Faculty Bibliography

BACKGROUND AND OBJECTIVES/OBJECTIVE:Traditional medical board examinations present clinical information in static vignettes with multiple-choices (MC), fundamentally different from how physicians gather and integrate data in practice. Recent advances in large language models (LLMs) offer promising approaches to creating more realistic clinical interactive conversations. However, these approaches are limited in neurosurgery, where patient communication capacity varies significantly and diagnosis heavily relies on objective data such as imaging and neurological examinations. We aimed to develop and evaluate a multi-artificial intelligence (AI) agent conversation framework for neurosurgical case assessment that enables realistic clinical interactions through simulated patients and structured access to objective clinical data. METHODS:We developed a framework to convert 608 Self-Assessment in Neurological Surgery first-order diagnosis questions into conversation sessions using 3 specialized AI agents: patient AI for subjective information, system AI for objective data, and clinical AI for diagnostic reasoning. We evaluated generative pretrained transformer 4o's (GPT-4o's) diagnostic accuracy across traditional vignettes, patient-only conversations, and patient + system AI interactions, with human benchmark testing from 10 neurosurgery residents. RESULTS:= .0030) using fewer interactions and reported high educational value of the interactive format. CONCLUSION/CONCLUSIONS:This multi-AI agent framework provides both a more challenging evaluation method for LLMs and an engaging educational tool for neurosurgical training. The significant performance drops in conversational formats suggest that traditional MC testing may overestimate LLMs' clinical reasoning capabilities, while the framework's interactive nature offers promising applications for enhancing medical education.

BACKGROUND AND OBJECTIVES/OBJECTIVE:articles into graphical abstracts using Cascade Styling Sheets (CSS) templates and iterative prompting of a frontier vision language model and to conduct a human evaluation of this pipeline. METHODS:We developed an automated pipeline to convert extracted manuscript content into standardized graphical abstracts. The pipeline implements a custom CSS profile designed to match existing journal standards. Using Claude Sonnet-3.5, we generated structured hypertext markup language summaries organized into 6 sections: Objectives, Background, Methods, Results, Discussion, and Conclusion. The model selected up to 2 representative figures per manuscript based on caption analysis. We evaluated performance using 100 randomly selected articles published between 2020 and 2024 (95 from Neurosurgery, 4 from Operative Neurosurgery, 1 from Neurosurgery Practice). Three Editorial Review Board members independently assessed abstracts using 3 binary criteria: (1) proper formatting, (2) factual accuracy, and (3) visual appeal. RESULTS:Generated graphical abstracts achieved proper formatting in 85% of cases (95% CI: 76.7%-90.7%), factual accuracy in 99% (95% CI: 94.4%-99.9%), and visual appropriateness in 82% (95% CI: 73.3%-88.3%). Overall, 70% of abstracts (95% CI: 60.5%-78.1%) met all 3 criteria and were deemed "publication ready" without manual intervention. Error analysis revealed poor figure selection (40.0%) as the most common failure mode, followed by title replacement errors from PDF extraction (26.7%). CONCLUSION/CONCLUSIONS:Our artificial intelligence-CSS pipeline demonstrates the feasibility of automating graphical abstract generation for neurosurgical manuscripts, achieving publication-ready quality in 70% of cases with 99% factual accuracy. This technology offers a scalable augmentation tool that can reduce the design burden for authors, enhancing visual scientific communication in neurosurgical publishing while complementing human expertise.

BACKGROUND AND PURPOSE/OBJECTIVE:IDH mutation & 1p/19q codeletion are critical biomarkers for glioma diagnosis & therapy. 1p/19q codeletion occurs exclusively in IDH-mutated gliomas. In this study, we developed a 2-stage, non-invasive, MRI-based deep learning method that leverages IDH status to enhance 1p/19q predictions. MATERIALS AND METHODS/METHODS:Predicted IDH-wildtype cases default to 1p/19q non-codeleted. Then the IDH-mutated cases were further classified for 1p/19q status using the 1p/19q-networks. RESULTS: CONCLUSIONS:to gate 1p/19q predictions. The developed method offers a reliable, non-invasive approach to determine important biomarkers for glioma diagnosis.

