.org/10.1212/WNL.00000000… 2. Qureshi A. Acute Hypertensive Response in Patients With Stroke: Pathophysiology and Management. Circulation. ahajournals.org/doi/10.1161/…
1
6
18
981
The 18th Annual Stroke & Cerebrovascular Disease Review is coming to The Ritz-Carlton Amelia Island this October 14-17. This year’s curriculum map: 🧠 Large vessel occlusion dynamics — from MeVO/M2 to basilar 💊 ICH reversal timing — when the 4-hour clock matters 🔬 CREST-2 keynote — the investigators, in person @Demaerschalk @ThomasBrott ⚡ Status epilepticus /stroke mimics— beyond first-line, Rapid EEG 🫀 Occult AF detection & management 🏥 Prehospital stroke systems of care 🔄 Minimally invasive ICH surgery — ARCH trial update 🧬 Cerebral amyloid angiopathy — emerging therapeutics 🏥 Stroke Economics and Social Determinants of Health (SDOH) Jointly accredited: CME / Pharmacy CE / Nursing CE. Built for the whole team — neurologists, neurosurgeons, APPs, nurses, pharmacists. Registration details below. 👇🏽 Save the dates: October 14-17, 2026. ce.mayo.edu/stroke2026 @tamrar1 @RabihTawkMD @wchrisfox @ThienHuynh15 #StrokeAmeliaIsland #MayoCPD #Stroke #Neurology #CME #NeuroCriticalCare
2
1
15
860
As one of the course directors, I look forward to cutting-edge cerebrovascular CME at Mayo Clinic's 18th Annual Stroke & Cerebrovascular Disease Review — Amelia Island, Oct 15–17. Four trials from the last 30 days we'll be arguing about: DETERMINE (n=433) — individualized BP during thrombectomy did NOT improve 90-day outcomes doi.org/10.1212/WNL.00000000… BRIDGE-TNK post hoc (n=550) — bridging tenecteplase signal favors ASPECTS <8, not 8–10 doi.org/10.1212/WNL.00000000… 2026 AHA/ASA Rehab & Recovery Guideline — first update since 2016 doi.org/10.1161/STR.00000000… DWI+ lesions after ICH (n=342) — independent predictor of poor 6-month outcome doi.org/10.1212/WNL.00000000… 31.5 AMA PRA Category 1 Credits. In-person or livestream. Ritz-Carlton room block closes Sept 22. ce.mayo.edu/stroke2026 @tamrar1 @RabihTawkMD @KevinMBarrettMD #StrokeAmeliaIsland #MayoCPD #Stroke #Neurology #CME
55
Calling all Neurocritical Care, Neurology, and Neurosurgery physicians! 🧠 Join national experts to master the latest evidence-based medical and surgical interventions for Intracerebral Hemorrhage (ICH) management. This multidisciplinary symposium features practical strategies and hands-on simulation training for endoscopic hematoma evacuation using 3D models. 🔬 📅 Date: Sunday, September 20, 2026 [1] 📍 Location: Northwestern NCASE in Chicago OR Livestream 🎓 Credits: Earn up to 6.25 AMA PRA Category 1 Credits Enhance your clinical decision-making and secure your spot today! 👇 @neurocritical @babaksjahromi @berejah @RabihTawkMD northwestern.cloud-cme.com/I…
2
4
11
975
How do we ensure ethical, family-centered care during organ donation in the neuro-ICU? 🧠🤍. The Neurocritical Care Society (NCS) has issued a new Ethical Guidance Statement addressing the complex intersection of end-of-life care and organ donation. Key takeaways include: 🤝 Enhancing collaborative communication between treatment teams and Organ Procurement Organizations (OPOs). ⚖️ Navigating the ethics of first-person authorization and addressing surrogate concerns. 🕊️ Prioritizing high-quality palliative care and patient dignity during the withdrawal of life-sustaining therapies. This is an essential framework for critical care professionals to maintain clear boundaries, build trust, and uphold patient autonomy while addressing the global organ shortage. Read the full guidelines below! 👇 Reference: Reynolds AS, Jaffa MN, McHugh D, et al. Education and Ethical Guidance for Organ Donation at the End-of-Life: An Ethical Guidance Statement from the Neurocritical Care Society. Neurocrit Care. 2026. doi.org/10.1007/s12028-026-0…
1
6
394
