Early prediction of hematoma expansion (HE) following nontraumatic intracerebral hemorrhage (ICH) may inform preemptive therapeutic interventions. We sought to identify how accurately machine learning (ML) radiomics models predict HE compared with expert clinicians using head computed tomography (HCT). We used data from 900 study participants with ICH enrolled in the Antihypertensive Treatment of Acute Cerebral Hemorrhage 2 Study. ML models were developed using baseline HCT images, as well as admission clinical data in a training cohort (n = 621), and their performance was evaluated in an independent test cohort (n = 279) to predict HE (defined as HE by 33

To describe an incidence of Lemierre syndrome due to blastomycosis, a rarely implicated pathogen.

Introduction: Hematoma expansion (HE) in patients with intracerebral hemorrhage (ICH) is a key predictor of poor prognosis and potentially amenable to treatment. This study aimed to build a classification model to predict HE in patients with ICH using deep learning algorithms without using advanced radiological features. Methods: Data from the ATACH-2 trial (Antihypertensive Treatment of Acute Cerebral Hemorrhage) was utilized. Variables included in the models were chosen as per literature consensus on salient variables associated with HE. HE was defined as increase in either >33% or 6 mL in hematoma volume in the first 24 h. Multiple machine learning algorithms were employed using iterative feature selection and outcome balancing methods. 70% of patients were used for training and 30% for internal validation. We compared the ML models to a logistic regression model and calculated AUC, accuracy, sensitivity and specificity for the internal validation models respective models. Results: Among 1000 patients included in the ATACH-2 trial, 924 had the complete parameters which were included in the analytical cohort. The median [interquartile range (IQR)] initial hematoma volume was 9.93. mm(3) [5.03-18.17] and 25.2% had HE. The best performing model across all feature selection groups and sampling cohorts was using an artificial neural network (ANN) for HE in the testing cohort with AUC 0.702 [95% CI, 0.631-0.774] with 8 hidden layer nodes The traditional logistic regression yielded AUC 0.658 [95% CI, 0.641-0.675]. All other models performed with less accuracy and lower AUC. Initial hematoma volume, time to initial CT head, and initial SBP emerged as most relevant variables across all best performing models. Conclusion: We developed multiple ML algorithms to predict HE with the ANN classifying the best without advanced radiographic features, although the AUC was only modestly better than other models. A larger, more heterogenous dataset is needed to further build and better generalize the models.

Introduction: Hematoma expansion (HE) in patients with intracerebral hemorrhage (ICH) is a key predictor of poor prognosis and potentially amenable to treatment. Current clinical scoring models, radiological markers, and labs have limited ability to predict HE. This study aimed to create a more robust classification model to predict HE in patients with ICH using a deep learning algorithm, an artificial neural network (ANN). Methods: Data from the ATACH 2 trial (Antihypertensive Treatment of Acute Cerebral Hemorrhage) was utilized. Variables included in the models were chosen as per literature consensus on salient factors associated with HE and clinical expertise. HE was defined as increase in either >33% or 6mL in hematoma volume in the first 24 hours. A multi-layer feedforward ANN was trained using the back-propagation method to minimize the loss function with 5-fold cross validation. 80% of patients were used for training and 20% for testing. The ANN was compared to a logistic regression model (GLM). Given the imbalance in patients with HE, AUPRC (area under the precision-recall curve), recall and precision were calculated for the respective models. Results: Of the 963 patients in the study (mean age 62±13.0, 38.5% female), 31% had hematoma expansion. The median [interquartile range (IQR)] initial hematoma volume was 10.5mm 3 [5.16 - 20.25], median admission SBP [IQR] was 200 [184-217], median platelet count [IQR] was 213 [178-256], median INR [IQR] was 1.0 [0.9-1.0], and median admission GCS [IQR] was 15 [13-15]. The GLM model had AUPRC of 0.38 with recall of 0.60 and precision of 0.38 in the testing cohort. The ANN training model demonstrated AUPRC of 0.92 with recall of 0.72 with precision of 0.39 during testing. Initial ICH volume, platelets, INR, and GCS had the highest feature importance. Conclusion: We developed an ANN to predict HE in ICH patients performing with improved sensitivity and similar positive predictive value when compared to GLM. We support that ANN can capture complex associations between variables not attainable in classic regression models. This model may help identify patients at risk for HE who warrant careful monitoring and aggressive treatment upfront including those suitable for clinical trials for treatments.

