Radiomic signatures of meningiomas using the Ki-67 proliferation index as a prognostic marker of clinical outcomes.

OBJECTIVE: The clinical behavior of meningiomas is not entirely captured by its designated WHO grade, therefore other factors must be elucidated that portend increased tumor aggressiveness and associated risk of recurrence. In this study, the authors identify multiparametric MRI radiomic signatures of meningiomas using Ki-67 as a prognostic marker of clinical outcomes independent of WHO grade. METHODS: A retrospective analysis was conducted of all resected meningiomas between 2012 and 2018. Preoperative MR images were used for high-throughput radiomic feature extraction and subsequently used to develop a machine learning algorithm to stratify meningiomas based on Ki-67 indices < 5% and ≥ 5%, independent of WHO grade. Progression-free survival (PFS) was assessed based on machine learning prediction of Ki-67 strata and compared with outcomes based on histopathological Ki-67. RESULTS: Three hundred forty-three meningiomas were included: 291 with WHO grade I, 43 with grade II, and 9 with grade III. The overall rate of recurrence was 19.8% (15.1% in grade I, 44.2% in grade II, and 77.8% in grade III) over a median follow-up of 28.5 months. Grade II and III tumors had higher Ki-67 indices than grade I tumors, albeit tumor and peritumoral edema volumes had considerable variation independent of meningioma WHO grade. Forty-six high-performing radiomic features (1 morphological, 7 intensity-based, and 38 textural) were identified and used to build a support vector machine model to stratify tumors based on a Ki-67 cutoff of 5%, with resultant areas under the curve of 0.83 (95% CI 0.78-0.89) and 0.84 (95% CI 0.75-0.94) achieved for the discovery (n = 257) and validation (n = 86) data sets, respectively. Comparison of histopathological Ki-67 versus machine learning-predicted Ki-67 showed excellent performance (overall accuracy > 80%), with classification of grade I meningiomas exhibiting the greatest accuracy. Prediction of Ki-67 by machine learning classifier revealed shorter PFS for meningiomas with Ki-67 indices ≥ 5% compared with tumors with Ki-67 < 5% (p < 0.0001, log-rank test), which corroborates divergent patient outcomes observed using histopathological Ki-67. CONCLUSIONS: The Ki-67 proliferation index may serve as a surrogate marker of increased meningioma aggressiveness independent of WHO grade. Machine learning using radiomic feature analysis may be used for the preoperative prediction of meningioma Ki-67, which provides enhanced analytical insights to help improve diagnostic classification and guide patient-specific treatment strategies.

Middle Meningeal Artery Embolization for Subdural Hematoma: An Institutional Cohort and Propensity Score-Matched Comparison with Conventional Management.

BACKGROUND: The treatment of non-acute subdural hematoma (NASDH) is challenging due to its high recurrence rates and comorbidities of mostly elderly patients. Middle meningeal artery embolization (MMAE) recently emerged as an alternative to surgery in the treatment of NASDH. OBJECTIVE: To describe a single center's experience of MMAE with Onyx for NASDH and compare it to a surgically treated historical cohort. METHODS: We performed a retrospective analysis of patients undergoing MMAE for NASDH from 2019 to 2021. MMAE was performed with ethylene vinyl alcohol copolymer (Onyx). Comparisons were made with a historical cohort from 2010 to 2018 who underwent surgical evacuation only, before and after propensity score matching. Outcomes were assessed clinically and radiographically. RESULTS: We included 44 consecutive patients (55 MMAEs) who underwent MMAE, with a median follow-up of 63.5 days. Twenty-four NASDHs underwent upfront embolization, 17 adjunctive, and 14 for recurrence after prior surgical evacuation, with no significant differences in hematoma and mRS reduction between them. Two patients died during hospitalization and 2 during follow-up, unrelated to the procedure. Mean SDH thickness decreased by 48.3% ± 38.1% (P < 0.001) on last follow-up, which did not correlate with the amount of Onyx injected. Six (13.6%) patients required surgical rescue after embolization. There were no procedure-related complications. The mean modified Rankin Scale (mRS) on admission was 2.8 ± 1.5, which decreased significantly to 1 [1,4] at the last follow-up (P = 0.033). The MMAE (41 hematomas; upfront and adjunctive embolization) and Surgical Evacuation-only (461 hematomas) cohorts were balanced with propensity score mathing. Matching was successful for 41 MMAE and 41 surgical-only hematoma pairs, and only hypertension remained significantly different between the two groups, but there was no significant difference in any outcome. CONCLUSION: MMAE for NASDH seems safe and effective in appropriately selected patients, non-inferior to surgery, and may become a minimally-invasive alternative. Given our encouraging results, large-scale clinical randomized trials are warranted.

