A5024004096- Acute Ischemic Stroke Management
- Cerebrovascular and Carotid Artery Diseases
- Medical Image Segmentation Techniques
- Retinal Imaging and Analysis
- Generative Adversarial Networks and Image Synthesis
- Machine Learning in Healthcare
- Artificial Intelligence in Healthcare and Education
- Sepsis Diagnosis and Treatment
- Radiomics and Machine Learning in Medical Imaging
- Medical Imaging and Analysis
- Stroke Rehabilitation and Recovery
- Intracranial Aneurysms: Treatment and Complications
- Traumatic Brain Injury and Neurovascular Disturbances
- Explainable Artificial Intelligence (XAI)
- Venous Thromboembolism Diagnosis and Management
- AI in cancer detection
- Intracerebral and Subarachnoid Hemorrhage Research
- MRI in cancer diagnosis
- Privacy-Preserving Technologies in Data
- Advanced Image Processing Techniques
- Cardiovascular Health and Disease Prevention
- Advanced MRI Techniques and Applications
- Neurosurgical Procedures and Complications
- Cardiac, Anesthesia and Surgical Outcomes
- Computer Graphics and Visualization Techniques
Charité - Universitätsmedizin Berlin
2020-2025
Humboldt-Universität zu Berlin
2023-2024
Freie Universität Berlin
2023-2024
Artificial Intelligence in Medicine (Canada)
2022
University of Amsterdam
2018-2022
Amsterdam University Medical Centers
2019-2022
Abstract Average Hausdorff distance is a widely used performance measure to calculate the between two point sets. In medical image segmentation, it compare ground truth images with segmentations allowing their ranking. We identified, however, ranking errors of average making less suitable for applications in segmentation assessment. To mitigate this error, we present modified calculation that have coined “balanced distance”. simulate ranking, manually created non-overlapping common magnetic...
Background: Endovascular treatment (EVT) is effective for stroke patients with a large vessel occlusion (LVO) of the anterior circulation. To further improve personalized care, it essential to accurately predict outcome after EVT. Machine learning might outperform classical prediction methods as capable addressing complex interactions and non-linear relations between variables. Methods: We included from Multicenter Randomized Clinical Trial Treatment Acute Ischemic Stroke in Netherlands (MR...
Introduction: Arterial brain vessel assessment is crucial for the diagnostic process in patients with cerebrovascular disease. Non-invasive neuroimaging techniques, such as time-of-flight (TOF) magnetic resonance angiography (MRA) imaging are applied clinical routine to depict arteries. They are, however, only visually assessed. Fully automated segmentation integrated into could facilitate time-critical diagnosis of abnormalities and might identification valuable biomarkers events. In...
To evaluate the transferability of deep learning (DL) models for early detection adverse events to previously unseen hospitals.
Deep learning requires large labeled datasets that are difficult to gather in medical imaging due data privacy issues and time-consuming manual labeling. Generative Adversarial Networks (GANs) can alleviate these challenges enabling synthesis of shareable data. While 2D GANs have been used generate images with their corresponding labels, they cannot capture the volumetric information 3D imaging. more suitable for this volumes but not labels. One reason might be synthesizing is challenging...
Reliable prediction of outcomes aneurysmal subarachnoid hemorrhage (aSAH) based on factors available at patient admission may support responsible allocation resources as well treatment decisions. Radiographic and clinical scoring systems help clinicians estimate disease severity, but their predictive value is limited, especially in devising strategies. In this study, we aimed to examine whether a machine learning (ML) approach using variables improve outcome aSAH compared established...
Background Accurate prediction of clinical outcome is utmost importance for choices regarding the endovascular treatment (EVT) acute stroke. Recent studies on modeling stroke focused mostly characteristics and radiological scores available at baseline. Radiological images are composed millions voxels, a lot information can be lost when representing this by single value. Therefore, in study we aimed developing models that take into account whole imaging data combined with Methods We included...
