Hounsfield Unit Value of Interpeduncular Cistern Hematomas Can Predict Symptomatic Vasospasm

Machine learning predicts cerebral vasospasm in patients with subarachnoid haemorrhage

BACKGROUND

Cerebral vasospasm (CV) is a feared complication which occurs after 20-40% of subarachnoid haemorrhage (SAH). It is standard practice to admit patients with SAH to intensive care for an extended period of resource-intensive monitoring. We used machine learning to predict CV requiring verapamil (CVRV) in the largest and only multi-center study to date.

METHODS

Patients with SAH admitted to UCLA from 2013 to 2022 and a validation cohort from VUMC from 2018 to 2023 were included. For each patient, 172 unique intensive care unit (ICU) variables were extracted through the primary endpoint, namely first verapamil administration or no verapamil. At each institution, a light gradient boosting machine (LightGBM) was trained using five-fold cross validation to predict the primary endpoint at various hospitalization timepoints.

FINDINGS

A total of 1750 patients were included from UCLA, 125 receiving verapamil. LightGBM achieved an area under the ROC (AUC) of 0.88 > 1 week in advance and ruled out 8% of non-verapamil patients with zero false negatives. Our models predicted "no CVRV" vs "CVRV within three days" vs "CVRV after three days" with AUCs = 0.88, 0.83, and 0.88, respectively. From VUMC, 1654 patients were included, 75 receiving verapamil. VUMC predictions averaged within 0.01 AUC points of UCLA predictions.

INTERPRETATION

We present an accurate and early predictor of CVRV using machine learning with multi-center validation. This represents a significant step towards optimized clinical management and resource allocation in patients with SAH.

FUNDING

Robert E. Freundlich is supported by National Center for Advancing Translational Sciences federal grant UL1TR002243 and National Heart, Lung, and Blood Institute federal grant K23HL148640; these funders did not play any role in this study. The National Institutes of Health supports Vanderbilt University Medical Center which indirectly supported these research efforts. Neither this study nor any other authors personally received financial support for the research presented in this manuscript. No support from pharmaceutical companies was received.

Red Blood Cells in the Cerebrospinal Fluid Compartment After Subarachnoid Haemorrhage: Significance and Emerging Therapeutic Strategies

Subarachnoid haemorrhage (SAH) is a subtype of stroke that predominantly impacts younger individuals. It is associated with high mortality rates and can cause long-term disabilities. This review examines the contribution of the initial blood load and the dynamics of clot clearance to the pathophysiology of SAH and the risk of adverse outcomes. These outcomes include hydrocephalus and delayed cerebral ischaemia (DCI), with a particular focus on the impact of blood located in the cisternal spaces, as opposed to ventricular blood, in the development of DCI. The literature described underscores the prognostic value of haematoma characteristics, such as volume, density, and anatomical location. The limitations of traditional radiographic grading systems are discussed, compared with the more accurate volumetric quantification techniques for predicting patient prognosis. Further, the significance of red blood cells (RBCs) and their breakdown products in secondary brain injury after SAH is explored. The review presents novel interventions designed to accelerate clot clearance or mitigate the effects of toxic byproducts released from erythrolysis in the cerebrospinal fluid following SAH. In conclusion, this review offers deeper insights into the complex dynamics of SAH and discusses the potential pathways available for advancing its management.

Machine Learning Predicts Cerebral Vasospasm in Subarachnoid Hemorrhage Patients

Background

Cerebral vasospasm (CV) is a feared complication occurring in 20-40% of patients following subarachnoid hemorrhage (SAH) and is known to contribute to delayed cerebral ischemia. It is standard practice to admit SAH patients to intensive care for an extended period of vigilant, resource-intensive, clinical monitoring. We used machine learning to predict CV requiring verapamil (CVRV) in the largest and only multi-center study to date.

Methods

SAH patients admitted to UCLA from 2013-2022 and a validation cohort from VUMC from 2018-2023 were included. For each patient, 172 unique intensive care unit (ICU) variables were extracted through the primary endpoint, namely first verapamil administration or ICU downgrade. At each institution, a light gradient boosting machine (LightGBM) was trained using five- fold cross validation to predict the primary endpoint at various timepoints during hospital admission. Receiver-operator curves (ROC) and precision-recall (PR) curves were generated.

