Objectives This study aimed to explore the patient, disease and service-level factors that may influence the psychosocial and functional impact of an unruptured AVM diagnosis. Methods A mixed-methods study was performed in a single-centre, high-volume, tertiary neurosurgical centre. The study comprised of psychological instruments and a semi-structured interview focusing on patient experiences in the early diagnostic journey. 37 patients completed the HADS and SF-36 questionnaires and of those, 33 took part in the semi-structured interviews. HADS and SF-36 questionnaire scores were compared against UK normative data and multiple regression analyses were performed to identify significant predictors of psychosocial and functional burden. Thematic analysis of semi-structured interviews was also performed. Results All of the HADS and the majority of SF-36 sub-domains were found to be significantly worse compared to UK normative data. Previous psychiatric history was found to be significantly associated with a worsened score in the HADS Anxiety sub-domain (p=0.024) as well as energy (p=0.001), social function (p= 0.03) and general health (p=0.004) SF-36 subdomains. Thematic analysis revealed key themes of functional impact, psychological impact, diagnosis and explanation, access to information and follow-up care. Conclusions This study identifies a significant psychosocial and functional burden on those diagnosed with an unruptured AVM. Our mixed-methods analysis provide potential avenues in mitigating this effect and tailoring neurosurgical practice to improve patient care. Recommendations for further research include longitudinal study to investigate effects of time and AVM treatment on patients’ psychosocial and functional burden.

OBJECTIVE:The incidence of vestibular schwannoma (VS) diagnosis among octogenarians has increased due to longer life expectancy and greater MRI accessibility and utilization. Stereotactic radiosurgery (SRS) offers a safe management strategy for these patients. METHODS:The authors retrospectively analyzed the data of octogenarians with VS managed with SRS across 27 institutions, assessing clinical and audiological outcomes, including survival rates, tumor control responses, complications, and posttreatment functional outcomes. RESULTS:Among 309 octogenarians with VS managed with single-fraction SRS, overall survival was 97.8% at 1 year, 88.5% at 3 years, and 78.1% at 5 years; the corresponding progression-free survival rates were 95.8%, 81.1%, and 71.1%, respectively. In competing risk analysis, tumor progression occurred in 5.2% of patients by 5 years, whereas death without progression reached 19.8%. Tumor control was achieved in 94.8% at 3 and 5 years. At last imaging follow-up (median 33.5 months, IQR 14-60 months), 29.1% showed volume regression. Communicating hydrocephalus developed in 4.2% and was successfully managed with CSF diversion. Age, sex, tumor volume, Koos grade, and margin dose were not significant predictors of patient survival or tumor progression. CONCLUSIONS:Single-fraction SRS is a safe and effective management modality for octogenarians with VS, providing durable tumor control with minimal morbidity.

Abstract Background and Objectives Recurrence following burr-hole drainage of chronic subdural haematoma (cSDH) occurs in 10-25% of cases, sustained by neovascularisation of the subdural neomembrane supplied by the middle meningeal artery (MMA). MMA embolisation reduces recurrence; whether incidental burr-hole intersection of MMA branches during drainage confers similar benefit is unknown. Methods We performed a multicentre retrospective cohort study of consecutive adults undergoing burr-hole drainage for cSDH at two UK tertiary neurosurgical centres. Postoperative thin-slice CT was used to classify burr-hole intersection of the underlying MMA groove (no hit, distal-branch hit or main-branch hit) and measure perpendicular burr-hole-to-MMA-groove distance. Co-primary outcomes were radiological recurrence and recurrence requiring intervention. Patient-clustered multivariable logistic regression adjusted for prespecified clinical covariates and treating site. Results 227 patients (284 operated hemispheres) were included. Radiological recurrence decreased from 34.4% with no branch hit to 22.9% with main-branch intersection, with the gradient confined predominantly to unilateral cSDH. Main-branch intersection was associated with lower adjusted odds of radiological recurrence in unilateral cSDH (adjusted OR 0.30, 95% CI 0.11-0.81; P = .018), with a similar but non-significant association in the overall cohort (adjusted OR 0.53, 95% CI 0.26-1.07; P = .075). Burr-hole-to-MMA-groove distance demonstrated a more consistent association: in the overall cohort, each 5-mm increase independently increased the odds of radiological recurrence (adjusted OR 1.38, 95% CI 1.04-1.82; P = .025). In unilateral cSDH, each 5-mm increase was independently associated with both radiological recurrence (adjusted OR 1.45, 95% CI 1.03-2.04; P = .034) and recurrence requiring intervention (adjusted OR 1.52, 95% CI 1.05-2.20; P = .027). Conclusion Main-branch intersection of the middle meningeal artery during routine burr-hole surgery is associated with lower recurrence of unilateral cSDH, while the accompanying burr-hole-to-MMA-groove distance gradient provides biologically plausible support for a dose-response relationship. Together, these findings provide mechanistic rationale for prospective evaluation of intentional neuronavigation-guided MMA targeting (BURR-MMA; NCT07549893 ).

