Improved Precision and Safety of Supra-Tentorial Intracranial Hematoma Puncture Using C-Arm CT Four-Dimensional Navigation.

BACKGROUND Spontaneous intracerebral hemorrhage has a high fatality rate within the initial month after onset. This study determined the safety and therapeutic efficacy of minimally invasive puncture for supra-tentorial intracranial hematoma under C-arm computed tomography (CT) 4-dimensional navigation. MATERIAL AND METHODS We retrospectively analyzed 64 patients with supra-tentorial cerebral hemorrhage from June 2020 to May 2023; 31 patients were assigned to the study group (C-arm CT navigation puncture) and 33 patients were in the control group (conventional CT-guided puncture). The analysis focused on assessment of puncture error, postoperative complication rate, and the Glasgow Outcome Scale (GOS) and National Institute of Health Stroke Scale (NIHSS) scores 30 and 90 days after surgery. RESULTS C-arm CT navigation puncture had improved precision, with significantly reduced transverse (3.17±1.75 mm) and longitudinal (1.83±1.21 mm) deviations, compared with the control group (7.88±1.74 mm and 5.50±1.84 mm, respectively; P<0.05). The overall postoperative complication rate was significantly lower in the study group than in the control group (12.90% vs 36.36%, P<0.05). The mean GOS score was higher in the study group than in the control group 30 and 90 days postoperatively (3.42±0.96 and 3.97±0.95 vs 2.94±0.79 and 3.46±0.90, respectively; P<0.05), while the mean NIHSS score was lower in the study group than in the control group 30 and 90 days postoperatively (10.58±6.52 and 5.97±4.55 vs 14.42±8.13 and 9.55±8.31, respectively; P<0.05). CONCLUSIONS Supra-tentorial intracranial hematoma puncture under C-arm CT 4-dimensional navigation is accurate, safe, and beneficial.

An interpretable artificial intelligence model based on CT for prognosis of intracerebral hemorrhage: a multicenter study.

OBJECTIVES: To develop and validate a novel interpretable artificial intelligence (AI) model that integrates radiomic features, deep learning features, and imaging features at multiple semantic levels to predict the prognosis of intracerebral hemorrhage (ICH) patients at 6 months post-onset. MATERIALS AND METHODS: Retrospectively enrolled 222 patients with ICH for Non-contrast Computed Tomography (NCCT) images and clinical data, who were divided into a training cohort (n = 186, medical center 1) and an external testing cohort (n = 36, medical center 2). Following image preprocessing, the entire hematoma region was segmented by two radiologists as the volume of interest (VOI). Pyradiomics algorithm library was utilized to extract 1762 radiomics features, while a deep convolutional neural network (EfficientnetV2-L) was employed to extract 1000 deep learning features. Additionally, radiologists evaluated imaging features. Based on the three different modalities of features mentioned above, the Random Forest (RF) model was trained, resulting in three models (Radiomics Model, Radiomics-Clinical Model, and DL-Radiomics-Clinical Model). The performance and clinical utility of the models were assessed using the Area Under the Receiver Operating Characteristic Curve (AUC), calibration curve, and Decision Curve Analysis (DCA), with AUC compared using the DeLong test. Furthermore, this study employs three methods, Shapley Additive Explanations (SHAP), Grad-CAM, and Guided Grad-CAM, to conduct a multidimensional interpretability analysis of model decisions. RESULTS: The Radiomics-Clinical Model and DL-Radiomics-Clinical Model exhibited relatively good predictive performance, with an AUC of 0.86 [95% Confidence Intervals (CI): 0.71, 0.95; P < 0.01] and 0.89 (95% CI: 0.74, 0.97; P < 0.01), respectively, in the external testing cohort. CONCLUSION: The multimodal explainable AI model proposed in this study can accurately predict the prognosis of ICH. Interpretability methods such as SHAP, Grad-CAM, and Guided Grad-Cam partially address the interpretability limitations of AI models. Integrating multimodal imaging features can effectively improve the performance of the model. CLINICAL RELEVANCE STATEMENT: Predicting the prognosis of patients with ICH is a key objective in emergency care. Accurate and efficient prognostic tools can effectively prevent, manage, and monitor adverse events in ICH patients, maximizing treatment outcomes.