Results

A total of 1,750 patients were included from UCLA, 125 receiving verapamil. LightGBM achieved an area under the ROC (AUC) of 0.88 an average of over one week in advance, and successfully ruled out 8% of non-verapamil patients with zero false negatives. Minimum leukocyte count, maximum platelet count, and maximum intracranial pressure were the variables with highest predictive accuracy. Our models predicted "no CVRV" vs "CVRV within three days" vs "CVRV after three days" with AUCs=0.88, 0.83, and 0.88, respectively. For external validation at VUMC, 1,654 patients were included, 75 receiving verapamil. Predictive models at VUMC performed very similarly to those at UCLA, averaging 0.01 AUC points lower.

Conclusions

We present an accurate (AUC=0.88) and early (>1 week prior) predictor of CVRV using machine learning over two large cohorts of subarachnoid hemorrhage patients at separate institutions. This represents a significant step towards optimized clinical management and improved resource allocation in the intensive care setting of subarachnoid hemorrhage patients.

Clinical Outcomes of Diffuse Angiogram-Negative Subarachnoid Hemorrhage Versus Aneurysmal Subarachnoid Hemorrhage: A Propensity Score-Matched Analysis

BACKGROUND

The outcome of diffuse angiogram-negative subarachnoid hemorrhage (dan-SAH) compared with aneurysmal SAH (aSAH) remains unclear. This study aimed to compare outcomes using propensity score matching.

METHODS AND RESULTS

Sixty-five patients with dan-SAH and 857 patients with aSAH admitted between January 2018 and December 2022 were retrospectively reviewed. Propensity score matching resulted in matching 65 patients with dan-SAH to 260 patients with aSAH, and clinical outcomes were compared between the groups. Compared with patients with dan-SAH, patients with aSAH were more likely to experience rehemorrhage (8.8% versus 0%, P=0.027), death (11.2% versus 1.5%; odds ratios [OR] 8.04 [95% CI, 1.07-60.12]; P=0.042), or delayed cerebral ischemia (12.3% versus 3.1%; OR, 4.42 [95% CI, 1.03-18.95]; P=0.045). Multivariate analysis revealed that Hunt-Hess grade 4 to 5 (OR, 3.13 [95% CI, 2.11-4.64]; P<0.001), presence of intraventricular hemorrhage (OR, 3.58 [95% CI, 1.72-7.46]; P=0.001), and smoking (OR, 2.44 [95% CI, 1.12-5.28]; P=0.024) were independently associated with the incidence of unfavorable outcomes (modified Rankin scale score >2 at 3 months), whereas dan-SAH was not (OR, 0.66 [95% CI, 0.25-1.73]; P=0.40).

CONCLUSIONS

Compared with patients with dan-SAH, patients with aSAH had higher rehemorrhage rates and in-hospital mortality, as well as a higher incidence of delayed cerebral ischemia. Unfavorable outcomes were associated with admission Hunt-Hess grade, the presence of intravenetricular hemorrhage, and smoking history, but there was no relation with the  pathogenesis of the hemorrhage (dan-SAH versus aSAH).

Development and external validation of a dynamic nomogram for delayed cerebral ischaemia after aneurysmal subarachnoid hemorrhage: a study protocol for a multicentre retrospective cohort study

INTRODUCTION

Delayed cerebral ischaemia (DCI) caused by aneurysmal subarachnoid haemorrhage (aSAH) is the most frequent complication and typically contributes to poor neurological outcome or deterioration of patients' condition. Therefore, early accurate and effective prediction of DCI is urgently needed. This study aims to construct a dynamic nomogram for precisely calculating the risk of DCI in patients with aSAH. Internal validation of this tool is conducted using the training cohort, and independent external validation is completed by using other medical centre datasets.

METHODS AND ANALYSIS

This study is a multicentre, retrospective, observational cohort study using data from patients with aSAH. The participants include all adult patients who received surgical treatment in neurosurgery of multiple medical centres from 1 September 2019 to 1 April 2021, including Renmin Hospital of Wuhan University, Huzhou Central Hospital, First Affiliated Hospital of Harbin Medical University, General Hospital of Northern Theatre Command and Affiliated Hospital of Panzhihua University. Clinical information is collected via the electronic medical record system, including demographic data, clinical state on admission and serum laboratory tests. Modified Fisher grade at admission, admission subarachnoid clot and cerebral oedema density, and residual postoperative subarachnoid clot density are determined using the electronic imagine record software. The primary outcome is DCI.