Machine learning (ML) models have been increasingly applied to predict postoperative facial nerve dysfunction and hearing preservation after vestibular schwannoma (VS) surgery. However, reported performance varies substantially, and the overall diagnostic accuracy and clinical reliability of these models remain uncertain. We conducted a systematic review and diagnostic test accuracy meta-analysis to characterise the current state and methodological readiness of ML-based prediction of these outcomes. PubMed, Embase, and CENTRAL were searched from inception to February 2026. Studies evaluating ML-based prediction of facial nerve function or hearing preservation following VS surgery were included. Diagnostic performance metrics were pooled using random-effects generalised linear mixed models. Sensitivity, specificity, diagnostic odds ratio, and AUC were synthesised, and SROC curves were constructed. The prespecified primary synthesis pooled the single best model per study; small-study effects were assessed with Deeks’ test. Risk of bias (PROBAST) and certainty of evidence (GRADE) were assessed. Ten retrospective cohort studies encompassing 1270 patients and 56 ML models met inclusion criteria. In the prespecified primary analysis pooling the single best model per study, the summary AUC was 0.91 for facial nerve dysfunction (sensitivity 0.89, specificity 0.86) and 0.92 for hearing preservation (sensitivity 0.88, specificity 0.96). Pooling all models on held-out test data gave a facial nerve AUC of 0.81; test-set data were too sparse for a stable hearing estimate, for which only training performance could be pooled (AUC 0.79). Tumour size, age, tumour location, and baseline hearing status were the most frequently identified influential predictors. Most studies were at unclear or high risk of bias (PROBAST has no intermediate “moderate” category), and certainty of evidence was moderate for facial nerve dysfunction and low for hearing preservation, the latter reflecting significant small-study effects (Deeks’ p = 0.004). ML-based models demonstrate promising discrimination for predicting postoperative facial nerve and hearing outcomes after VS surgery. However, heterogeneity, limited external validation, and inconsistent reporting of calibration constrain inference regarding transportability and clinical implementation.

Microvascular neurosurgery training is limited by low case volumes, high complexity, and significant risks to the patient. Simulation offers a safe, high‐repetition alternative for skills acquisition for trainees. The purpose of this study was to evaluate the UpSurgeOn Mycro Box for end-to-end microvascular anastomosis by establishing quantitative predictive performance thresholds of surgical expertise. Twenty-nine participants performed a standardised end-to-end microvascular anastomosis using the Mycro Box simulator. Quantitative performance metrics were extracted for each trial. A multivariable logistic regression model with ridge penalization was trained using nested crossvalidation to classify participants as expert (i.e. status as attending neurosurgeon) versus non-expert based on these metrics. Model discrimination was assessed by balanced accuracy and area under the ROC curve (AUC), and 95% confidence intervals were obtained via bootstrapped resampling. The best performing model was used to derive quantitative performance thresholds corresponding to a 50% probability of expert-level classification. The best-performing model incorporated operative time, number of sutures, and Objective Structured Assessment of Technical Skills (OSATS) score as predictors. This model achieved outstanding discriminative ability with a balanced classification accuracy of 94% and a ROC-AUC of 0.98. All three performance metrics were independent predictors of expertise. From the model, we generated objective proficiency benchmarks on the simulator, visualised with a faceted bivariate regression plot. Performance on the Mycro Box simulator could predict a surgeon’s level of expertise, supporting the simulator’s construct validity. An interpretable regression model provided quantitative performance thresholds for proficiency. These findings suggest that objective metrics from simulation can define expert-level performance benchmarks, which may be used to guide training curricula and competency-based progression in neurosurgery. N/A