🧠 A massive breakthrough in clinical AI and neurology is here! Introducing MORGOTH, the first unified foundation model designed for comprehensive clinical EEG interpretation across all major medical settings. EEG interpretation is a scarce resource, often leading to preventable overdiagnosis or misdiagnosis of epilepsy. Existing AI models only target single tasks or settings, but MORGOTH changes everything by providing an all-in-one system. Key Highlights: 1️⃣ Expert-Level Accuracy: Achieved AUC-ROC scores between 0.86 and 0.98 across 17 different clinical EEG findings. 2️⃣ Human-Grade Performance: Outperformed at least 90% of human experts on three multi-expert-annotated datasets and surpassed at least 20% on all 17 tasks. 3️⃣ Massive Validation: Rigorously evaluated using 12 datasets containing 34,602 patients across 52 hospitals in 8 countries. 4️⃣ Robust and Reliable: Maintains stable performance across different age groups, sexes, and even moderate EEG channel loss. 5️⃣ Clinically Friendly: Automatically handles raw EEG preprocessing and channel selection without manual intervention. This foundation model is designed to augment clinical workflows, boosting diagnostic access in resource-limited areas and enhancing efficiency in high-volume medical centers. Reference: Sun C, Karakis I, Herlopian A, Ng MC, Osman G, Sheikh Z, et al. Toward unified and comprehensive automated electroencephalogram interpretation: a multicentre development and validation of an electroencephalogram foundation model. Lancet Digit Health. Published online 2026. doi.org/10.1016/j.landig.202… Data sharing: All data needed to reproduce the results is available at bdsp.io/content/morgoth1/1.0… Code is available at github.com/bdsp-core/morgoth
1
1
10
713
🧠 PREDICTORS OF SHUNT DEPENDENCY AFTER ANEURYSMAL SUBARACHNOID HEMORRHAGE A dual-center study introduces a new Machine Learning model to predict Shunt-Dependent Hydrocephalus (SDHC) after aneurysmal subarachnoid hemorrhage (aSAH). Traditional prediction tools rely on early admission snapshots, limiting their predictive accuracy (AUC 0.60 to 0.70). This model uses a Day-21 landmark design to capture dynamic, weekly patient trajectories. 📊 KEY FINDINGS 🔹 The Random Forest model achieved a high internal validation AUC of 0.894 and external validation AUC of 0.867 🔹 Adding dynamic features significantly outperformed static-only models (AUC 0.894 vs. 0.831) 🔹 The model provides clear risk separation, allowing precise patient stratification. 🎯 TOP PREDICTORS OF SDHC 1. Glasgow Coma Scale (GCS) score at admission. 2. Hunt-Hess grade. 3. Evans index total change (progressive ventricular dilation). 4. Patient age 5. Intracranial pressure (ICP) coefficient of variation. 6. Postoperative 14-day CSF drainage volume. 🏥 CLINICAL IMPACT Using this model at Day 21, clinicians can stratify patients into low (5.3% risk), intermediate (35.7% risk), and high-risk (64.7% risk) tiers. This supports personalized, early shunt planning, optimized drainage, and shorter ICU stays. Reference: Zheng C, Huang J, Wang X, Ye Q, Lin C, Gu X. Development and external validation of a machine learning model based on multimodal dynamic data for predicting shunt-dependent hydrocephalus after aneurysmal subarachnoid hemorrhage: a dual-center retrospective cohort study. Front Neurol. 2026;17:1837898. doi:10.3389/fneur.2026.1837898 Read the full paper here: doi.org/10.3389/fneur.2026.1…
1
4
11
592
Calling all neurosurgeons and neurocritical care specialists! 🧠 The ACE Society Annual Meeting 2026 is happening on September 19 in Chicago. This event provides a vital forum to collaborate on clinical trial initiatives, analyze interim study data, and explore cutting-edge advancements in Active Cerebrospinal Fluid (CSF) exchange technology. Event Highlights: 📍 Location: Hyatt Central Chicago 🗓️ Date and Time: Sep 19, 2026, 7:00 AM to 4:00 PM CDT 🎓 CME Credit: Available through Northwestern University. 🗣️ Sessions: Includes presentations like "Active CSF Exchange vs Passive Drainage: a Meta Analysis" by Dr Behnam Rezai Jahromi Join the pivotal discussion aimed at advancing neurosurgical practices and improving patient outcomes by RSVPing today. @RabihTawkMD @babaksjahromi @NickBrandmeir @berejah @NorthwesternU For more info and how to register visit the ACE Society Annual Meeting 2026 website: ace-society.org/event-detail…