This is a patient with multiple meningoencephaloceles which resulted in bacterial meningitis and subsequent status epilepticus. We identify impressive imaging findings demonstrating herniation of the meninges from nasal and bitemporal skull base defects possibly as a result of intracranial hypertension.

A 29-year-old man with no prior known medical diagnoses presented to the emergency department with 5 days of fevers, chills, vomiting, headache, and neck pain followed by voice changes, difficulty swallowing, and intractable hiccups. The patient denied vision changes, weakness, numbness, recent travel, new medications, or illicit drug use. He had no personal or family history of neurologic illness. On questioning concerning possible toxic exposure, he noted recent consumption of sushi and raw meat at a restaurant; a friend who had eaten with him also experienced gastrointestinal illness after that meal.

Introduction: While the thrombotic complications of COVID-19 have been described, there are limited data on its implications in hemorrhagic stroke. The clinical characteristics, underlying stroke mechanism, and outcomes in this group of patients are especially salient as empiric therapeutic anticoagulation becomes increasingly common in the treatment and prevention of thrombotic complications of COVID-19. Methods: We conducted a retrospective cohort study of patients with hemorrhagic stroke (both non-traumatic intracerebral hemorrhage and spontaneous non-aneurysmal subarachnoid hemorrhage) who were hospitalized between 3/1/20-5/15/20 at a NYC hospital system, during the coronavirus pandemic. We compared the demographic and clinical characteristics of patients with hemorrhagic stroke and COVID-19 to those without COVID-19 admitted to our hospital between 3/1/20-5/15/20 (contemporary controls) and 3/1/19-5/15/19 (historical controls), using Fischer’s exact test and non-parametric testing. We adjusted for multiple comparisons using the Bonferroni method. Results: During the study period, 19 out of 4071 (0.5%) patients who were hospitalized with COVID-19 had hemorrhagic stroke on imaging. Of all COVID-19 with hemorrhagic stroke, only 3 had non-aneurysmal SAH without intraparenchymal hemorrhage. Among hemorrhagic stroke and COVID-19 patients, coagulopathy was the most common etiology (73.7%); empiric anticoagulation was started in 89.5% vs 4.2% of contemporary and 10.0% of historical controls (both with p =<0.001). Compared to contemporary and historical controls, COVID-19 patients had higher initial NIHSS scores, INR, PTT and fibrinogen levels. These patients also had higher rates of in-hospital mortality [84.6% vs. 4.6%, p =<0.001]. Sensitivity analyses excluding patients with strictly subarachnoid hemorrhage yielded similar results. Conclusion: We observed an overall low rate of imaging-confirmed hemorrhagic stroke among patients hospitalized with COVID-19. Most hemorrhages in COVID-19 patients occurred in the setting of therapeutic anticoagulation and were associated with increased mortality. Further studies are needed to evaluate the safety and efficacy of therapeutic anticoagulation in COVID-19 patients.