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.

Associations between diffusion kurtosis imaging metrics and neurodevelopmental outcomes in neonates with low-grade germinal matrix and intraventricular hemorrhage.

Diffusion Kurtosis Imaging (DKI)-derived metrics are recognized as indicators of maturation in neonates with low-grade germinal matrix and intraventricular hemorrhage (GMH-IVH). However, it is not yet known if these factors are associated with neurodevelopmental outcomes. The objective of this study was to acquire DKI-derived metrics in neonates with low-grade GMH-IVH, and to demonstrate their association with later neurodevelopmental outcomes. In this prospective study, neonates with low-grade GMH-IVH and control neonates were recruited, and DKI were performed between January 2020 and March 2021. These neonates underwent the Bayley Scales of Infant Development test at 18 months of age. Mean kurtosis (MK), radial kurtosis (RK) and gray matter values were measured. Spearman correlation analyses were conducted for the measured values and neurodevelopmental outcome scores. Forty controls (18 males, average gestational age (GA) 30 weeks ± 1.3, corrected GA at MRI scan 38 weeks ± 1) and thirty neonates with low-grade GMH-IVH (13 males, average GA 30 weeks ± 1.5, corrected GA at MRI scan 38 weeks ± 1). Neonates with low-grade GMH-IVH exhibited lower MK and RK values in the PLIC and the thalamus (P < 0.05). The MK value in the thalamus was associated with Mental Development Index (MDI) (r = 0.810, 95% CI 0.695-0.13; P < 0.001) and Psychomotor Development Index (PDI) (r = 0.852, 95% CI 0.722-0.912; P < 0.001) scores. RK value in the caudate nucleus significantly and positively correlated with MDI (r = 0.496, 95% CI 0.657-0.933; P < 0.001) and PDI (r = 0.545, 95% CI 0.712-0.942; P < 0.001) scores. The area under the curve (AUC) were used to assess diagnostic performance of MK and RK in thalamus (AUC = 0.866, 0.787) and caudate nucleus (AUC = 0.833, 0.671) for predicting neurodevelopmental outcomes. As quantitative neuroimaging markers, MK in thalamus and RK in caudate nucleus may help predict neurodevelopmental outcomes in neonates with low-grade GMH-IVH.

Purely Endoscopic Evacuation of Intracranial Hematomas.

Cerebral hemorrhage is a frequent disease and one of the main causes of disabilities. Even in the case of cerebral hemorrhage, if there were a treatment that would improve the functional prognosis, the benefits would be immeasurable. Although there are limited reports with a high level of evidence in past studies, it has been found that surgery can be effective if a large amount of hematoma can be removed in a minimally invasive manner. Also, it has become clear that the control of bleeding becomes a problem when surgery is performed within 2 days after the onset of stroke and that the therapeutic time window might be longer. In Japan, since the introduction of the transparent sheath by Nishihara et al., endoscopic hematoma removal has been widely performed and has become the standard surgical procedure. The three basic equipment needed for this surgery are a rigid scope, a suction coagulator, and a transparent sheath. The key point of the surgery is to search for hematomas and bleeding points through a transparent sheath and coagulate the bleeding vessels. In this chapter, we will introduce surgical techniques using these devices, but it is important to carefully decide on surgical options by considering your own technique, the condition of each patient, and the devices available in the area.

Hematoma expansion prediction based on SMOTE and XGBoost algorithm.