BACKGROUND AND PURPOSE/OBJECTIVE:-wt GBMs to help radiologists in differentiating them from nonmalignant mimic diagnoses (eg, encephalitis). Additionally, the histologic, genomic, and survival profiles of nonenhancing GBMs were compared with those of enhancing GBMs. MATERIALS AND METHODS/METHODS:-wt GBMs with nonmalignant mimics. Histopathologic and genomic analyses were performed on institutional cases. Overall survival between nonenhancing and enhancing GBMs was compared using Kaplan-Meier analysis. RESULTS:= .078). CONCLUSIONS:Nonenhancing GBMs demonstrate distinct MRI features that must be recognized for early diagnosis and differentiation from nonmalignant mimics. Nonenhancing GBMs demonstrated longer overall survival compared with enhancing GBMs, though they were not statistically significant.

BACKGROUND AND OBJECTIVES/OBJECTIVE:Surgery in NF2-related schwannomatosis (NF2-SWN) vestibular schwannoma (VS) carries a higher risk of facial nerve damage, hearing loss, and partial resection, than in sporadic cases. Radiosurgery is also associated with higher failure compared with sporadic schwannomas. Nowadays, bevacizumab (BEV) is frequently considered in the NF2-SWN population. However, some patients experience progression despite treatment. Among other surgical risks, in BEV-treated patients, hemorrhage and impaired healing are specific considerations. These concerns have led manufacturers to recommend stopping BEV 6 to 8 weeks preoperatively. The aim of our multicentric study was to assess the perioperative bleeding risk and postoperative outcomes in NF2-SWN patients undergoing VS surgery after preoperative BEV treatment. METHODS:Our retrospective analysis included medical and surgical records along with imaging reviews from 4 high-volume tertiary academic referral centers for NF2-SWN and VS. RESULTS:A total of 21 patients met the inclusion criteria. VS had a mean volume of 13.2 ±7.6 cm3 corresponding to 1 KOOS III and 20 KOOS IV. BEV was stopped at a mean of 5.8 ± 4.0 months before surgery with a total mean treatment duration of 33.7 ± 20.7 months and a monthly dose of 10.2 ± 4.1 mg/kg. Intraoperatively, the tumor was assessed to be bloody by the operating surgeons in 7 patients. Late BEV discontinuation and high cumulative dose independently predicted perioperative bleeding and longer surgery duration. No other complication such as wound dehiscence was reported. CONCLUSION/CONCLUSIONS:Our findings suggest that a higher cumulative BEV dose (∼600 mg/kg) and a longer interval between BEV discontinuation and surgery (∼8 months) are associated with a modest but statistically significant increase in intraoperative bleeding risk. Based on these observations, a BEV-free window between 6 weeks and 6 months (depending on the clinical scenario) before tumor resection seems optimal, particularly for patients with high cumulative exposure.

PURPOSE/OBJECTIVE:We describe a case of bilateral, multiple, branch retinal artery occlusions (BRAO) associated with carotid webs. METHODS:A thorough chart review was conducted for the patient. Relevant literature was systematically reviewed. RESULTS:Eight cases of fibromuscular dysplasia (FMD) associated with retinal artery occlusions have been reported. Two additional cases of FMD with other ocular involvement have been described. No cases of carotid webs associated with retinal artery occlusions were found. CONCLUSION/CONCLUSIONS:Carotid webs, an uncommon variant of FMD, are a recognized causative etiology of arterial, ischemic stroke. The case described here of bilateral, multifocal BRAOs represents a unique manifestation of this variant of FMD. This diagnosis should be considered in the setting of an otherwise unrevealing BRAO workup, as recognition of this association may be sight and life-saving.