Machine learning for outcome prediction of neurosurgical aneurysm treatment: Current methods and future directions.

INTRODUCTION: Machine learning algorithms have received increased attention in neurosurgical literature for improved accuracy over traditional predictive methods. In this review, the authors sought to assess current applications of machine learning for outcome prediction of neurosurgical treatment of intracranial aneurysms and identify areas for future research. METHODS: A PRISMA-compliant systematic review of the PubMed, MEDLINE, and EMBASE databases was conducted for all studies utilizing machine learning for outcome prediction of intracranial aneurysm treatment. Patient characteristics, machine learning methods, outcomes of interest, and accuracy metrics were recorded from included studies. RESULTS: 16 studies were ultimately included in qualitative synthesis. Studies primarily analyzed angiographic outcomes, functional outcomes, or complication prediction using clinical, radiological, or composite variables. The majority of included studies utilized supervised learning algorithms for analysis of dichotomized outcomes. CONCLUSIONS: Commonly included variables were demographics, presentation variables (including ruptured or unruptured status), and treatment used. Areas for future research include increased generalizability across institutions and for smaller datasets, as well as development of front-end tools for clinical applicability of published algorithms.

Improved Functional Outcomes of Stroke Patients Undergoing Mechanical Thrombectomy After Arriving via a Mobile Stroke Unit.

BACKGROUND: Mobile stroke units (MSUs) have been implemented worldwide for stroke care, but outcome data are lacking to show their efficacy specifically in patients undergoing mechanical thrombectomy (MT). Here, we include patients from our stroke network MSU and compare them to patients who arrived conventionally. METHODS: A retrospective review of a stroke database was performed to identify patients who underwent MT after arrival via an MSU from August 2019 to December 2020. Demographic factors, past medical history, stroke characteristics, treatment variables, complications, and functional outcomes were recorded. These were compared to date-matched patients who underwent MT after arrival via conventional means. RESULTS: Seven patients were treated with MT after arriving by an MSU. These patients were compared to 50 date-matched patients who underwent thrombectomy after arrival through conventional means. No statistically significant difference between cohorts was observed in terms of demographic variables, comorbidities, stroke characteristics, or tissue plasminogen activator administration. Patients from the MSU cohort had significantly shorter time from symptom onset to groin puncture time (191.33 minutes ±77.53 vs. 483.51 minutes ±322.66, P = 0.034). Importantly, MSU-transferred patients had significantly better discharge functional status measured by using the modified Rankin Scale (1.86 ± 1.35 vs. 3.57 ± 1.88, P = 0.024). No significant difference in final thrombolysis in cerebral infarction score, complications, length of stay, or mortality was observed. CONCLUSIONS: Our pilot study demonstrates the efficacy of the MSU in decreasing door-to-puncture time and a concordant improvement in the discharge modified Rankin Scale score. Further prospective studies are needed to assess cost-efficacy and optimal protocol for MSUs in stroke care.

Predictors of first-pass reperfusion for mechanical thrombectomy in acute ischemic stroke.

BACKGROUND AND OBJECTIVE: Rapid reperfusion of ischemic penumbra in patients with acute stroke is critical to neurological recovery. Achieving reperfusion after first-pass mechanical thrombectomy has been associated with improved patient outcomes. However, the predictors for obtaining first-pass reperfusion are not well known. METHODS: A single-institution retrospective study of all patients who underwent mechanical thrombectomy at a tertiary care center from January 2010 until March 2019 was conducted to assess for predictors of first-pass reperfusion. RESULTS: A total of 257 patients were reviewed. Successful reperfusion was obtained in 63.4% of patients, and in 38% of patients on the first pass. On multivariate analysis, increasing door-to-puncture time was a negative predictor of FPR (OR 0.989, 95% CI = 0.980-0.997) and use of combined thrombectomy technique with stent-retriever and aspiration was a positive predictor of FPR compared to aspiration or stent-retriever alone (OR 4.441, 95% CI = 1.001-19.699). CONCLUSIONS: Combination therapy using stent-retriever and aspiration may increase the chance of obtaining FPR, whereas delays in starting the procedure after patient arrival may decrease the odds of FPR. Rapid thrombectomy initiation and procedural technique may play in optimizing rates of FPR and ultimately patient outcomes, however, randomized controlled trials assessing these variables are necessary to determine optimal treatment strategies.