ETHICS AND DISSEMINATION

This study protocol was reviewed and approved by the Medical Ethics Committee of Renmin Hospital of Wuhan University, which is the principal affiliation of this study (approval number: WDRM2021-K022). The other Ethics Committees, including Huzhou Central Hospital (approval number: 202108005-01), First Affiliated Hospital of Harbin Medical University (approval number: H202156), General Hospital of Northern Theater Command (approval number: Y2021060) and Affiliated Hospital of Panzhihua University (approval number: 202105002), also approved the protocol. The results of this research will be published in a peer-reviewed medical journal.

TRIAL REGISTRATION NUMBER

ChiCTR2100044448.

Predictive effects of admission white blood cell counts and hounsfield unit values on delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage

OBJECTIVE

Neuroinflammatory response is deemed the primary pathogenesis of delayed cerebral ischemia (DCI) caused by aneurysmal subarachnoid hemorrhage (aSAH). Both white blood cell (WBC) count and Hounsfield Unit (HU) are gradually considered can reflect inflammation in DCI. This study aims to identify the relationship between WBC count and HU value and investigate the effects of both indicators in predicting DCI after aSAH.

METHODS

We enrolled 109 patients with aSAH admitted within 24 h of onset in our study. A multivariate logistic regression analysis was used to evaluate the admission WBC count, HU value, and combined WBC-HU associated with DCI. The receiver operating characteristic curve and area under the curve (AUC) were used to determine thresholds and detect the predictive ability of these predictors. These indicators were also compared with the established inflammation markers.

RESULTS

Thirty-six (33%) patients developed DCI. Both WBC count and HU value were strongly associated with the admission glucose level (ρ = .303, p = .001; ρ = .273, p = .004), World Federation of Neurosurgical Societies grade (ρ = .452, p < .001; ρ = .578; p < .001), Hunt-Hess grade (ρ = .450, p < .001; ρ = .510, p < .001), and modified Fisher scale score (ρ = .357, p < .001; ρ = .330, p < .001). After controlling these public variables, WBC count (ρ = .300, p = .002) positively correlated with HU value. An early elevated WBC (odds ratio [OR] 1.449, 95% confidence interval [CI]: 1.183-1.774, p < .001) count and HU value (OR 1.304, 95%CI: 1.149-1.479, p < .001) could independently predict the occurrence of DCI. However, only these patients with both WBC count and HU value exceeding the cut-off points (OR 36.89, 95%CI: 5.606-242.78, p < .001) were strongly correlated with DCI. Compared with a single WBC count (AUC 0.811, 95%CI: 0.729-0.892, p < .001) or HU value (AUC 0.869, 95%CI: 0.803-0.936, p < .001), the combined WBC-HU (AUC 0.898, 95%CI: 0.839-0.957, p < .001) demonstrated a better ability to predict the occurrence of DCI. Inspiringly, the prediction performance of these indicators outperformed the established inflammatory markers.

CONCLUSION

An early elevated WBC count and HU value could independently predict DCI occurrence between 4 and 30 days after aSAH. Furthermore, WBC count was positively correlated with HU value, and the combined WBC-HU demonstrated a superior prediction ability for DCI development compared with the individual indicator.

The influence of aneurysm morphology on the volume of hemorrhage after rupture

OBJECTIVE

Factors determining the risk of rupture of intracranial aneurysms have been extensively studied; however, little attention is paid to variables influencing the volume of bleeding after rupture. In this study the authors aimed to evaluate the impact of aneurysm morphological variables on the amount of hemorrhage.

METHODS

This was a retrospective cohort analysis of a prospectively collected data set of 116 patients presenting at a single center with subarachnoid hemorrhage due to aneurysmal rupture. A volumetric assessment of the total hemorrhage volume was performed from the initial noncontrast CT. Aneurysms were segmented and reproduced from the initial CT angiography study, and morphology indexes were calculated with a computer-assisted approach. Clinical and demographic characteristics of the patients were included in the study. Factors influencing the volume of hemorrhage were explored with univariate correlations, multiple linear regression analysis, and graphical probabilistic modeling.

RESULTS

The univariate analysis demonstrated that several of the morphological variables but only the patient's age from the clinical-demographic variables correlated (p < 0.05) with the volume of bleeding. Nine morphological variables correlated positively (absolute height, perpendicular height, maximum width, sac surface area, sac volume, size ratio, bottleneck factor, neck-to-vessel ratio, and width-to-vessel ratio) and two correlated negatively (parent vessel average diameter and the aneurysm angle). After multivariate analysis, only the aneurysm size ratio (p < 0.001) and the patient's age (p = 0.023) remained statistically significant. The graphical probabilistic model confirmed the size ratio and the patient's age as the variables most related to the total hemorrhage volume.

