Background: The objective of this study was to assess the performance of the first publicly available automated 3D segmentation for spontaneous intracerebral hemorrhage (ICH) based on a 3D neural network before and after retraining. Methods: We performed an independent validation of this model using a multicenter retrospective cohort. Performance metrics were evaluated using the dice score (DSC), sensitivity, and positive predictive values (PPV). We retrained the original model (OM) and assessed the performance via an external validation design. A multivariate linear regression model was used to identify independent variables associated with the model’s performance. Agreements in volumetric measurements and segmentation were evaluated using Pearson’s correlation coefficients (r) and intraclass correlation coefficients (ICC), respectively. With 1040 patients, the OM had a median DSC, sensitivity, and PPV of 0.84, 0.79, and 0.93, compared to thoseo f 0.83, 0.80, and 0.91 in the retrained model (RM). However, the median DSC for infratentorial ICH was relatively low and improved significantly after retraining, at p < 0.001. ICH volume and location were significantly associated with the DSC, at p < 0.05. The agreement between volumetric measurements (r > 0.90, p > 0.05) and segmentations (ICC ≥ 0.9, p < 0.001) was excellent. Conclusion: The model demonstrated good generalization in an external validation cohort. Location-specific variances improved significantly after retraining. External validation and retraining are important steps to consider before applying deep learning models in new clinical settings.

Intracerebral hemorrhage (ICH) and perihematomal edema (PHE) are key imaging markers of primary and secondary brain injury in hemorrhagic stroke. Accurate segmentation and quantification of ICH and PHE can help with prognostication and guide treatment planning. In this study, we combined Swin-Unet Transformers with nnU-NETv2 convolutional network for segmentation of ICH and PHE on non-contrast head CTs. We also applied test-time data augmentations to assess individual-level prediction uncertainty, ensuring high confidence in prediction. The model was trained on 1782 CT scans from a multicentric trial and tested in two independent datasets from Yale (n = 396) and University of Berlin Charité Hospital and University Medical Center Hamburg-Eppendorf (n = 943). Model performance was evaluated with the Dice coefficient and Volume Similarity (VS). Our dual Swin-nnUNET model achieved a median (95% confidence interval) Dice = 0.93 (0.90–0.95) and VS = 0.97 (0.95–0.98) for ICH, and Dice = 0.70 (0.64–0.75) and VS = 0.87 (0.80–0.93) for PHE segmentation in the Yale cohort. Dice = 0.86 (0.80–0.90) and VS = 0.91 (0.85–0.95) for ICH and Dice = 0.65 (0.56–0.70) and VS = 0.86 (0.77–0.93) for PHE segmentation in the Berlin/Hamburg-Eppendorf cohort. Prediction uncertainty was associated with lower segmentation accuracy, smaller ICH/PHE volumes, and infratentorial location. Our results highlight the benefits of a dual transformer-convolutional neural network architecture for ICH/PHE segmentation and test-time augmentation for uncertainty quantification.

Background and Purpose: Fully automated methods for segmentation and volume quantification of intraparenchymal hemorrhage (ICH), intraventricular hemorrhage extension (IVH), and perihematomal edema (PHE) are gaining increasing interest. Yet, reliabilities demonstrate considerable variances amongst each other. Our aim was therefore to evaluate both the intra- and interrater reliability of ICH, IVH and PHE on ground-truth segmentation masks. Methods: Patients with primary spontaneous ICH were retrospectively included from a German tertiary stroke center (Charité Berlin; January 2016–June 2020). Baseline and follow-up non-contrast Computed Tomography (NCCT) scans were analyzed for ICH, IVH, and PHE volume quantification by two radiology residents. Raters were blinded to all demographic and outcome data. Inter- and intrarater agreements were determined by calculating the Intraclass Correlation Coefficient (ICC) for a randomly selected set of patients with ICH, IVH, and PHE. Results: 100 out of 670 patients were included in the analysis. Interrater agreements ranged from an ICC of 0.998 for ICH (95% CI [0.993; 0.997]), to an ICC of 0.979 for IVH (95% CI [0.984; 0.993]), and an ICC of 0.886 for PHE (95% CI [0.760; 0.938]), all p-values < 0.001. Intrarater agreements ranged from an ICC of 0.997 for ICH (95% CI [0.996; 0.998]), to an ICC of 0.995 for IVH (95% CI [0.992; 0.996]), and an ICC of 0.980 for PHE (95% CI [0.971; 0.987]), all p-values < 0.001. Conclusion Manual segmentations of ICH, IVH, and PHE demonstrate good-to-excellent inter- and intrarater reliabilities, with the highest agreement for ICH and IVH and lowest for PHE. Therefore, the degree of variances reported in fully automated quantification methods might be related amongst others to variances in ground-truth masks.

