Search Results (6)

Background/Objectives: Recent studies reveal an “obesity paradox”, suggesting better clinical outcomes after intracranial hemorrhage for obese patients compared to patients with a healthy BMI. While this paradox indicates improved survival rates for obese individuals in stroke cases, it is unknown whether this trend remains true across all forms of intracranial hemorrhage. Therefore, the objective of our study was to investigate the incidence, characteristics, and outcomes of hospitalized obese patients with intracranial hemorrhage. Methods: The National Inpatient Sample (NIS) database was queried for data from 2015 to 2019 to identify adult patients aged 18 years and older with a primary diagnosis of non-traumatic intracranial hemorrhage. Using International Classification of Disease 10th Edition codes, patients were stratified by BMI categories: healthy weight, overweight, class I–II obesity, and class III obesity. The cohorts were examined for demographic characteristics, comorbidities, stroke severity, inpatient complications, interventions, and clinical outcomes, including length of stay (LOS), discharge disposition, and inpatient mortality. Results: Of 41,960 intracranial hemorrhage patients identified, 13,380 (33.0%) also had an obese BMI. Class I–II obese intracranial hemorrhage patients were more likely to be of white race (OR: 1.101, 95% CI: 1.052, 1.152, p < 0.001), less likely to be female (OR: 0.773, 95% CI: 0.740, 0.808, p < 0.001), and more likely to have diabetes mellitus (OR: 1.545, 95% CI: 1.477, 1.616, p < 0.001) and hypertension (OR: 1.828, 95% CI: 1.721, 1.943, p < 0.001) in comparison to healthy-weight patients. In a matched cohort analysis adjusting for demographics and severity, intracranial hemorrhage patients with class I–II obesity had a shorter length of stay (LOS) (OR 0.402, 95% CI: 0.118, 0.705, p < 0.001), reduced inpatient mortality (OR 0.847, 95% CI: 0.798, 0.898, p < 0.001), and more favorable discharge disposition (OR 1.395, 95% CI: 1.321, 1.474, p < 0.001) compared to their healthy-weight counterparts. Furthermore, these patients were less likely to require decompressive hemicraniectomy (OR 0.697, 95% CI: 0.593, 0.820, p < 0.001). Following an analysis of individual ICH subtypes, obese subarachnoid hemorrhage (SAH) patients demonstrated reduced mortality (OR: 0.692, 95% CI: 0.577–0.831, p < 0.001) and LOS (OR: 0.070, 95% CI: 0.466–0.660, p = 0.039), with no differences in discharge disposition. Similarly, intracerebral hemorrhage patients demonstrated reduced mortality (OR: 0.891, 95% CI: 0.827–0.959, p = 0.002) and LOS (OR: 0.480, 95% CI: 0.216–0.743, p < 0.001). Other ICH subtypes showed improved discharge outcomes (OR: 1.504, 95% CI: 1.368–1.654, p < 0.001), along with decreased mortality (OR: 0.805, 95% CI: 0.715–0.907, p < 0.001) and LOS (OR: −10.313, 95% CI: −3.599 to −2.449, p < 0.001). Conclusions: Intracranial hemorrhage patients with class I–II obesity exhibited more favorable clinical outcomes than those who were of a healthy weight or overweight. Despite its association with risk factors contributing to intracranial hemorrhage, class I–II obesity was associated with improved outcomes, lending support to the existence of the obesity paradox in stroke. Full article

Primary Angiitis of the Central Nervous System (PACNS) is a rare cerebrovascular disease involving the arteries of the leptomeninges, brain and spinal cord. Its diagnosis can be challenging, and the current diagnostic criteria show several limitations. Among the clinical and neuroimaging manifestations of PACNS, intracranial bleeding, particularly intracerebral hemorrhage (ICH), is poorly described in the available literature, and it is considered infrequent. This review aims to summarize the available data addressing this issue with a dedicated focus on the clinical, neuroradiological and neuropathological perspectives. Moreover, the limitations of the actual data and the unanswered questions about hemorrhagic PACNS are addressed from a double point of view (PACNS subtyping and ICH etiology). Fewer than 20% of patients diagnosed as PACNS had an ICH during the course of the disease, and in cases where ICH was reported, it usually did not occur at presentation. As trigger factors, both sympathomimetic drugs and illicit drugs have been proposed, under the hypothesis of an inflammatory response due to vasoconstriction in the distal cerebral arteries. Most neuroradiological descriptions documented a lobar location, and both the large-vessel PACNS (LV-PACNS) and small-vessel PACNS (SV-PACNS) subtypes might be the underlying associated phenotypes. Surprisingly, amyloid beta deposition was not associated with ICH when histopathology was available. Moreover, PACNS is not explicitly included in the etiological classification of spontaneous ICH. This issue has received little attention in the past, and it could be addressed in future prospective studies. Full article