Frameless Angiography-Based Gamma Knife Stereotactic Radiosurgery for Cerebral Arteriovenous Malformations: A Proof-of-Concept Study.

BACKGROUND: Traditional Gamma Knife radiosurgery (GKRS) of brain arteriovenous malformations (AVMs) using digital subtraction angiography (DSA) requires head immobilization using a stereotactic frame. OBJECTIVE: We describe our protocol of frameless GKRS using DSA while maintaining high spatial resolution for precision. METHODS: This study is a retrospective review of patients with unruptured AVMs who underwent frameless GKRS. Magnetic resonance imaging and 3-dimensional DSA were obtained without a stereotactic frame for all patients. The imaging studies were merged for contouring of the AVM nidus. During GKRS treatment, patients were immobilized using an individually molded thermoplastic mask. RESULTS: Thirty-one patients were included in the analysis. The median age is 45.0 years (interquartile range [IQR]: 28.0-55.0). The median nidus size is 3.0 cm (IQR: 2.0-3.4). One patient had a Spetzler-Martin grade I, 11 had a grade II, 11 had a grade III, 6 had a grade IV, and 2 had a grade V AVM. Eleven patients underwent preradiosurgical embolization, 3 patients had previous microsurgical resection and/or embolization, and 1 patient had prior radiosurgery. The median administered dose was 20 Gy (IQR: 18.0-21.0). All patients completed their treatment with the planned radiation dose without complications. CONCLUSION: This is the first study that integrates DSA in the treatment planning of brain AVMs using GKRS without utilizing a stereotactic head frame. Frameless GKRS provides numerous advantages over frame-based techniques including improved patient experience and the capability of fractionation and thus expanding the eligibility of more AVMs for radiosurgery, while maintaining high spatial resolution of the AVM using angiography data.

Patient and procedure selection for mechanical thrombectomy: Toward personalized medicine and the role of artificial intelligence.

Mechanical thrombectomy (MT) for ischemic stroke due to large vessel occlusion is standard of care. Evidence-based guidelines on eligibility for MT have been outlined and evidence to extend the treatment benefit to more patients, particularly those at the extreme ends of a stroke clinical severity spectrum, is currently awaited. As patient selection continues to be explored, there is growing focus on procedure selection including the tools and techniques of thrombectomy and associated outcomes. Artificial intelligence (AI) has been instrumental in the area of patient selection for MT with a role in diagnosis and delivery of acute stroke care. Machine learning algorithms have been developed to detect cerebral ischemia and early infarct core, presence of large vessel occlusion, and perfusion deficit in acute ischemic stroke. Several available deep learning AI applications provide ready visualization and interpretation of cervical and cerebral arteries. Further enhancement of AI techniques to potentially include automated vessel probe tools in suspected large vessel occlusions is proposed. Value of AI may be extended to assist in procedure selection including both the tools and technique of thrombectomy. Delivering personalized medicine is the wave of the future and tailoring the MT treatment to a stroke patient is in line with this trend.

Minimally Invasive Sacroiliac Fusion-a Review.

PURPOSE OF REVIEW: Sacroiliac dysfunction is an important cause of low back pain with significant impact on quality of life and daily activities. Minimally invasive sacroiliac joint fusion (MIS SI fusion) is an effective treatment for patients who failed non-surgical strategies. The purpose of this article is to review the clinical outcomes and complications of this surgical technique. RECENT FINDINGS: For patients with SI joint dysfunction, MIS SI fusion reduced pain and disability as measured by Visual Analog Scale and Oswestry Index and improved quality of life as measured by Short-Form 36 and EuroQol-5D questionnaires. Satisfaction rates were higher in the SI fusion group when compared to the conservative management. In recent clinical trials, adverse events occurred with a similar rate in the first 6 months for patients assigned in the conservative management versus patients assigned to MIS SI fusion. MIS SI fusion is an effective and safe procedure for patients with sacroiliac dysfunction who failed non-surgical strategies. This procedure provides rapid as well as sustained pain relief, improvement in back function, high patient satisfaction, with low rate of complications.