2
7
398
🧠 Head of bed positioning in stroke care: When does flat vs elevated actually matter? Two landmark trials provide a clear timeline for treating large vessel occlusion strokes. Phase 1: PRE-Thrombectomy (ZODIAC Trial) 🛑 Keep the bed FLAT. The ZODIAC trial showed that maintaining a 0-degree head position is a vital protective maneuver while awaiting surgery. Elevating the head to 30 degrees before clot removal increased the hazard of early neurological worsening by 34.4 times. Phase 2: POST-Thrombectomy (HeadSOAR Trial) 🛏️ You have flexibility. Hospitals often default to a 30 to 45-degree head elevation to prevent aspiration, especially in ventilated ICU patients. However, in non-intubated patients, the HeadSOAR trial adds highly useful information for routine stroke care. Once the clot is successfully removed, head positioning matters less. The trial evaluated 1,368 patients and found no significant difference in 90-day functional recovery or safety when comparing flat (0-10 degrees) versus elevated (30-40 degrees) positions for 72 hours. The Takeaway: New EMS stroke mantra? HOB Flat til Cath lab, then individualize. 🎬 Note: The attached infographic and video focus specifically on the newer HeadSOAR trial! Drop your comments and thoughts below 👇🏽 References: 1. Yuan Z, Peng J, Gu L, et al. Head positioning after endovascular therapy for acute stroke due to large vessel occlusion (HeadSOAR): multicentre randomised controlled trial. BMJ. 2026;394:e100363. doi.org/10.1136/bmj-2026-100… 2. Alexandrov AW, Shearin AJ, Mandava P, et al. Optimal Head-of-Bed Positioning Before Thrombectomy in Large Vessel Occlusion Stroke: A Randomized Clinical Trial. JAMA Neurol. 2025;82(9):905-914. doi.org/10.1001/jamaneurol.2… #MedTwitter
1
5
15
1,237
🧠 Predicting delirium in emergency critical care is a major hurdle. A new AI system called DeLiriuMAgents is changing that by simulating multidisciplinary clinical consultations. 🏥 Published in Cell Reports Medicine, this multi-agent LLM framework combines data-driven prediction with clinical specialist reasoning to catch delirium risk early in the emergency department. How it works: 1. EHRPromptAgent: Translates tabular EHR data into clinical narratives. 2. ModelPhysicianAgent: Employs a machine learning model to estimate baseline risk. 3. Multidisciplinary Specialists: Emergency, neurology, and psychiatry agents provide independent clinical reasoning. 4. MedEvidenceAgent: Pulls real-time guidelines and literature via Retrieval-Augmented Generation to reduce hallucinations. 5. DeliriumDecisionAgent: Coordinates all inputs to deliver an explainable, evidence-backed risk report. 📈 Performance and Validation: - Validated across MIMIC-IV, eICU-CRD, and a Chinese Peking University cohort. - Achieved predictive accuracies of 74.9% on MIMIC-IV, 73.1% on Peking University, and 67.0% on eICU-CRD. - Clinicians aided by these AI reports saw diagnostic accuracy rise from 62.0% to 65.0%. - 96% of clinicians surveyed are willing to use the system in real-world practice. 🧑‍⚕️ By shifting from post-onset treatment to early ED-ICU prediction, DeLiriuMAgents provides clinicians with the warnings they need to start early prevention. AMA Reference: Shang W, Shi T, Ma Q, Kong G. A large language model-driven multidisciplinary AI agent system predicts delirium in emergency critically ill patients. Cell Rep Med. 2026;7:102986. Article link: doi.org/10.1016/j.xcrm.2026.… @anirbanb_007 @STRXLabs
1
3
6
404
🎥 Explainer video by @Gemini_Notebook
65