Introduction: Vestibular schwannoma (VS) is the most common pathology in the lateral skull base, with a rising prevalence resulting in over 3,500 cases per 100,000 per year in the United States between 2004 and 2016. Management varies by tumor size; smaller tumors are typically managed with serial imaging, while larger, symptomatic tumors often require surgery. The primary goal of VS surgery is maximal tumor removal while preserving neurological function. Facial nerve palsy remains a significant concern, with large VS (> 30 mm in diameter) significantly more likely to result in facial paralysis compared to small tumors.

INTRODUCTION:Following endovascular treatment (EVT) or microsurgical treatment (MT) of intracranial aneurysms (IA), radiological follow-up is performed to assess for recurrence and to determine the need for re-treatment. There is a paucity of evidence describing the long-term results of EVT and MT for IA and therefore data to inform the design of follow-up protocols are lacking. The overarching aim of the META study is to determine the clinically relevant long-term outcomes of EVT and MT for IA, and use this data to create evidence based radiological and clinical follow-up protocols for these aneurysms. METHODS AND ANALYSIS:The META study will be a multicentre, retrospective cohort study. Data collection will begin in June 2024 and all IA treated with EVT or MT meeting inclusion criteria between 1st January 2013 and 31st of December 2013 will be included, to allow for a maximum of 10 years of radiological and clinical follow-up. Clinical and radiological data will be collected and stored on a secure online database. Following the completion of data collection, factors associated with re-treatment or subarachnoid haemorrhage from an aneurysm treated with EVT or MT will be identified and used to risk stratify IAs, with a view to developing an evidence-based follow-up protocol of IA treated with EVT or MT. ETHICS AND DISSEMINATION:This project will be registered with the Royal College of Surgeons in Ireland (RCSI) Research Ethics Committee (REC). It will also be registered locally at each participating centre and appropriate local approvals will be obtained. The results of the study will be disseminated through presentation at national and international meetings, and publication in peer reviewed journals. STRENGTHS AND LIMITATIONS:This study will be the first contemporary multicentre study examining the long-term outcomes following treatment of ruptured and unruptured intracranial aneurysms.Our study will follow up treated aneurysms over a prolonged period of up to ten years; such prolonged follow-up is essential in the counselling of patients with this pathology, the majority of whom are in the fifth and sixth decades of life.The multicentre study design will increase the external validity and applicability of the results.The study will not assess aneurysm occlusion directly; data on significant aneurysm recurrences requiring re-intervention or leading to aneurysm rupture will be collected. PLAIN ENGLISH SUMMARY:Intracranial aneurysms (IAs) are abnormal outpouchings or dilations on the main blood vessels supplying the brain. 3% of the general population have IAs and the majority will remain asymptomatic however a proportion will go on to rupture and cause subarachnoid haemorrhage (SAH), which is a condition associated with a significant rate of death and disability. The treatment of both ruptured and unruptured IAs aims to prevent future aneurysm rupture leading to SAH. Historically, the only treatment modality for IAs was a neurosurgical operation called microsurgical clipping but more recently endovascular treatment, whereby the aneurysms can be treated using catheters inserted into the arteries of the groyne or wrist has become the predominant treatment. However, there are very few studies describing the long-term outcomes of microsurgical clipping or endovascular treatment, particularly with respect to the requirement for aneurysm re-treatment or SAH due to rupture of a previously treated aneurysm. The aim of this study is to assess the outcomes following microsurgical clipping and endovascular treatment of a large number of both ruptured and unruptured IA over an extended follow-up period, with a view to providing data that will permit us to optimise the follow-up of patients with IA following treatment.