Robot-assisted carotid artery stenting: outcomes, safety, and operational learning curve.

OBJECTIVE: Over the past 2 decades, robots have been increasingly used in surgeries to help overcome human limitations and perform precise and accurate tasks. Endovascular robots were pioneered in interventional cardiology, however, the CorPath GRX was recently approved by the FDA for peripheral vascular and extracranial interventions. The authors aimed to evaluate the operational learning curve for robot-assisted carotid artery stenting over a period of 19 months at a single institution. METHODS: A retrospective analysis of a prospectively maintained database was conducted, and 14 consecutive patients who underwent robot-assisted carotid artery stenting from December 2019 to June 2021 were identified. The metrics for proficiency were the total fluoroscopy and procedure times, contrast volume used, and radiation dose. To evaluate operator progress, the patients were divided into 3 groups of 5, 4, and 5 patients based on the study period. RESULTS: A total of 14 patients were included. All patients received balloon angioplasty and stent placement. The median degree of stenosis was 95%. Ten patients (71%) were treated via the transradial approach and 4 patients (29%) via the transfemoral approach, with no procedural complications. The median contrast volume used was 80 mL, and the median radiation dose was 38,978.5 mGy/cm2. The overall median fluoroscopy and procedure times were 24.6 minutes and 70.5 minutes, respectively. Subgroup analysis showed a significant decrease in these times, from 32 minutes and 86 minutes, respectively, in group 1 to 21.9 minutes and 62 minutes, respectively, in group 3 (p = 0.002 and p = 0.008, respectively). CONCLUSIONS: Robot-assisted carotid artery stenting was found to be safe and effective, and the learning curve for robotic procedures was overcome within a short period of time at a high-volume cerebrovascular center.

Predictors of aneurysm occlusion following treatment with the WEB device: systematic review and case series.

The Woven EndoBridge (WEB) device is becoming increasingly popular for treatment of wide-neck aneurysms. As experience with this device grows, it is important to identify factors associated with occlusion following WEB treatment to guide decision making and screen patients at high risk for recurrence. The aim of this study was to identify factors associated with adequate aneurysm occlusion following WEB device treatment in the neurosurgical literature and in our case series. A systematic review of the present literature was conducted to identify studies related to the prediction of WEB device occlusion. In addition, a retrospective review of our institutional data for patients treated with the WEB device was performed. Demographics, aneurysm characteristics, procedural variables, and 6-month follow-up angiographic outcomes were recorded. Seven articles totaling 450 patients with 456 aneurysms fit our criteria. Factors in the literature associated with inadequate occlusion included larger size, increased neck width, partial intrasaccular thrombosis, irregular shape, and tobacco use. Our retrospective review identified 43 patients with 45 aneurysms. A total of 91.1% of our patients achieved adequate occlusion at a mean follow-up time of 7.32 months. Increasing degree of contrast stasis after WEB placement on the post-deployment angiogram was significantly associated with adequate occlusion on follow-up angiogram (p = 0.005) and with Raymond-Roy classification (p = 0.048), but not with retreatment (p = 0.617). In our systematic review and case series totaling 450 patients with 456 aneurysms, contrast stasis on post-deployment angiogram was identified as a predictor of adequate aneurysm occlusion, while morphological characteristics such as larger size and wide neck negatively impact occlusion.

Glioblastoma with deep supratentorial extension is associated with a worse overall survival.

Glioblastoma (GBM) with deep-supratentorial extension (DSE) involving the thalamus, basal ganglia and corpus collosum, poses significant challenges for clinical management. In this study, we present our outcomes in patients who underwent resection of supratentorial GBM with associated involvement of deep brain structures. We conducted a retrospective review of patients who underwent resection of GBM at our institution between 2012 and 2018. A total of 419 patients were included whose pre-operative MRI scans were reviewed. Of these, 143 (34.1%) had GBM with DSE. There were similar rates of IDH-1 mutation (9% versus 7.6%, p = 0.940) and MGMT methylation status (35.7% versus 45.2%, p = 0.397) between the two cohorts. GBM patients without evidence of DSE had higher rates of radiographic gross total resection (GTR) compared to those with DSE: 70.6% versus 53.1%, respectively (p = 0.002). The presence of DSE was not associated with decreased progression-free survival (PFS) compared to patients without DSE (mean 7.24 ± 0.97 versus 8.89 ± 0.76 months, respectively; p = 0.276), but did portend a worse overall survival (OS) (mean 10.55 ± 1.04 versus 15.02 ± 1.05 months, respectively; p = 0.003). There was no difference in PFS or OS amongst DSE and non-DSE patients who underwent GTR, but patients who harbored DSE and underwent subtotal resection had worse OS (mean 8.26 ± 1.93 versus 12.96 ± 1.59 months, p = 0.03). Our study shows that GBM patients with DSE have lower OS compared to those without DSE. This survival difference appears to be primarily related to the limited surgical extent of resection owing to the neurological deficits that may be incurred with involvement of eloquent deep brain structures.