Intracranial hemorrhage (ICH) can lead to death or disability, which requires immediate action from radiologists. Due to the heavy workload, less experienced staff, and the complexity of subtle hemorrhages, a more intelligent and automated system is necessary to detect ICH. In literature, many artificial-intelligence-based methods are proposed. However, they are less accurate for ICH detection and subtype classification. Therefore, in this paper, we present a new methodology to improve the detection and subtype classification of ICH based on two parallel paths and a boosting technique. The first path employs the architecture of ResNet101-V2 to extract potential features from windowed slices, whereas Inception-V4 captures significant spatial information in the second path. Afterwards, the detection and subtype classification of ICH is performed by the light gradient boosting machine (LGBM) using the outputs of ResNet101-V2 and Inception-V4. Thus, the combined solution, known as ResNet101-V2, Inception-V4, and LGBM (Res-Inc-LGBM), is trained and tested over the brain computed tomography (CT) scans of CQ500 and Radiological Society of North America (RSNA) datasets. The experimental results state that the proposed solution efficiently obtains 97.7% accuracy, 96.5% sensitivity, and 97.4% F1 score using the RSNA dataset. Moreover, the proposed Res-Inc-LGBM outperforms the standard benchmarks for the detection and subtype classification of ICH regarding the accuracy, sensitivity, and F1 score. The results prove the significance of the proposed solution for its real-time application. Full article

Generally, traumatic and aneurysmal brain injuries cause intracranial hemorrhages, which is a severe disease that results in death, if it is not treated and diagnosed properly at the early stage. Compared to other imaging techniques, Computed Tomography (CT) images are extensively utilized by clinicians for locating and identifying intracranial hemorrhage regions. However, it is a time-consuming and complex task, which majorly depends on professional clinicians. To highlight this problem, a novel model is developed for the automatic detection of intracranial hemorrhages. After collecting the 3D CT scans from the Radiological Society of North America (RSNA) 2019 brain CT hemorrhage database, the image segmentation is carried out using Fuzzy C Means (FCM) clustering algorithm. Then, the hybrid feature extraction is accomplished on the segmented regions utilizing the Histogram of Oriented Gradients (HoG), Local Ternary Pattern (LTP), and Local Binary Pattern (LBP) to extract discriminative features. Furthermore, the Cuckoo Search Optimization (CSO) algorithm and the Optimized Gated Recurrent Unit (OGRU) classifier are integrated for feature selection and sub-type classification of intracranial hemorrhages. In the resulting segment, the proposed ORGU-CSO model obtained 99.36% of classification accuracy, which is higher related to other considered classifiers. Full article

In this paper, we present our system for the RSNA Intracranial Hemorrhage Detection challenge, which is based on the RSNA 2019 Brain CT Hemorrhage dataset. The proposed system is based on a lightweight deep neural network architecture composed of a convolutional neural network (CNN) that takes as input individual CT slices, and a Long Short-Term Memory (LSTM) network that takes as input multiple feature embeddings provided by the CNN. For efficient processing, we consider various feature selection methods to produce a subset of useful CNN features for the LSTM. Furthermore, we reduce the CT slices by a factor of $2\times$, which enables us to train the model faster. Even if our model is designed to balance speed and accuracy, we report a weighted mean log loss of $0.04989$ on the final test set, which places us in the top 30 ranking (2%) from a total of 1345 participants. While our computing infrastructure does not allow it, processing CT slices at their original scale is likely to improve performance. In order to enable others to reproduce our results, we provide our code as open source. After the challenge, we conducted a subjective intracranial hemorrhage detection assessment by radiologists, indicating that the performance of our deep model is on par with that of doctors specialized in reading CT scans. Another contribution of our work is to integrate Grad-CAM visualizations in our system, providing useful explanations for its predictions. We therefore consider our system as a viable option when a fast diagnosis or a second opinion on intracranial hemorrhage detection are needed. Full article

Women are known to have particular heterogeneity in stroke etiology related to childbearing and hormonal factors. Although there are continued acute stroke treatment advances focusing on clot dissolution or extraction, effective secondary prevention of stroke, however, is dependent on an accurate etiological determination of the stroke. Otherwise, more strokes are likely to follow. Analysis of young women’s stroke etiology in a large stroke registry incorporating contemporary neurovascular and parenchymal imaging and cardiac imaging. Young people (18-49 years old) with stroke were consecutively accrued over a 4 year period and an investigative protocol prospectively applied that incorporated multimodality magnetic resonance imaging, angiography, cardiac echo and stroke relevant blood investigations. All patients were classified according to an expanded Trial of Org 10172 in Acute Stroke Treatment − TOAST − classification and neurological deficit by the National Institute of Health stroke admission scores. In 511 registry derived, young stroke patients (mean age 39.8 years, 95% confidence interval: 39.1; 40.7 years), gender (women n=269, 53%) the etiological categories (women; men) included: i) small vessel disease (30/55;25/55), ii) cardioembolic (16/42;26/42), iii) large vessel cervical and intracranial disease (24/43;19/43), the other category (132/226; 91/226), which included, iv) substance abuse (15/41; 26/41, 4.6), v) prothrombotic states (22/37;15/37), vi) dissection (11/30;19/30), vii) cerebral venous thrombosis (15/19; 4/19, 12.4), viii) vasculitis (8/12; 4/12), ix) migraine related (10/11, 1/11) and x) miscellaneous vasculopathy (38/52;14/52). The latter entities comprised of aortic arch atheroma, vessel redundancy syndrome, vertebrobasilar hypoplasia, arterial fenestrations and dolichoectasia. Some conditions occurred solely in women, such as eclampsia (5), Call Fleming syndrome (4), fibromuscular dysplasia (3) and Moya Moya syndrome (2). Categories aside from bland infarction included: ii) intracerebral hemorrhage (43/106; 63/106) and xiii) stroke of undetermined etiology (6/10; 4/10). Admission mean National Institute of Health Stroke Scale scores differed significantly between women and men (4.7; 6.0 t=1.8, P=0.03). Young women’s stroke is significantly different from men in 7/12 stroke etiological categories in addition to 4 unique subtypes that require specific management. Full article