BACKGROUND AND OBJECTIVES:Machine learning (ML) in surgical video analysis offers promising prospects for training and decision support in surgery. The past decade has seen key advances in ML-based operative workflow analysis, though existing applications mostly feature shorter surgeries (<2 hours) with limited scene changes. The aim of this study was to develop and evaluate a ML model capable of automated operative workflow recognition for retrosigmoid vestibular schwannoma (VS) resection. In doing so, this project furthers previous research by applying workflow prediction platforms to lengthy (median >5 hours duration), data-heavy surgeries, using VS resection as an exemplar. METHODS:A video dataset of 21 microscopic retrosigmoid VS resections was collected at a single institution over 3 years and underwent workflow annotation according to a previously agreed expert consensus (Approach, Excision, and Closure phases; and Debulking or Dissection steps within the Excision phase). Annotations were used to train a ML model consisting of a convolutional neural network and a recurrent neural network. 5-fold cross-validation was used, and performance metrics (accuracy, precision, recall, F1 score) were assessed for phase and step prediction. RESULTS:Median operative video time was 5 hours 18 minutes (IQR 3 hours 21 minutes-6 hours 1 minute). The "Tumor Excision" phase accounted for the majority of each case (median 4 hours 23 minutes), whereas "Approach and Exposure" (28 minutes) and "Closure" (17 minutes) comprised shorter phases. The ML model accurately predicted operative phases (accuracy 81%, weighted F1 0.83) and dichotomized steps (accuracy 86%, weighted F1 0.86). CONCLUSION:This study demonstrates that our ML model can accurately predict the surgical phases and intraphase steps in retrosigmoid VS resection. This demonstrates the successful application of ML in operative workflow recognition on low-volume, lengthy, data-heavy surgical videos. Despite this, there remains room for improvement in individual step classification. Future applications of ML in low-volume high-complexity operations should prioritize collaborative video sharing to overcome barriers to clinical translation.

Background and objectives:It is important to establish a platform that allows methodical recording of treatments provided for brain arteriovenous malformations (bAVMs) which is a complex and heterogenous disease. In this preliminary report, the authors present the early analysis of the treatment of bAVMs from the British & Irish Brain AVM Registry (BIBAR). Research question:Can a multicenter registry effectively capture bAVMs presentation and treatment data? Materials & methods:The British Neurovascular Group (BNVG) set up a bAVMs registry working group in November 2018, with the primary aim of trying to ascertain the number and types of treatments provided for bAVMs across the United Kingdom. Results:Between January 1, 2019 to December 31, 2023, treatment decisions were recorded for 1969 registered patients with bAVMs, of which 1713 patients received treatment at the time of the analysis. 56.28 % (964) patients had no evidence of rupture at the time of the initial treatment decision, whilst 43.72 % (749) presented with evidence of rupture at initial presentation. Of these, 83.31 % (624) were treated with radiosurgery, 13.62 % (102) with surgery and 0.93 % (7) underwent embolization. Age was negatively correlated with likelihood of surgical treatment. Patients who did not receive any treatment at the time of this analysis were not included. Discussion and conclusion:We have shown that with a collective, collaborative effort, a national bAVM registry is feasible and as data capture becomes more complete, can provide valuable data on treatment types and volunes and provide an insight into the decision making underlying those treatments.

Giant intracranial aneurysms (>25 mm) represent 5% of aneurysms1 and have a poor natural history with high risk of rupture.2 We describe a 74-year-old woman presenting with cognitive decline and dysphasia, initially having been referred to a memory clinic for a potential diagnosis of dementia. Magnetic resonance imaging demonstrated a left temporal lobe mass with surrounding edema, confirmed as a partially thrombosed giant middle cerebral artery aneurysm on catheter angiography. Awake balloon test occlusion was passed without deficit preoperatively. She then underwent craniotomy with opening of the aneurysm sac under temporary clipping and removal of the thrombus followed by clip reconstruction of the aneurysm neck. Postoperative magnetic resonance imaging demonstrated resolution of the edema and mass effect, and angiography confirmed aneurysm occlusion. Dysphasia and cognitive deficits resolved on postoperative psychometric testing. Written patient consent for surgical video publication and institutional approval was obtained (Video 1).