BACKGROUND AND OBJECTIVES/OBJECTIVE:The rapid rise of electric and mechanical bikes and scooters has transformed urban transportation, but their neurosurgical consequences remain underexplored. This study aimed to evaluate micromobility-related injuries over time, examining mechanisms of injury, patient risk factors, injury patterns, and associated clinical outcomes at a Level-1 trauma center over a 5-year period. METHODS:We performed a retrospective review of patients who sustained micromobility-related injuries and presented to the Bellevue Hospital Center between 2018 and 2023. The cohort included riders of electric or mechanical bikes and scooters, as well as pedestrians struck by these devices. Key clinical variables and outcomes were compared across device types, both before and after propensity score matching. Unlike national database studies, this hospital-based analysis provides detailed clinical and neurosurgical outcome data. RESULTS:A total of 914 patients presented with micromobility-related injuries, accounting for 6.9% of all trauma admissions. Annual case volume and electric device involvement increased over time. The most common mechanism was collision with a motor vehicle (49.9%). Most patients (68.7%) required admission; 30.2% required intensive care. The median length of hospital stay was 3 days [IQR 1-5]. Half underwent a surgical intervention or procedure, and the overall mortality was 1.2%. Helmet use was low (31.7%). Pedestrians experienced the most severe outcomes, particularly when struck by electric devices. Injuries clustered during evening hours, suggesting modifiable environmental and behavioral risk factors. CONCLUSION/CONCLUSIONS:Micromobility-related trauma imposes a substantial neurosurgical burden, with frequent traumatic brain injury, intensive care unit utilization, and operative intervention. Unlike previous database studies, this hospital-based analysis provides detailed neurosurgical outcome data and identifies prevention targets-including helmet use, intoxication, and urban infrastructure-to reduce morbidity and resource utilization.

Polyaminopathies are a recently described family of rare genetic neurodevelopmental disorders. Polyaminopathies disrupt the biosynthesis of the primary polyamines: putrescine, spermidine, and spermine. Snyder-Robinson syndrome results from hemizygous loss-of-function variants in the spermine synthase (SMS) gene, resulting in decreased or complete loss of spermine synthase enzyme activity. Bachmann-Bupp syndrome results from heterozygous gain-of-function variants in the ornithine decarboxylase 1 (ODC1) gene, resulting in increased ornithine decarboxylase enzyme activity. Faundes-Banka syndrome results from heterozygous loss-of-function variants in the eukaryotic translation initiation factor 5A (EIF5A) gene, impairing eIF5A protein function. DHPS (deoxyhypusine synthase) deficiency is an autosomal recessive disease and results from bi-allelic hypomorphic variants in the deoxyhypusine synthase (DHPS) gene, which results in reduced deoxyhypusine synthase enzyme activity. Finally, DOHH (deoxyhypusine hydroxylase) disorder is an autosomal recessive disorder caused by bi-allelic loss-of-function variants in the deoxyhypusine hydroxylase (DOHH) gene, which causes decreased deoxyhypusine hydroxylase enzyme activity. Snyder-Robinson syndrome was first described in 1969, while the other four syndromes have only been identified in the past 7 years. A comprehensive phenotypic and genotypic description of these five syndromes is needed. We review the clinical and genetic features of these five polyaminopathies to create an inclusive clinical resource. A systematic keyword search strategy was used to identify all published cases in PubMed, Web of Science, and Scopus databases. The five known syndromes associated with the polyamine pathway share many similar clinical phenotypes, and yet patients with each syndrome present with distinctive syndromic features. This review will serve as a valuable resource for clinicians diagnosing and caring for patients with these rare polyaminopathies.

The frequency and nature of neurological exams (neuro-checks) in patients with severe acquired brain injury resulting in coma or disorders of consciousness (DoC) remain variable, with limited evidence guiding practice and poor understanding of their role in predicting and preventing neurological deterioration, functional recovery and adverse effects such as delirium. This scoping review aims to explore the frequency of bedside neurological exams within the first 7 days of injury impact on clinical outcomes in adult patients with severe acquired brain injury including mortality, neurological deterioration, long-term function, and delirium. METHODS: A comprehensive literature search was conducted using the PubMed, CINAHL, Medline and EMBASE databases from 2003 to 2023. Search terms captured a range of acute brain injuries and neuro-assessment tools. Eligible studies included adult patients with severe traumatic or non-traumatic brain injury or stroke that addressed frequency of bedside neurological exams within the first 7 days of admission. RESULTS: Of 1327 studies screened, 20 met inclusion criteria, representing over 16,000 patients across 14 countries. Assessment tools varied, but use of the Glasgow Coma Scale was prevalent. Frequency of neuro-checks ranged from hourly to daily. Multiple outcome measures were utilized. Some studies found that continuing hourly neuro-checks beyond the first 48 hours did not provide additional clinical benefit. Others associated excessive assessment with increased stress or delirium. CONCLUSION: There is very low evidence supporting an association between the frequency of neuro-checks and functional outcomes, mortality, length of stay, or delirium. Although early assessments may aid prognostication, excessive exams may not improve outcomes and may contribute to harm. The heterogeneity, lack of evidence, and limited standardization of neuro-check frequency highlight the need for clinical research to guide future practice.