Introduction: Emergency department observation units (ED-OU) allow patients with a suspected transient ischemic attack (TIA) an expedited workup without the need for a prolonged inpatient admission. Despite risk stratification scores and physician evaluation, however, the reliability in diagnosis of TIA remains poor, which may lead to unnecessary testing. This study aimed to identify and compare the diagnostic workup between patients with final diagnosis of true vascular events (TIA or minor stroke, TIAMS) versus nonischemic transient neurological attacks (NI-TNA) in suspected TIA patients admitted to an ED observation unit. Methods: A retrospective analysis was performed on consecutive patients who were admitted to an ED-OU at a single center for suspected TIA. All diagnostic testing obtained during observation stay was abstracted from chart review. Final discharge diagnosis was dichotomized to either TIAMS or NI-TNA. Standard statistical tests were used for comparison testing between the two groups with significance defined as p<0.05. Results: Of 186 suspected TIA patients admitted to an ED-OU, median ABCD2 score was 4 [IQR 3-4]. Final diagnosis was TIAMS in 85 (46%) patients and NI-TNA in 101 (54%) patients. A total of 182 (98%) patients had non-contrast head CT (NCHCT); 160 (86%) brain MRI; 117 (63%) extracranial vessel imaging; 116 (62%) transthoracic echocardiogram (TTE); and 108 (58%) intracranial vessel imaging. Assessing diagnostic work-up by final diagnosis, TTE (78% vs 40%, p<0.01), and extracranial imaging (75% vs 55%, p<0.01) were more common in patients with TIAMS. Restricted diffusion on MRI (27% vs. 2%, p<0.01) and abnormality on TTE (50% vs. 28% p=0.02) were more common in TIAMS patients. The overall rate of symptomatic stenosis was low: 1 patient had a symptomatic carotid and 4 patients had symptomatic intracranial stenosis. Conclusion: Extensive diagnostic testing is done on patients with suspected TIA admitted to ED-OU, with more studies acquired on patients with true ischemic events as compared to NI-TNA. As the use of ED-OUs increases, refinement of current diagnostic testing algorithms to reduce workup for cerebrovascular disease among patients with NI-TIA and among different ischemic stroke subtypes is warranted.

While the thrombotic complications of COVID-19 have been well described, there are limited data on clinically significant bleeding complications including hemorrhagic stroke. The clinical characteristics, underlying stroke mechanism, and outcomes in this particular subset of patients are especially salient as therapeutic anticoagulation becomes increasingly common in the treatment and prevention of thrombotic complications of COVID-19. We conducted a retrospective cohort study of patients with hemorrhagic stroke (both non-traumatic intracerebral hemorrhage and spontaneous non-aneurysmal subarachnoid hemorrhage) who were hospitalized between March 1, 2020, and May 15, 2020, within a major healthcare system in New York, during the coronavirus pandemic. Patients with hemorrhagic stroke on admission and who developed hemorrhage during hospitalization were both included. We compared the clinical characteristics of patients with hemorrhagic stroke and COVID-19 to those without COVID-19 admitted to our hospital system between March 1, 2020, and May 15, 2020 (contemporary controls), and March 1, 2019, and May 15, 2019 (historical controls). Demographic variables and clinical characteristics between the individual groups were compared using Fischer’s exact test for categorical variables and nonparametric test for continuous variables. We adjusted for multiple comparisons using the Bonferroni method. During the study period in 2020, out of 4071 patients who were hospitalized with COVID-19, we identified 19 (0.5%) with hemorrhagic stroke. Of all COVID-19 with hemorrhagic stroke, only three had isolated non-aneurysmal SAH with no associated intraparenchymal hemorrhage. Among hemorrhagic stroke in patients with COVID-19, coagulopathy was the most common etiology (73.7%); empiric anticoagulation was started in 89.5% of these patients versus 4.2% in contemporary controls (p ≤ .001) and 10.0% in historical controls (p ≤ .001). Compared to contemporary and historical controls, patients with COVID-19 had higher initial NIHSS scores, INR, PTT, and fibrinogen levels. Patients with COVID-19 also had higher rates of in-hospital mortality (84.6% vs. 4.6%, p ≤ 0.001). Sensitivity analyses excluding patients with strictly subarachnoid hemorrhage yielded similar results. We observed an overall low rate of imaging-confirmed hemorrhagic stroke among patients hospitalized with COVID-19. Most hemorrhages in patients with COVID-19 infection occurred in the setting of therapeutic anticoagulation and were associated with increased mortality. Further studies are needed to evaluate the safety and efficacy of therapeutic anticoagulation in patients with COVID-19.

