Prediction of hematoma expansion in spontaneous intracerebral hemorrhage using a multimodal neural network.

Hematoma expansion occasionally occurs in patients with intracerebral hemorrhage (ICH), associating with poor outcome. Multimodal neural networks incorporating convolutional neural network (CNN) analysis of images and neural network analysis of tabular data are known to show promising results in prediction and classification tasks. We aimed to develop a reliable multimodal neural network model that comprehensively analyzes CT images and clinical variables to predict hematoma expansion. We retrospectively enrolled ICH patients at four hospitals between 2017 and 2021, assigning patients from three hospitals to the training and validation dataset and patients from one hospital to the test dataset. Admission CT images and clinical variables were collected. CT findings were evaluated by experts. Three types of models were developed and trained: (1) a CNN model analyzing CT images, (2) a multimodal CNN model analyzing CT images and clinical variables, and (3) a non-CNN model analyzing CT findings and clinical variables with machine learning. The models were evaluated on the test dataset, focusing first on sensitivity and second on area under the receiver operating curve (AUC). Two hundred seventy-three patients (median age, 71 years [59-79]; 159 men) in the training and validation dataset and 106 patients (median age, 70 years [62-82]; 63 men) in the test dataset were included. Sensitivity and AUC of a CNN model were 1.000 (95% confidence interval [CI] 0.768-1.000) and 0.755 (95% CI 0.704-0.807); those of a multimodal CNN model were 1.000 (95% CI 0.768-1.000) and 0.799 (95% CI 0.749-0.849); and those of a non-CNN model were 0.857 (95% CI 0.572-0.982) and 0.733 (95% CI 0.625-0.840). We developed a multimodal neural network model incorporating CNN analysis of CT images and neural network analysis of clinical variables to predict hematoma expansion in ICH. The model was externally validated and showed the best performance of all the models.

Three-dimensional Images Fusion Method Useful for Preoperative Simulation of Clipping Surgery for Cerebral Aneurysms.

This study aimed to introduce a three-dimensional (3D) images fusion method for preoperative simulation of aneurysm clipping. Consecutive unruptured aneurysm cases treated with surgical clipping from March 2021 to October 2023 were included. In all cases, preoperative images of plain computed tomography (CT), CT angiography, magnetic resonance imaging (MRI) 3D fluid-attenuated inversion recovery, 3D heavily T2-weighted images, and 3D rotational angiography were acquired and transported into a commercial software (Ziostation2 Plus, Ziosoft, Inc. Tokyo, Japan). The software provided 3D images of skull, arteries including aneurysms, veins, and brain tissue that were freely rotated, magnified, trimmed, and superimposed. Using the 3D images fusion method, two operators predicted clips to be used in the following surgery. The predicted clips and actually used ones were compared to give agreement scores for the following factors: (1) type of clips (simple or fenestrated), (2) shape of clips (straight, curved, angled, or bayonet), and (3) clipping strategy (single or multiple). The agreement score ranged from 0 to 3 because a score of 1 or 0 was given for agreement or disagreement on each factor. Interoperator reproducibility was also evaluated. During the study period, 44 aneurysms from 37 patients were clipped. All procedures were successfully completed, thanks to the precisely reproduced surgical corridors with the 3D images fusion method. Agreement in clip prediction was good with mean agreement score of 2.4. Interobserver reproducibility was also high with the kappa value of 0.79. The 3D images fusion method was useful for preoperative simulation of aneurysm clipping.

Effects of New-Generation Antiepileptic Drug Prophylaxis on Delayed Neurovascular Events After Aneurysmal Subarachnoid Hemorrhage.

Neuroelectric disruptions such as seizures and cortical spreading depolarization may contribute to the development of delayed cerebral ischemia (DCI) after aneurysmal subarachnoid hemorrhage (SAH). However, effects of antiepileptic drug prophylaxis on outcomes remain controversial in SAH. The authors investigated if prophylactic administration of new-generation antiepileptic drugs levetiracetam and perampanel was beneficial against delayed neurovascular events after SAH. This was a retrospective single-center cohort study of 121 consecutive SAH patients including 56 patients of admission World Federation of Neurological Surgeons grades IV - V who underwent aneurysmal obliteration within 72 h post-SAH from 2013 to 2021. Prophylactic antiepileptic drugs differed depending on the study terms: none (2013 - 2015), levetiracetam for patients at high risks of seizures (2016 - 2019), and perampanel for all patients (2020 - 2021). The 3rd term had the lowest occurrence of delayed cerebral microinfarction on diffusion-weighted magnetic resonance imaging, which was related to less development of DCI. Other outcome measures were similar among the 3 terms including incidences of angiographic vasospasm, computed tomography-detectable delayed cerebral infarction, seizures, and 3-month good outcomes (modified Rankin Scale 0 - 2). The present study suggests that prophylactic administration of levetiracetam and perampanel was not associated with worse outcomes and that perampanel may have the potential to reduce DCI by preventing microcirculatory disturbances after SAH. Further studies are warranted to investigate anti-DCI effects of a selective α-amino-3-hydroxy-5-methyl-4-isoxazole propionate receptor antagonist perampanel in SAH patients in a large-scale prospective study.