Background Natural language processing (NLP), a subset of artificial intelligence (AI), aims to decipher unstructured human language. This study showcases NLP's application in surgical health care, focusing on vestibular schwannoma (VS). By employing an NLP platform, we identify prevalent text concepts in VS patients' electronic health care records (EHRs), creating concept panels covering symptomatology, comorbidities, and management. Through a case study, we illustrate NLP's potential in predicting postoperative cerebrospinal fluid (CSF) leaks. Methods An NLP model analyzed EHRs of surgically managed VS patients from 2008 to 2018 in a single center. The model underwent unsupervised (trained on one million documents from EHR) and supervised (300 documents annotated in duplicate) learning phases, extracting text concepts and generating concept panels related to symptoms, comorbidities, and management. Statistical analysis correlated concept occurrences with postoperative complications, notably CSF leaks. Results Analysis included 292 patients' records, yielding 6,901 unique concepts and 360,929 occurrences. Concept panels highlighted key associations with postoperative CSF leaks, including "antibiotics," "sepsis," and "intensive care unit admission." The NLP model demonstrated high accuracy (precision 0.92, recall 0.96, macro F1 0.93). Conclusion Our NLP model effectively extracted concepts from VS patients' EHRs, facilitating personalized concept panels with diverse applications. NLP shows promise in surgical settings, aiding in early diagnosis, complication prediction, and patient care. Further validation of NLP's predictive capabilities is warranted.

Objective Despite advances in skull-base reconstruction techniques, cerebrospinal fluid (CSF) leaks remain a common complication following retrosigmoid (RS) vestibular schwannoma (VS) surgery. We aimed to review and classify the available strategies used to prevent CSF leaks following RS VS surgery.Methods A systematic review, including studies of adults undergoing RS VS surgery since 2000, was conducted. Repair protocols were synthesized into a narrative summary, and a taxonomic classification of techniques and materials was produced. Additionally, the advantages, disadvantages, and associated CSF leak rates of different repair protocols were described.Results All 42 studies were case series, of which 34 were retrospective, and eight were prospective. Repair strategies included heterogeneous combinations of autografts, xenografts, and synthetic materials. A repair taxonomy was produced considering seven distinct stages to CSF leak prevention, including intraoperative approaches to the dura, internal auditory canal (IAC), air cells, RS bony defect, extracranial soft tissue, postoperative dressings, and CSF diversion. Notably, there was significant heterogeneity among institutions, particularly in the dural and IAC stages. The median postoperative incidence of CSF leaks was 6.3% (IQR: 1.3-8.44%).Conclusions The intraoperative strategies used to prevent CSF leaks during RS VS surgery vary between and within institutions. As a result of this heterogeneity and inconsistent reporting of CSF leak predictive factors, a meaningful comparative analysis of repair protocols was not feasible. Instead, we propose the development of a prospective multicenter observational evaluation designed to accurately capture a comprehensive dataset of potential CSF risk factors, including all stages of the operative repair protocol.

Non-traumatic posterior fossa haemorrhage accounts for approximately 10% of all intracranial haematomas, and 1.5% of all strokes. In the posterior fossa, a small amount of mass effect can have dramatic effects, due to its small volume. This can be due to immediate transmission of pressure to the brainstem, or via occlusion of the aqueduct of Sylvius or compression of the fourth ventricle, leading to acute obstructive hydrocephalus, with the risk of tonsillar herniation. Timely investigations and management are essential to maximise good outcomes. This Element offers a brief overview of posterior fossa haemorrhage. It looks at the anatomy, aetiology, management, and surgical options, with a review of the available evidence to guide practice.