Hematoma expansion (HE) is a high risky symptom with high rate of occurrence for patients who have undergone spontaneous intracerebral hemorrhage (ICH) after a major accident or illness. Correct prediction of the occurrence of HE in advance is critical to help the doctors to determine the next step medical treatment. Most existing studies focus only on the occurrence of HE within 6 h after the occurrence of ICH, while in reality a considerable number of patients have HE after the first 6 h but within 24 h. In this study, based on the medical doctors recommendation, we focus on prediction of the occurrence of HE within 24 h, as well as the occurrence of HE every 6 h within 24 h. Based on the demographics and computer tomography (CT) image extraction information, we used the XGBoost method to predict the occurrence of HE within 24 h. In this study, to solve the issue of highly imbalanced data set, which is a frequent case in medical data analysis, we used the SMOTE algorithm for data augmentation. To evaluate our method, we used a data set consisting of 582 patients records, and compared the results of proposed method as well as few machine learning methods. Our experiments show that XGBoost achieved the best prediction performance on the balanced dataset processed by the SMOTE algorithm with an accuracy of 0.82 and F1-score of 0.82. Moreover, our proposed method predicts the occurrence of HE within 6, 12, 18 and 24 h at the accuracy of 0.89, 0.82, 0.87 and 0.94, indicating that the HE occurrence within 24 h can be predicted accurately by the proposed method.

Necrotizing enterocolitis in a term newborn after spontaneous cerebral parenchymal hemorrhage: a case report.

BACKGROUND: Necrotizing enterocolitis (NEC) and intracranial hemorrhage are severe emergencies in the neonatal period. The two do not appear to be correlated. However, our report suggests that parenchymal brain hemorrhage in full-term newborns may put patients at risk for NEC by altering intestinal function through the brain-gut axis. CASE PRESENTATION: We present a case of spontaneous parenchymal cerebral hemorrhage in a full-term newborn who developed early-stage NEC on Day 15. CONCLUSIONS: It is possible to consider brain parenchymal hemorrhage as a risk factor for the appearance of NEC. Clinicians should be highly cautious about NEC in infants who have experienced parenchymal hemorrhage. This article is the first to discuss the relationship between parenchymal hemorrhage and NEC in full-term newborns.

HE-Mind: A model for automatically predicting hematoma expansion after spontaneous intracerebral hemorrhage.

PURPOSE: To develop and validate an end-to-end model for automatically predicting hematoma expansion (HE) after spontaneous intracerebral hemorrhage (sICH) using a novel deep learning framework. METHODS: This multicenter retrospective study collected cranial noncontrast computed tomography (NCCT) images of 490 patients with sICH at admission for model training (n = 236), internal testing (n = 60), and external testing (n = 194). A HE-Mind model was designed to predict HE, which consists of a densely connected U-net for segmentation process, a multi-instance learning strategy for resolving label ambiguity and a Siamese network for classification process. Two radiomics models based on support vector machine or logistic regression and two deep learning models based on residual network or Swin transformer were developed for performance comparison. Reader experiments including physician diagnosis mode and artificial intelligence mode were conducted for efficiency comparison. RESULTS: The HE-Mind model showed better performance compared to the comparative models in predicting HE, with areas under the curve of 0.849 and 0.809 in the internal and external test sets respectively. With the assistance of the HE-Mind model, the predictive accuracy and work efficiency of the emergency physician, junior radiologist, and senior radiologist were significantly improved, with accuracies of 0.768, 0.789, and 0.809 respectively, and reporting times of 7.26 s, 5.08 s, and 3.99 s respectively. CONCLUSIONS: The HE-Mind model could rapidly and automatically process the NCCT data and predict HE after sICH within three seconds, indicating its potential to assist physicians in the clinical diagnosis workflow of HE.

The Usefulness of the 3.1-mm-diameter 4K Rigid Endoscope for Intracerebral Hematoma Evacuation.

The indication for surgical intervention in spontaneous intracerebral hemorrhage remains controversial. Although many clinical trials have failed to demonstrate its efficacy over medical treatment, less invasive endoscopic treatment is expected to demonstrate its superiority. A novel endoscopic system for hematoma removal consisting of a 3.1-mm-diameter 4K high-resolution rigid endoscope was used.The system was used in eight cases of spontaneous intracerebral hemorrhage. It provided improved maneuverability of the surgical instrument while maintaining satisfactory image quality. The surgical goal was achieved in all cases without any complications, including perioperative rebleeding.Endoscopic hematoma removal using the 3.1 mm high-resolution endoscope is an alternative minimally invasive approach to spontaneous intracerebral hemorrhage with improved reliability.

Nomogram established on account of Lasso-logistic regression for predicting hemorrhagic transformation in patients with acute ischemic stroke after endovascular thrombectomy.