Comparison of Anesthetic Agents Dexmedetomidine and Midazolam During Mechanical Thrombectomy.

OBJECTIVES: The ideal anesthetic for mechanical thrombectomy (MT) is a subject of debate. Recent studies have supported the use of monitored anesthesia care (MAC), but few have attempted to compare MAC neuroanesthetics. Our study directly compares midazolam and dexmedetomidine (DEX) on blood pressure control during thrombectomy and functional outcomes at discharge. MATERIALS AND METHODS: We performed a retrospective review of an MT database, which consisted of 612 patients admitted between 2010-2019 to our tertiary stroke center. 193 patients who received either midazolam or DEX for MAC induction were identified. Primary and secondary outcomes were >20% maximum decrease in mean arterial pressure during MT and functional independence respectively. RESULTS: 146 patients were administered midazolam, while 47 were administered DEX. Decrease in blood pressure (BP) during MT was associated with lower rates of functional independence at last follow-up (p=0.034). When compared to midazolam, DEX had significantly higher rates of intraprocedural decrease in MAP at the following cut-offs: >20% (p<0.001), >30% (p=0.001), and >40% (p=0.006). On multivariate analysis, DEX was an independent predictor of >20% MAP decrease (OR 7.042, p<0.001). At time of discharge, NIHSS scores and functional independence (mRS 0-2) were statistically similar between DEX and midazolam. Functional independence at last known follow-up was statistically similar between DEX and midazolam (p = 0.643). CONCLUSIONS: Use of DEX during MT appears to be associated with increased blood pressure volatility when compared to midazolam. Further investigation is needed to determine the impact of MAC agents on functional independence.

Decompressive Hemicraniectomy in the Modern Era of Mechanical Thrombectomy.

OBJECTIVES: We aim to determine the incidence of decompressive hemicraniectomy (DHC) in the modern era of mechanical thrombectomy techniques and improved revascularization outcomes. METHODS: We performed a retrospective analysis of 512 patients admitted with acute ischemic strokes with anterior circulation large-vessel occlusion that were treated by mechanical thrombectomy from 2010-2019. The primary endpoint was the need for surgical decompression. Secondary endpoints were infarct size, hemorrhagic conversion, and functional outcome at hospital discharge. RESULTS: Of the 512 patients, 18 (3.5%) underwent DHC at a median 2.0 days from stroke onset. The DHC group was significantly younger than the non-DHC group (P < 0.001), had worse reperfusion rates (P = 0.024) and larger infarct size (P < 0.001). Hemorrhagic conversion was more frequent in the DHC group but did not reach statistical significance (P = 0.08). From 2010-2015, 196 patients underwent a mechanical thrombectomy, 13 of whom (6.6%) required a DHC, while 316 patients underwent mechanical thrombectomy from 2016-2019 and only 5 patients required a DHC (1.6%; P = 0.002). Younger age (P < 0.001), urinary tract infection (P < 0.001) and increasing infarct size were significantly associated with needing a DHC. When controlling for other risk factors, higher thrombolysis in cerebral infarction score significantly reduced the need for decompressive hemicraniectomy (P = 0.004). CONCLUSIONS: This is one of the largest single-center experiences demonstrating that improved recanalization decreased the need for DHC without increasing the risk of hemorrhagic conversion.

Transvenous Onyx Embolization of a Type D Carotid-Cavernous-Fistula: Operative Video.