BACKGROUND: The introduction of the electronic health record (EHR) has improved the collection and storage of patient information, enhancing clinical communication and academic research. However, EHRs are limited by data quality and the time-consuming task of manual data extraction. This study aimed to use process mapping to help identify critical data entry points within the clinical pathway for patients with vestibular schwannoma (VS) ideal for structured data entry and automated data collection to improve patient care and research. METHODS: A 2-stage methodology was used at a neurosurgical unit. Process maps were developed using semi-structured interviews with stakeholders in the management of VS resection. Process maps were then retrospectively validated against EHRs for patients admitted between August 2019 and December 2021, establishing critical data entry points. RESULTS: In the process map development, 20 stakeholders were interviewed. Process maps were validated against EHRs of 36 patients admitted for VS resection. Operative notes, surgical inpatient reviews (including ward rounds), and discharge summaries were available for all patients, representing critical data entry points. Areas for documentation improvement were in the preoperative clinics (30/36; 83.3%), preoperative skull base multidisciplinary team (32/36; 88.9%), postoperative followup clinics (32/36; 88.9%), and postoperative skull base multidisciplinary team meeting (29/36; 80.6%). CONCLUSIONS: This is a first use to our knowledge of a 2-stage methodology for process mapping the clinical pathway for patients undergoing VS resection. We identified critical data entry points that can be targeted for structured data entry and for automated data collection tools, positively impacting patient care and research.

Objectives Despite advances in skull base reconstruction techniques, cerebrospinal fluid (CSF) leaks remain a relatively common complication after translabyrinthine (TL) vestibular schwannoma (VS) surgery. We conducted a systematic review to synthesize the repair techniques and materials used in TL VS surgery to prevent CSF leaks.Design A systematic review of studies published since 2000 reporting techniques to prevent CSF leaks during adult TL VS surgery was conducted. A narrative synthesis of primary repair protocols was produced, and a taxonomy was established. Additionally, the advantages, disadvantages, and associated CSF leak rates of different repair protocols were extracted.Results All 43 studies were case series, and 39 were retrospective. Repair strategies included heterogeneous combinations of autografts, xenografts, and synthetic materials. A taxonomy was produced, classifying repairs into seven distinct stages, including approaches to the dura, middle ear cleft, air cells, TL bony defect, extra-cranial soft tissue, postoperative dressings, and CSF diversion. The median postoperative incidence of CSF leaks was 6% (interquartile range: 0-10%).Conclusions This systematic review reveals substantial inter-institutional heterogeneity in intraoperative strategies to prevent CSF leaks following TL VS surgery. However, comparing these techniques is challenging due to the multiple predictive factors for CSF leaks and their inconsistent reporting. We propose a taxonomy of seven stages to classify operative techniques and materials aimed at preventing CSF leaks. We recommend that future evaluations should adopt a prospective approach encompassing data collection strategies that considers all operative stages described by our taxonomy.

Objective: To compare the ability of a deep-learning platform (the MACSSwin-T model) with health care professionals in detecting cerebral aneurysms from operative videos. Secondly, we aimed to compare health care professionals' ability to detect cerebral aneurysms with and without artificial intelligence (AI) assistance. Background: Modern microscopic surgery enables the capture of operative video data on an unforeseen scale. Advances in computer vision, a branch of AI, have enabled automated analysis of operative video. These advances are likely to benefit clinicians, health care systems, and patients alike, yet such benefits are yet to be realized. Methods: In a cross-sectional comparative study, neurosurgeons, anesthetists, and operating room nurses, all at varying stages of training and experience, reviewed still frames of aneurysm clipping operations and labeled frames as "aneurysm not in frame" or "aneurysm in frame." Frames then underwent analysis by the AI platform. A second round of data collection was performed, whereby the neurosurgical team had AI assistance. The accuracy of aneurysm detection was calculated for human-only, AI-only, and AI-assisted human groups. Results: A total of 5154 individual frame reviews were collated from 338 health care professionals. Health care professionals correctly labeled 70% of frames without AI assistance, compared with 78% with AI assistance (odds ratio: 1.77, P < 0.001). Neurosurgical Attendings showed the greatest improvement, from 77% to 92% correct predictions with AI assistance (odds ratio: 4.24, P = 0.003). Conclusions: AI-assisted human performance surpassed both human and AI alone. Notably, across health care professionals surveyed, frame accuracy improved across all subspecialties and experience levels, particularly among the most experienced health care professionals. These results challenge the prevailing notion that AI primarily benefits junior clinicians, highlighting its crucial role throughout the surgical hierarchy as an essential component of modern surgical practice.