Sunday, April 26April 14, 2020Free AccessIdentifying Predictors for Final Diagnosis of Ischemic Events in an Emergency Department Observation Unit (4985)Arooshi Kumar, Cen Zhang, Ava Liberman, Koto Ishida, Jose Torres, and Sara RostanskiAuthors Info & AffiliationsApril 14, 2020 issue94 (15_supplement)https://doi.org/10.1212/WNL.94.15_supplement.4985 Letters to the Editor

Monday, April 27April 14, 2020Free AccessDiagnostic Evaluation of Patients Admitted to Emergency Department Observation Unit for Suspected TIA (3969)Arooshi Kumar, Koto Ishida, Ava Liberman, Cen Zhang, Shadi Yaghi, Jose Torres, and Sara RostanskiAuthors Info & AffiliationsApril 14, 2020 issue94 (15_supplement)https://doi.org/10.1212/WNL.94.15_supplement.3969 Letters to the Editor

•A patient developed encephalopathy two months after a mild case of COVID-19.•MRI Brain demonstrated extensive white matter lesions juxtacortically and subcortically.•Workup for infectious, paraneoplastic, toxic process was negative.•MBP was elevated, prompting diagnosis of post-infectious ADEM following COVID-19.•COVID-19 patients without hypoxia may have post-infectious leukoencephalopathy.

Introduction: Transient neurologic events have high rates of diagnostic uncertainty. Emergency department observation units (ED-OU) allow an accelerated diagnostic work up for suspected transient ischemic attacks (TIAs). However, clinical decision support regarding which patients to admit to these units is lacking. This study aimed to identify clinical features that differentiate true ischemic events from nonischemic transient neurological attacks (NI-TNA) among patients admitted to an ED-OU for suspected TIA. Methods: A retrospective analysis was performed on consecutive patients admitted to the ED-OU at a single academic center for suspected TIA. Demographics, vascular risk factors, presenting symptoms, and details of the clinical presentation were abstracted from chart review. Final discharge diagnosis was dichotomized to either ischemic event (TIA or minor stroke, TIAMS) or NI-TNA based on the treating vascular neurologist’s final diagnosis. Standard statistical tests were used for comparison testing between the two groups. Significantly different factors with p<0.2 on univariate analysis were carried forward in a multivariable logistic regression model. Results: Of 186 consecutive patients, 101 (54%) had a final diagnosis of NI-TNA and 85 (46%) of TIAMS. The median population ABCD2 score was 4 [IQR 3-4]. On univariate analysis, older age (63 vs. 70, p<0.01), history of atrial fibrillation (AF) (12% vs. 26%, p=0.01), and facial weakness (5% vs. 14% p=0.03) were associated with TIAMS. Headache (24% vs. 12%, p=0.04) and symptom duration>60min (57% vs. 40%, p=0.02) were associated with NI-TNA. On multivariable analysis, only symptom duration>60 minutes predicted NI-TNA (OR 0.39, p=0.04) and only history of AF (OR 2.53, p=0.03) predicted TIAMS. Facial weakness was strongly predictive of TIAMS (OR 3.22, p=0.05), but not significant. Conclusion: We identified two clinical features that distinguished TIAMS from NI-TNA among patients admitted to an ED-OU for suspected TIA.These may be helpful in emergency room triage of TIAMS. Data from ED-OU can be used to identify factors associated with cerebral ischemia and improve current care pathways for patients with suspected TIA, so diagnostic evaluation is received in the most appropriate setting.