BACKGROUND: Hemorrhagic transformation (HT) is a common and serious complication in patients with acute ischemic stroke (AIS) after endovascular thrombectomy (EVT). This study was performed to determine the predictive factors associated with HT in stroke patients with EVT and to establish and validate a nomogram that combines with independent predictors to predict the probability of HT after EVT in patients with AIS. METHODS: All patients were randomly divided into development and validation cohorts at a ratio of 7:3. The least absolute shrinkage and selection operator (LASSO) regression was used to select the optimal factors, and multivariate logistic regression analysis was used to build a clinical prediction model. Calibration plots, decision curve analysis (DCA) and receiver operating characteristic curve (ROC) were generated to assess predictive performance. RESULTS: LASSO regression analysis showed that Alberta Stroke Program Early CT Scores (ASPECTS), international normalized ratio (INR), uric acid (UA), neutrophils (NEU) were the influencing factors for AIS with HT after EVT. A novel prognostic nomogram model was established to predict the possibility of HT with AIS after EVT. The calibration curve showed that the model had good consistency. The results of ROC analysis showed that the AUC of the prediction model established in this study for predicting HT was 0.797 in the development cohort and 0.786 in the validation cohort. CONCLUSION: This study proposes a novel and practical nomogram based on ASPECTS, INR, UA, NEU, which can well predict the probability of HT after EVT in patients with AIS.

Machine learning-based model for predicting outcomes in cerebral hemorrhage patients with leukemia.

BACKGROUND AND PURPOSE: Intracranial hemorrhage (ICH) in leukemia patients progresses rapidly with high mortality. Limited data are available on imaging studies in this population. The study aims to develop prediction models for 7-day and short-term mortality risk based on the non-contrast computed tomography (NCCT) image features. METHODS: The NCCT image features of ICH in 135 leukemia patients between 2007-2023 were retrospectively extracted using manual assessment and radiomics methods. After multiple imputation of missing laboratory data, univariate logistic regression and least absolute shrinkage and selection operator (LASSO) were used for feature selection. Random forest models were built with comprehensive evaluation and ranking of feature importance. RESULT: 135 and 129 patients were included in the studies for 7-day and short-term prognostic models, respectively. The median age of all enrolled patients was 35 years, and there were 86 male patients (63.7 %). Clinical models (validation: AUC [area under the curve] = 0.78, AUPRC [area under the precision-recall curve] = 0.73; AUC = 0.84, AUPRC = 0.86), radiomics models (validation: AUC = 0.82, AUPRC = 0.78; AUC = 0.75, AUPRC = 0.77), and the combined models (validation: AUC = 0.84, AUPRC = 0.83; AUC = 0.87, AUPRC = 0.89) predicted 7-day and short-term mortality with good predictive efficacy. Clinical decision curve analysis showed that the combined models predicted 7-day and 30-day risk of death would be more beneficial than other models. Shape features contributed significantly more than semantic features in both radiomics models and combined models (93.3 %, 52.1 %, as well as 85.2 %,37.4 %, respectively) for 7-day and 30-day mortality. CONCLUSIONS: Combined models constructed based on NCCT perform well in predicting the risk of 7-day and short-term mortality in ICH patients with leukemia. Shape features extracted by radiomics are important markers for modeling the prognosis.

Associations of Cerebral Small Vessel Disease and Chronic Kidney Disease in Patients With Acute Intracerebral Hemorrhage: A Cross-Sectional Study.