Objective: This study aimed to compare the ability of a deep-learning platform (the MACSSwin-T model) with healthcare professionals in detecting cerebral aneurysms from operative videos. Secondly, we aimed to compare the neurosurgical team’s ability to detect cerebral aneurysms with and without AI-assistance. Background: Modern microscopic surgery enables the capture of operative video data on an unforeseen scale. Advances in computer vision, a branch of artificial intelligence (AI), have enabled automated analysis of operative video. These advances are likely to benefit clinicians, healthcare systems, and patients alike, yet such benefits are yet to be realised. Methods: In a cross-sectional comparative study, neurosurgeons, anaesthetists, and operating room (OR) nurses, all at varying stages of training and experience, reviewed still frames of aneurysm clipping operations and labelled frames as “aneurysm not in frame” or “aneurysm in frame”. Frames then underwent analysis by the AI platform. A second round of data collection was performed whereby the neurosurgical team had AI-assistance. Accuracy of aneurysm detection was calculated for human only, AI only, and AI-assisted human groups. Results: 5,154 individual frame reviews were collated from 338 healthcare professionals. Healthcare professionals correctly labelled 70% of frames without AI assistance, compared to 78% with AI-assistance (OR 1.77, P<0.001). Neurosurgical Attendings showed the greatest improvement, from 77% to 92% correct predictions with AI-assistance (OR 4.24, P=0.003). Conclusion: AI-assisted human performance surpassed both human and AI alone. Notably, across healthcare professionals surveyed, frame accuracy improved across all subspecialties and experience levels, particularly among the most experienced healthcare professionals. These results challenge the prevailing notion that AI primarily benefits junior clinicians, highlighting its crucial role throughout the surgical hierarchy as an essential component of modern surgical practice.

Background: Cerebral cavernous malformations (CCM) may undergo a period of clinical and/or radiographical surveillance that precedes or follows definitive treatment. There are no international guidelines on the optimal surveillance strategy. This study describes the surveillance strategies at our centre and explore the related clinical outcomes.Methods: We performed a retrospective study of adult patients with CCMs referred to a neurovascular service over an 8-year period, to determine the frequency and type of surveillance, intervention, and explore the associated outcomes. We report our findings adhering to STROBE guidelines. Results: 133 patients (Male:Female 73:60; men age 42 years; range 12-82) were included. CCMs were identified in patients first presenting with symptomatic intracerebral haemorrhage (42.11%); headache, focal neurological deficit, or seizure without haemorrhage (41.35%); or, as an incidental finding (16.54%). The most common CCM location was supratentorial (59.40%), followed by brain stem (21.80%), cerebellum (10.53%) and basal ganglia (6.02%). Of the 133 patients, 77 patients (57.89%) were managed conservatively, 49 patients (36.84%) were managed by surgical resection alone, and seven patients (5.26%) were managed with stereotactic radiosurgery (SRS).Patients follow-up had a mean duration of 65.94 months, and varied widely (SD = 52.59; range 0-265), for a total of 730.83 person-years of follow up. During surveillance, 16 patients suffered an ICH equating to a bleeding rate of 2.19 per 100 patient years. CCMs that increased in size had a higher bleeding rate (p = 8.58 x10- 4). There were 8 (6.02%) cases where routine clinic review or MRI resulted in a change in management.Conclusions: Our single centre retrospective study supports existing literature relating to presentation and sequalae of CCM, with an increase in CCM size being associated with higher rates of detected bleeding. There remains heterogeneity, even within a single centre, on the frequency and modality of surveillance. Further, there are no international guidelines or high-quality data that recommends the optimal duration and frequency of surveillance, and its effect on clinical outcomes. This is a future research direction.