Glioblastoma with concomitant moyamoya vasculopathy in neurofibromatosis type 1: illustrative case.

BACKGROUND: In a case of concurrent glioblastoma and moyamoya vasculopathy, it is arduous to safely perform surgery because the brain is highly vulnerable and collaterals are sometimes well developed. In addition, radiotherapy carries a risk of aggravating moyamoya vasculopathy, and chemotherapeutic agents also have a risk of interfering with collateral development. OBSERVATIONS: A 48-year-old woman with neurofibromatosis type 1 was admitted because of left hemiparesis and hemispatial neglect. Brain imaging studies revealed a large mass with peripheral enhancement in the right frontal lobe and occlusion of the bilateral middle cerebral arteries with an abnormal vascular network at the base of the brain. Total tumor resection was performed, and the pathological diagnosis was isocitrate dehydrogenase-mutant glioblastoma. Radiotherapy with a total dose of 60 Gy was delivered with concurrent temozolomide, and thereafter six cycles of adjuvant temozolomide were given. Progression of moyamoya vasculopathy without symptoms was observed after the completion of each of radiotherapy and adjuvant temozolomide. LESSONS: The authors present the first adult case of glioblastoma with moyamoya vasculopathy. Careful consideration and attention should be given throughout treatment to avoiding moyamoya vasculopathy-related ischemic and hemorrhagic events. Although the patient did not exhibit neurological deterioration, progression of moyamoya vasculopathy occurred early after radiotherapy and continued thereafter.

Machine learning prediction of hematoma expansion in acute intracerebral hemorrhage.

To examine whether machine learning (ML) approach can be used to predict hematoma expansion in acute intracerebral hemorrhage (ICH) with accuracy and widespread applicability, we applied ML algorithms to multicenter clinical data and CT findings on admission. Patients with acute ICH from three hospitals (n = 351) and those from another hospital (n = 71) were retrospectively assigned to the development and validation cohorts, respectively. To develop ML predictive models, the k-nearest neighbors (k-NN) algorithm, logistic regression, support vector machines (SVMs), random forests, and XGBoost were applied to the patient data in the development cohort. The models were evaluated for their performance on the patient data in the validation cohort, which was compared with previous scoring methods, the BAT, BRAIN, and 9-point scores. The k-NN algorithm achieved the highest area under the receiver operating characteristic curve (AUC) of 0.790 among all ML models, and the sensitivity, specificity, and accuracy were 0.846, 0.733, and 0.775, respectively. The BRAIN score achieved the highest AUC of 0.676 among all previous scoring methods, which was lower than the k-NN algorithm (p = 0.016). We developed and validated ML predictive models of hematoma expansion in acute ICH. The models demonstrated good predictive ability, showing better performance than the previous scoring methods.

A case presenting as fatal subarachnoid hemorrhage due to segmental arterial mediolysis associated with Crohn's disease.

BACKGROUND: Segmental arterial mediolysis (SAM) is a rare arterial pathology and can cause rupture or dissection of the intracranial arterial wall. The etiology is unveiled, but vasospastic stimuli such as migraine are considered as a possible cause of SAM. We present the first case of subarachnoid hemorrhage (SAH) due to SAM associated with Crohn's disease and migraine, and discuss the possible contribution of Crohn's disease to the development of SAM besides migraine. CASE DESCRIPTION: A 33-year-old man with Crohn's disease, which had been treated with adalimumab, repeatedly underwent 3-tesla magnetic resonance (MR) imaging and angiography for severe headache due to migraine and the subsequent development of fatigue in the left arm and both legs. At 7 months after the last MR imaging studies showing no abnormalities, he had a sudden onset of severe SAH, which was caused by rupture or dissection of the terminal portion in the right internal carotid artery. As his brain-stem reflexes were absent, the patient was conservatively treated and died 6 days after the ictus. By postmortem histopathological examination, SAM was diagnosed as the cause of SAH. Vasa vasorum was also observed around the rupture point. CONCLUSIONS: Our case suggests that: 1) the formation of vasa vasorum may be an antecedent pathology for vessel rupture of the fragile arterial wall affected by SAM, and 2) vasospastic nature of both Crohn's disease and migraine may contribute to the development of intracranial SAM.