BACKGROUND AND OBJECTIVES: Chronic kidney disease (CKD) may be associated with the pathogenesis and phenotype of cerebral small vessel disease (SVD), which is the commonest cause of intracerebral hemorrhage (ICH). The purpose of this study was to investigate the associations of CKD with ICH neuroimaging phenotype, volume, and location, total burden of small vessel disease, and its individual components. METHODS: In 2 cohorts of consecutive patients with ICH evaluated with MRI, we investigated the frequency and severity of CKD based on established Kidney Disease Improving Global Outcomes criteria, requiring estimated glomerular filtration rate (eGFR) measurements <60 mL/min/1.732 ≥ 3 months apart to define CKD. MRI scans were rated for ICH neuroimaging phenotype (arteriolosclerosis, cerebral amyloid angiopathy, mixed location SVD, or cryptogenic ICH) and the presence of markers of SVD (white matter hyperintensities [WMHs], cerebral microbleeds [CMBs], lacunes, and enlarged perivascular spaces, defined according to the STandards for ReportIng Vascular changes on nEuroimaging criteria). We used multinomial, binomial logistic, and ordinal logistic regression models adjusted for age, sex, hypertension, and diabetes to account for possible confounding caused by shared risk factors of CKD and SVD. RESULTS: Of 875 patients (mean age 66 years, 42% female), 146 (16.7%) had CKD. After adjusting for age, sex, and comorbidities, patients with CKD had higher rates of mixed SVD than those with eGFR >60 (relative risk ratio 2.39, 95% CI 1.16-4.94, p = 0.019). Severe WMHs, deep microbleeds, and lacunes were more frequent in patients with CKD, as was a higher overall SVD burden score (odds ratio 1.83 for each point on the ordinal scale, 95% CI 1.31-2.56, p < 0.001). Patients with eGFR ≤30 had more CMBs (median 7 [interquartile range 1-23] vs 2 [0-8] for those with eGFR >30, p = 0.007). DISCUSSION: In patients with ICH, CKD was associated with SVD burden, a mixed SVD phenotype, and markers of arteriolosclerosis. Our findings indicate that CKD might independently contribute to the pathogenesis of arteriolosclerosis and mixed SVD, although we could not definitively account for the severity of shared risk factors. Longitudinal and experimental studies are, therefore, needed to investigate causal associations. Nevertheless, stroke clinicians should be aware of CKD as a potentially independent and modifiable risk factor of SVD.

Neurodevelopmental outcome in preterm infants with intraventricular hemorrhages: the potential of quantitative brainstem MRI.

OBJECTIVES: This retrospective study aimed to identify quantitative magnetic resonance imaging markers in the brainstem of preterm neonates with intraventricular hemorrhages. It delves into the intricate associations between quantitative brainstem magnetic resonance imaging metrics and neurodevelopmental outcomes in preterm infants with intraventricular hemorrhage, aiming to elucidate potential relationships and their clinical implications. MATERIALS AND METHODS: Neuroimaging was performed on preterm neonates with intraventricular hemorrhage using a multi-dynamic multi-echo sequence to determine T1 relaxation time, T2 relaxation time, and proton density in specific brainstem regions. Neonatal outcome scores were collected using the Bayley Scales of Infant and Toddler Development. Statistical analysis aimed to explore potential correlations between magnetic resonance imaging metrics and neurodevelopmental outcomes. RESULTS: Sixty preterm neonates (mean gestational age at birth 26.26 ± 2.69 wk; n = 24 [40%] females) were included. The T2 relaxation time of the midbrain exhibited significant positive correlations with cognitive (r = 0.538, P < 0.0001, Pearson's correlation), motor (r = 0.530, P < 0.0001), and language (r = 0.449, P = 0.0008) composite scores at 1 yr of age. CONCLUSION: Quantitative magnetic resonance imaging can provide valuable insights into neurodevelopmental outcomes after intraventricular hemorrhage, potentially aiding in identifying at-risk neonates. Multi-dynamic multi-echo sequence sequences hold promise as an adjunct to conventional sequences, enhancing the sensitivity of neonatal magnetic resonance neuroimaging and supporting clinical decision-making for these vulnerable patients.

Swirl sign score system: a novel and practical tool for predicting hematoma expansion risk after spontaneous intracerebral haemorrhage.