Carotid-cavernous fistulas (CCFs) are acquired pathologic shunts between the carotid circulation and the cavernous sinus that result in venous congestion.1 They often present with ocular symptoms, such as chemosis, proptosis, and blurry vision. Cranial nerve deficits and increased intraocular pressure are often seen on the neuro-ophthalmologic examination.2 If left untreated, they can lead to cortical venous reflux and intracranial hemorrhage. A cerebral angiogram is the gold standard to diagnose these lesions. The hallmark of dural CCF is opacification of venous structures in the arterial phase of the angiogram. Dependent on carotid branches contributing to the fistula, 4 types are classically defined by Barrow et al.3 When the fistula is indirect (types B-D), the goal of treatment is obliteration via the transvenous route.4 We present the case of a patient who had chemosis and proptosis of the left eye with imaging findings concerning for dural CCF (Video 1). An informed consent was obtained and the patient underwent a cerebral angiogram and treatment of the CCF. In the operative video, we showcase the treatment of a type D CCF using transvenous embolization with Onyx (Covidien, Irvine, CA) and achieve angiographic cure of the fistula. We were able to use Onyx for embolization since the superselective injection did not show cortical venous drainage. This is important as obliteration of cortical veins with liquid embolisate could cause venous infarcts. To our knowledge, this is the first video article that illustrates the endovascular embolization of a CCF and highlights the angiographic findings pre- and post-embolization.

Alternate Seizure Spread with Agenesis of the Corpus Callosum.

Agenesis of the corpus callosum is a brain malformation that can occur in isolation or in conjunction with other congenital or developmental defects. The clinical sequelae of this condition include epilepsy, cognitive deficits, developmental delay, and various neurological and psychiatric disorders. Here we present the case of a patient with congenital complete agenesis of the corpus callosum and medically refractory epilepsy who underwent stereoelectroencephalography. This identified a left frontal ictal focus and revealed that contralateral spread occurred though the anterior commissure, a rare and interesting occurrence. Left frontal resection resulted in significant improvement. This demonstrates the role of the anterior commissure in ictal spread and the potential for novel methods of seizure spread in patients with temporal lobe epilepsy that must be considered in a surgical approach.

Operative versus Nonoperative Management of Idiopathic Spinal Cord Herniation: Effect on Symptomatology and Disease Progression.

BACKGROUND: Idiopathic spinal cord herniation (ISCH) is a rare pathology characterized by extravasation of the spinal cord through a dural defect. The optimal algorithm for choosing operative or nonoperative management is not well elucidated, partially because of the rarity of this pathology. We present the largest single-center series of ISCH and compare operative treatment to conservative management. METHODS: A retrospective case series of all patients evaluated for treatment of ISCH at our institution between 2010 and 2019 was conducted. Demographic variables, presenting symptoms, and imaging characteristics were assessed for all patients. For patients who underwent operative treatment, surgical approach, postoperative course, and discharge outcomes were recorded. Follow-up notes were reviewed for status of symptoms and functional capabilities, which were synthesized into Odom's criteria score. RESULTS: Sixteen patients met the inclusion criteria for this study, 8 of whom underwent operative treatment. No significant differences were found between operative and nonoperative groups with regard to demographic variables or pathology characteristics. Odom's criteria scores for the operative cohort were 12.5% (1 of 8) Excellent, 62.5% (5 of 8) Good, 12.5% (1 of 8) Fair, and 12.5% (1 of 8) Poor. Odom's criteria scores for the nonoperative cohort were 16.7% (1 of 6) Excellent, 33.3% (2 of 6) Good, 16.7% (1 of 6) Fair, and 33.3% (2 of 6) Poor. There was no significant difference between Odom's criteria score distribution between the operative and nonoperative groups at latest follow up (P = 0.715). CONCLUSIONS: Conservative management of spinal cord herniation is an option that does not preclude symptomatic improvement in patients with idiopathic spinal cord herniation.

Discrepancies in Stroke Distribution and Dataset Origin in Machine Learning for Stroke.