CONCLUSIONS

A greater aneurysm size ratio and an older patient age are likely to entail a greater volume of bleeding after subarachnoid hemorrhage.

Value of Hounsfield Units in the Evaluation of Isolated Sphenoid Sinus Lesions

Radiologic findings of fungal sinus disease are generally opacification in paranasal computed tomography (CT) images. The Hounsfield unit (HU) is a standardized objective unit that is also suitable for measuring remodeling and opacifications on CT scans of bone sections of patients with chronic rhinosinusitis. We hypothesized that HU values could provide valuable information in isolated sphenoid sinus lesions before surgery. Between 2012 and 2019, 35 patients underwent functional endoscopic sinus surgery for sphenoid sinus lesions. Tissues obtained from the sphenoid sinus were divided into two groups, fungal and nonfungal, according to the findings of histopathologic examinations. HU values were measured in sphenoid sinus sections on paranasal CT scans of these two groups. Differences in mean and maximum HU values between the two groups were statistically significant (p < .05). The maximum HU values calculated from the sphenoid sinus were 435.08 and 196.23 (p ≤ .05) in the fungal group and nonfungal group, respectively. The mean HU values calculated from the sphenoid sinus were 64.31 and 29 (p ≤ .05) in the fungal and nonfungal groups, respectively. At the maximum cutoff value of 241, the sensitivity and specificity of the HU maximum were 84.6% and 77.3%, respectively. At the mean cutoff value of 41.5, the sensitivity and specificity of the HU mean were 76.9% and 86.4%, respectively. HU is an objective value used in radiographic density measurement. The HU values were higher in fungal lesions than in nonfungal inflammations, and they are useful in preoperative measurement.

When the Blood Hits Your Brain: The Neurotoxicity of Extravasated Blood

Hemorrhage in the central nervous system (CNS), including intracerebral hemorrhage (ICH), intraventricular hemorrhage (IVH), and aneurysmal subarachnoid hemorrhage (aSAH), remains highly morbid. Trials of medical management for these conditions over recent decades have been largely unsuccessful in improving outcome and reducing mortality. Beyond its role in creating mass effect, the presence of extravasated blood in patients with CNS hemorrhage is generally overlooked. Since trials of surgical intervention to remove CNS hemorrhage have been generally unsuccessful, the potent neurotoxicity of blood is generally viewed as a basic scientific curiosity rather than a clinically meaningful factor. In this review, we evaluate the direct role of blood as a neurotoxin and its subsequent clinical relevance. We first describe the molecular mechanisms of blood neurotoxicity. We then evaluate the clinical literature that directly relates to the evacuation of CNS hemorrhage. We posit that the efficacy of clot removal is a critical factor in outcome following surgical intervention. Future interventions for CNS hemorrhage should be guided by the principle that blood is exquisitely toxic to the brain.

The Hounsfield Unit of Perihematomal Edema Is Associated With Poor Clinical Outcomes in Intracerebral Hemorrhage

BACKGROUND

Hounsfield unit (HU) of perihematomal edema (PHE) may be a predictor of prognosis of intracerebral hemorrhage (ICH). Our study evaluated whether PHE mean HU at the 72 hours after ICH predicts outcome, and how it compares against other PHE measures.

METHODS

Patients with ICH from a tertiary medical institution were included. PHE was segmented by the semiautomatic plane method to measure volume and mean HU. Outcomes of interest was poor 90-day prognosis (modified Rankin Scale score ≥3). Logistic regression was used to assess relationships with outcome.

RESULTS

Data from a total of 159 patients with ICH were collected. The median mean HU of PHE at 72 hours was 22.1 (IQR: 19.2-25.0). Binary logistic regression showed that the 72-hour PHE mean HU was negatively correlated with the poor prognosis of patients with ICH (OR 0.59, 95% CI 0.47-0.75, P < 0.05). The receiver operator curves of meaningful indicators revealed that the area under the curve (AUC) of PHE mean HU at 72 hours was larger and the difference of AUC between PHE mean HU with PHE absolute volume or extension distance were statistically significant (P < 0.05). The 72-hour PHE mean HU has a higher value in predicting adverse prognosis of patients with ICH.

CONCLUSIONS

The PHE mean HU at 72 hours was negatively correlated with the poor prognosis of patients with ICH. The prediction ability of PHE mean HU at 72 hours was better than PHE absolute volume and extension distance, contributing to a rather good index for predicting outcome of ICH.