Background: Noncontrast Computed Tomography (NCCT) features are promising markers for acute hematoma expansion (HE) in patients with intracerebral hemorrhage (ICH). It remains unclear whether accurate identification of these markers is also reliable in raters with different levels of experience. Methods: Patients with acute spontaneous ICH admitted at four tertiary centers in Germany and Italy were retrospectively included from January 2017 to June 2020. In total, nine NCCT markers were rated by one radiology resident, one radiology fellow, and one neuroradiology fellow with different levels experience in ICH imaging. Interrater reliabilities of the resident and radiology fellow were evaluated by calculated Cohen’s kappa (κ) statistics in reference to the neuroradiology fellow who was referred as the gold standard. Gold-standard ratings were evaluated by calculated interrater κ statistics. Global interrater reliabilities were evaluated by calculated Fleiss kappa statistics across all three readers. A comparison of receiver operating characteristics (ROCs) was used to evaluate differences in the diagnostic accuracy for predicting acute hematoma expansion (HE) among the raters. Results: Substantial-to-almost-perfect interrater concordance was found for the resident with interrater Cohen’s kappa from 0.70 (95% CI 0.65–0.81) to 0.96 (95% CI 0.94–0.98). The interrater Cohen’s kappa for the radiology fellow was moderate to almost perfect and ranged from 0.58 (95% CI 0.52–0.65) to 94 (95% CI 92–0.97). The intrarater gold-standard Cohen’s kappa was almost perfect and ranged from 0.79 (95% CI 0.78–0.90) to 0.98 (95% CI 0.78–0.90). The global interrater Fleiss kappa ranged from 0.62 (95%CI 0.57–0.66) to 0.93 (95%CI 0.89–0.97). The diagnostic accuracy for the prediction of acute hematoma expansion (HE) was different for the island sign and fluid sign, with p-values < 0.05. Conclusion: The NCCT markers had a substantial-to-almost-perfect interrater agreement among raters with different levels of experience. Differences in the diagnostic accuracy for the prediction of acute HE were found in two out of nine NCCT markers. The study highlights the promising utility of NCCT markers for acute HE prediction.

Hematoma expansion (HE) is an independent predictor of poor outcomes and a modifiable treatment target in intracerebral hemorrhage (ICH). Evaluating HE in large datasets requires segmentation of hematomas on admission and follow-up CT scans, a process that is time-consuming and labor-intensive in large-scale studies. Automated segmentation of hematomas can expedite this process; however, cumulative errors from segmentation on admission and follow-up scans can hamper accurate HE classification. In this study, we combined a tandem deep-learning classification model with automated segmentation to generate probability measures for false HE classifications. With this strategy, we can limit expert review of automated hematoma segmentations to a subset of the dataset, tailored to the research team’s preferred sensitivity or specificity thresholds and their tolerance for false-positive versus false-negative results. We utilized three separate multicentric cohorts for cross-validation/training, internal testing, and external validation (n = 2261) to develop and test a pipeline for automated hematoma segmentation and to generate ground truth binary HE annotations (≥3, ≥6, ≥9, and ≥12.5 mL). Applying a 95% sensitivity threshold for HE classification showed a practical and efficient strategy for HE annotation in large ICH datasets. This threshold excluded 47–88% of test-negative predictions from expert review of automated segmentations for different HE definitions, with less than 2% false-negative misclassification in both internal and external validation cohorts. Our pipeline offers a time-efficient and optimizable method for generating ground truth HE classifications in large ICH datasets, reducing the burden of expert review of automated hematoma segmentations while minimizing misclassification rate.

Background: Digital twin technology holds promise for personalized stroke care, but current applications remain fragmented. This systematic review investigates how digital twins are currently utilized in the stroke care continuum. Methods: Following PRISMA guidelines, we conducted a systematic search of PubMed, Web of Science, and the Cochrane Library through April 2025. Studies applying digital twins to acute ischemic stroke care were included. Each study was categorized along the stroke care continuum (pre-stroke, in-hospital, post-stroke) and assessed using a digital twin maturity framework (L0 to L3). We extracted data on clinical intent, modeling approach, validation strategy, study design, population, sample size, and key outcomes to enable structured synthesis. Results: Eight studies met inclusion criteria. Half targeted pre-stroke risk prediction (e.g., modeling atherosclerosis or atrial fibrillation), 2 simulated mechanical thrombectomy, 1 supported prehospital diagnosis, and 1 predicted post-stroke disease progression. Most models remained at maturity levels L1 to L2, lacking real-time updating or workflow integration. Technologies included machine learning ( n = 3), computational fluid dynamics ( n = 3), and hybrid or rule-based approaches ( n = 2). Conclusions: Digital twins in stroke care are promising but remain preclinical. Current models predominantly address pre-stroke risk prediction or procedural simulation, with limited representation of acute decision support or post-stroke monitoring. Clinical integration is constrained by low technological maturity, limited real-world validation, and a lack of interoperability.