April 23, 2018April 10, 2018Free AccessAdmission white blood cell count is associated with stroke severity and outcome independent of acute infarct size (P2.236)Anne-Katrin Giese, Arooshi Kumar, Cathy Zhang, Sarah E. Nelson, Kelsey Shideler, Lisa Cloonan, Allison Kanakis, Kaitlin Fitzpatrick, Karen Furie, Fanny Herisson, Ona Wu, and Natalia RostAuthors Info & AffiliationsApril 10, 2018 issue90 (15_supplement)https://doi.org/10.1212/WNL.90.15_supplement.P2.236 Letters to the Editor

Introduction: Drip and ship (DS) thrombolysis provides immediate acute ischemic stroke treatment and follow-up tertiary stroke care at a certified Comprehensive Stroke Center (CSC) for one quarter of the US population living in rural areas. Studies reveal that patients with stroke mimic (SM) inadvertently receive treatment due to lack of immediate access to specialists and the limited treatment time window. Hypothesis: We hypothesized that a higher percentage of SM would receive thrombolysis via the DS paradigm than those directly presenting to the CSC. Methods: We reviewed consecutive DS tPA cases transferred to the University of Louisville Hospital (ULH) and tPA cases originating at ULH from January, 2013 to June, 2015. ULH is a CSC that provides rural Kentucky and Southern Indiana hospitals with 24-hour telephone access to stroke specialists. We compared the percentage of SMs via a DS paradigm to those originating at ULH. SM data collected included demographics, medical history, NIHSS, complications, discharge diagnosis, discharge disposition, and the length of hospitalization. Etiology of SM was evaluated in the Old (≥ 65 years old) and the Young (<65 years old) group, respectively. Comparative analyses with t-tests and Fisher Exact tests were performed. Results: Total numbers of tPA cases were similar between the DS (201) and the ULH (200) groups, but the percentage of SM in the DS group was double the ULH group (27.4% vs 13.5%). Clinical features, NIHSS on admission, and percentage of SM patients who were 65 years or older were similar in both groups. None of SM had intracranial hemorrhage or severe adverse events. One patient in the DS had minor hematemesis without transfusion. All except one patient returned home or to an assisted living facility. One patient who was from home was discharged to a nursing home due to Parkinson’s disease. Psychiatric disease was more common in the Young SM than the old (45.3% vs 7.4%, p < 0.05). Encephalopathy for various reasons (25.9%) and seizures (22.2%) were the two most common causes in the Old SM. Conclusions: SMs are treated with tPA more often in the DS paradigm than when presenting to a CSC. Although the thrombolysis caused no harm, adequate access to specialists (i.e telestroke) may decrease unnecessary treatment with tPA.

Radiation necrosis (RN) is a serious complication that can occur in up to 10% of brain radiotherapy cases, with the incidence dependent on both dose and brain location. Available medical treatment for RN includes steroids, vitamin E, pentoxifylline, and hyperbaric oxygen. In a significant number of patients, however, RN is medically refractory and the patients experience progressive neurological decline, disabling headaches, and decreased quality of life.Vascular endothelial growth factor (VEGF) is a known mediator of cerebral edema in RN. Recent reports have shown successful treatment of RN with intravenous bevacizumab, a monoclonal antibody for VEGF. Bevacizumab, however, is associated with significant systemic complications including sinus thrombosis, pulmonary embolus, gastrointestinal tract perforation, wound dehiscence, and severe hypertension. Using lower drug doses may decrease systemic exposure and reduce complication rates. By using an intraarterial route for drug administration following blood-brain barrier disruption (BBBD), the authors aim to lower the bevacizumab dose while increasing target delivery.In the present report, the authors present the cases of 2 pediatric patients with cerebral arteriovenous malformations, who presented with medically intractable RN following stereotactic radiosurgery. They received a single intraarterial infusion of 2.5 mg/kg bevacizumab after hyperosmotic BBBD.At mean follow-up duration of 8.5 months, the patients had significant and durable clinical and radiographic response. Both patients experienced resolution of their previously intractable headaches and reversal of cushingoid features as they were successfully weaned off steroids. One of the patients regained significant motor strength. There was an associated greater than 70% reduction in cerebral edema.Intraarterial administration of a single low dose of bevacizumab after BBBD was safe and resulted in durable clinical and radiographic improvements at concentrations well below those required for the typical systemic intravenous route. Advantages over the intravenous route may include higher concentration of drug delivery to the affected brain, decreased systemic toxicity, and a significantly lower cost.