OBJECTIVE: To methodically analyse the swirl sign and construct a scoring system to predict the risk of hematoma expansion (HE) after spontaneous intracerebral haemorrhage (sICH). METHODS: We analysed 231 of 683 sICH patients with swirl signs on baseline noncontrast CT (NCCT) images. The characteristics of the swirl sign were analysed, including the number, maximum diameter, shape, boundary, minimum CT value of the swirl sign, and the minimum distance from the swirl sign to the edge of the hematoma. In the development cohort, univariate and multivariate analyses were used to identify independent predictors of HE, and logistic regression analysis was used to construct the swirl sign score system. The swirl sign score system was verified in the validation cohort. RESULTS: The number and the minimum CT value of the swirl sign were independent predictors of HE. The swirl sign score system was constructed (2 points for the number of swirl signs >1 and 1 point for the minimum CT value ≤41 Hounsfield units). The area under the curve of the swirl sign score system in predicting HE was 0.773 and 0.770 in the development and validation groups, respectively. CONCLUSIONS: The swirl sign score system is an easy-to-use radiological grading scale that requires only baseline NCCT images to effectively identify subjects at high risk of HE. ADVANCES IN KNOWLEDGE: Our newly developed semiquantitative swirl sign score system greatly improves the ability of swirl sign to predict HE.

Abnormal dynamic functional connectivity in young nondisabling intracerebral hemorrhage patients.

OBJECTIVE: Previous resting-state functional magnetic resonance imaging studies on intracerebral hemorrhage patients have focused more on the static characteristics of brain activity, while the time-varying effects during scanning have received less attention. Therefore, the current study aimed to explore the dynamic functional network connectivity changes of intracerebral hemorrhage patients. METHODS: Using independent component analysis, the sliding window approach, and the k-means clustering analysis method, different dynamic functional network connectivity states were detected from resting-state functional magnetic resonance imaging data of 37 intracerebral hemorrhage patients and 44 healthy controls. The inter-group differences in dynamic functional network connectivity patterns and temporal properties were investigated, followed by correlation analyses between clinical scales and abnormal functional indexes. RESULTS: Ten resting-state networks were identified, and the dynamic functional network connectivity matrices were clustered into four different states. The transition numbers were decreased in the intracerebral hemorrhage patients compared with healthy controls, which was associated with trail making test scores in patients. The cerebellar network and executive control network connectivity in State 1 was reduced in patients, and this abnormal dynamic functional connectivity was positively correlated with the animal fluency test scores of patients. INTERPRETATION: The current study demonstrated the characteristics of dynamic functional network connectivity in intracerebral hemorrhage patients and revealed that abnormal temporal properties and functional connectivity may be related to the performance of different cognitive domains after ictus. These results may provide new insights into exploring the neurocognitive mechanisms of intracerebral hemorrhage.

Predicting hematoma expansion using machine learning: An exploratory analysis of the ATACH 2 trial.

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.mm3 [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.

Boston Criteria v2.0 for Cerebral Amyloid Angiopathy Without Hemorrhage: An MRI-Neuropathologic Validation Study.

BACKGROUND AND OBJECTIVES: Updated criteria for the clinical-MRI diagnosis of cerebral amyloid angiopathy (CAA) have recently been proposed. However, their performance in individuals without symptomatic intracerebral hemorrhage (ICH) presentations is less defined. We aimed to assess the diagnostic performance of the Boston criteria version 2.0 for CAA diagnosis in a cohort of individuals ranging from cognitively normal to dementia in the community and memory clinic settings. METHODS: Fifty-four participants from the Mayo Clinic Study of Aging or Alzheimer's Disease Research Center were included if they had an antemortem MRI with gradient-recall echo sequences and a brain autopsy with CAA evaluation. Performance of the Boston criteria v2.0 was compared with v1.5 using histopathologically verified CAA as the reference standard. RESULTS: The median age at MRI was 75 years (interquartile range 65-80) with 28/54 participants having histopathologically verified CAA (i.e., moderate-to-severe CAA in at least 1 lobar region). The sensitivity and specificity of the Boston criteria v2.0 were 28.6% (95% CI 13.2%-48.7%) and 65.3% (95% CI 44.3%-82.8%) for probable CAA diagnosis (area under the receiver operating characteristic curve [AUC] 0.47) and 75.0% (55.1-89.3) and 38.5% (20.2-59.4) for any CAA diagnosis (possible + probable; AUC 0.57), respectively. The v2.0 Boston criteria were not superior in performance compared with the prior v1.5 criteria for either CAA diagnostic category. DISCUSSION: The Boston criteria v2.0 have low accuracy in patients who are asymptomatic or only have cognitive symptoms. Additional biomarkers need to be explored to optimize CAA diagnosis in this population.