BACKGROUND: Machine learning algorithms depend on accurate and representative datasets for training in order to become valuable clinical tools that are widely generalizable to a varied population. We aim to conduct a review of machine learning uses in stroke literature to assess the geographic distribution of datasets and patient cohorts used to train these models and compare them to stroke distribution to evaluate for disparities. AIMS: 582 studies were identified on initial searching of the PubMed database. Of these studies, 106 full texts were assessed after title and abstract screening which resulted in 489 papers excluded. Of these 106 studies, 79 were excluded due to using cohorts from outside the United States or being review articles or editorials. 27 studies were thus included in this analysis. SUMMARY OF REVIEW: Of the 27 studies included, 7 (25.9%) used patient data from California, 6 (22.2%) were multicenter, 3 (11.1%) were in Massachusetts, 2 (7.4%) each in Illinois, Missouri, and New York, and 1 (3.7%) each from South Carolina, Washington, West Virginia, and Wisconsin. 1 (3.7%) study used data from Utah and Texas. These were qualitatively compared to a CDC study showing the highest distribution of stroke in Mississippi (4.3%) followed by Oklahoma (3.4%), Washington D.C. (3.4%), Louisiana (3.3%), and Alabama (3.2%) while the prevalence in California was 2.6%. CONCLUSIONS: It is clear that a strong disconnect exists between the datasets and patient cohorts used in training machine learning algorithms in clinical research and the stroke distribution in which clinical tools using these algorithms will be implemented. In order to ensure a lack of bias and increase generalizability and accuracy in future machine learning studies, datasets using a varied patient population that reflects the unequal distribution of stroke risk factors would greatly benefit the usability of these tools and ensure accuracy on a nationwide scale.

Long-segment posterior cervical decompression and fusion: does caudal level affect revision rate?

OBJECTIVE: Posterior cervical decompression and fusion (PCDF) is a commonly performed procedure to address cervical myelopathy. A significant number of these patients require revision surgery for adjacent-segment disease (ASD) or pseudarthrosis. Currently, there is no consensus among spine surgeons on the inclusion of proximal thoracic spine instrumentation. This study investigates the benefits of thoracic extension in long-segment cervical fusions and the potential drawbacks. The authors compare outcomes in long-segment subaxial cervical fusion for degenerative cervical myelopathy with caudal vertebral levels of C6, C7, and T1. METHODS: A retrospective analysis identified 369 patients who underwent PCDF. Patients were grouped by caudal fusion level. Reoperation rates for ASD and pseudarthrosis, infection, and blood loss were examined. Data were analyzed with chi-square, 1-way ANOVA, and logistic regression. RESULTS: The total reoperation rate for symptomatic pseudarthrosis or ASD was 4.8%. Reoperation rates, although not significant, were lower in the C3-6 group (2.6%, vs 8.3% for C3-7 and 3.8% for C3-T1; p = 0.129). Similarly, rates of infection were lower in the shorter-segment fusion without achieving statistical significance (2.6% for C3-6, vs 5.6% for C3-7 and 5.5% for C3-T1; p = 0.573). The mean blood loss was documented as 104, 125, and 224 mL for groups 1, 2, and 3, respectively (p < 0.001). CONCLUSIONS: Given the lack of statistical difference in reoperation rates for long-segment cervical fusions ending at C6, C7, or T1, shorter fusions in high-risk surgical candidates or elderly patients may be performed without higher rates of reoperation.

A Machine Learning Approach to First Pass Reperfusion in Mechanical Thrombectomy: Prediction and Feature Analysis.

INTRODUCTION: Novel machine learning (ML) methods are being investigated across medicine for their predictive capabilities while boasting increased adaptability and generalizability. In our study, we compare logistic regression with machine learning for feature importance analysis and prediction in first-pass reperfusion. METHODS: We retrospectively identified cases of ischemic stroke treated with mechanical thrombectomy (MT) at our institution from 2012-2018. Significant variables used in predictive modeling were demographic characteristics, medical history, admission NIHSS, and stroke characteristics. Outcome was binarized TICI on first pass (0-2a vs 2b-3). Shapley feature importance plots were used to identify variables that strongly affected outcomes. RESULTS: Accuracy for the Random Forest and SVM models were 67.1% compared to 65.8% for the logistic regression model. Brier score was lower for the Random Forest model (0.329 vs 0.342) indicating better predictive capability. Other supervised learning models performed worse than the logistic regression model, with accuracy of 56.2% for Naïve Bayes and 61.6% for XGBoost. Shapley plots for the Random Forest model showed use of aspiration, hyperlipidemia, hypertension, use of stent retriever, and time between symptom onset and catheterization as the top five predictors of first pass reperfusion. CONCLUSION: Use of machine learning models, such as Random Forest, for the study of MT outcomes, is more accurate than logistic regression for our dataset, and identifies new factors that contribute to achieving first pass reperfusion. The benefits of machine learning, such as improved predictive capabilities, integration of new data, and generalizability, establish ML as the preferred model for studying outcomes in stroke.