Outcome prediction after mechanical thrombectomy (MT) in patients with acute ischemic stroke (AIS) and large vessel occlusion (LVO) is commonly performed by focusing on favorable outcome (modified Rankin Scale, mRS 0-2) 3 months but poor representing severe disability mortality (mRS 5 6) might be of equal importance for clinical decision-making.We retrospectively analyzed AIS LVO undergoing MT from 2009 to 2018. Prognostic variables were grouped baseline (A), MRI-derived including mismatch...
Hematoma expansion occasionally occurs in patients with intracerebral hemorrhage (ICH), associating poor outcome. Multimodal neural networks incorporating convolutional network (CNN) analysis of images and tabular data are known to show promising results prediction classification tasks. We aimed develop a reliable multimodal model that comprehensively analyzes CT clinical variables predict hematoma expansion. retrospectively enrolled ICH at four hospitals between 2017 2021, assigning from...
Abstract Early and reliable prediction of shunt-dependent hydrocephalus (SDHC) after aneurysmal subarachnoid hemorrhage (aSAH) may decrease the duration in-hospital stay reduce risk catheter-associated meningitis. Machine learning (ML) improve predictions SDHC in comparison to traditional non-ML methods. ML models were trained for CHESS SDASH two combined individual feature sets with clinical, radiographic, laboratory variables. Seven different algorithms used including three types...
Abstract Introduction Arterial brain vessel assessment is crucial for the diagnostic process in patients with cerebrovascular disease. Noninvasive neuroimaging techniques such as time-of-flight (TOF) magnetic resonance angiography (MRA) imaging are applied clinical routine to depict arteries. They are, however, only visually assessed. Fully automated segmentation integrated into could facilitate time-critical diagnosis of abnormalities and might identification valuable biomarkers events. In...
Sharing labeled data is crucial to acquire large datasets for various Deep Learning applications. In medical imaging, this often not feasible due privacy regulations. Whereas anonymization would be a solution, standard techniques have been shown partially reversible. Here, synthetic using Generative Adversarial Network (GAN) with differential guarantees could solution ensure the patient's while maintaining predictive properties of data. study, we implemented Wasserstein GAN (WGAN) and...
Stroke is a major cause of death or disability. As imaging-based patient stratification improves acute stroke therapy, dynamic susceptibility contrast magnetic resonance imaging (DSC-MRI) interest in image brain perfusion. However, expert-level perfusion maps require manual semi-manual post-processing by medical expert making the procedure time-consuming and less-standardized. Modern machine learning methods such as generative adversarial networks (GANs) have potential to automate map...
When time since stroke onset is unknown, DWI-FLAIR mismatch rating an established technique for patient stratification. A visible DWI lesion without corresponding parenchymal hyperintensity on FLAIR suggests of under 4.5 h and thus a potential benefit from intravenous thrombolysis. To improve accuracy availability the concept, deep learning might be able to augment human support decision-making in these cases. We used unprocessed coregistered imaging data train model predict dichotomized...
Elderly patients receiving lumbar fusion surgeries present with a higher risk profile, which necessitates robust predictor of postoperative outcomes. The Red Distribution Width (RDW) is preoperative routinely determined parameter that reflects the degree heterogeneity red blood cells. Thereby, RDW associated frailty in hospital-admitted patients. This study aims to elucidate potential as biomarker predictive prolonged hospital stays following elective mono-segmental surgery elderly In this...
Brain arteries are routinely imaged in the clinical setting by various modalities, e.g., time-of-flight magnetic resonance angiography (TOF-MRA). These imaging techniques have great potential for diagnosis of cerebrovascular disease, disease progression, and response to treatment. Currently, however, only qualitative assessment is implemented applications, relying on visual inspection. While manual or semi-automated approaches quantification exist, such solutions impractical as they...
Abstract Background Arterial brain vessel segmentation allows utilising clinically relevant information contained within the cerebral vascular tree. Currently, however, no standardised performance measure is available to evaluate quality of segmentations. Thus, we developed a selection framework based on manual visual scoring simulated variations find most suitable for segmentation. Methods To simulate variations, manually created non-overlapping errors common in magnetic resonance...
Abstract Whether endovascular thrombectomy (EVT) improves functional outcome in patients with large-vessel occlusion (LVO) stroke that do not comply inclusion criteria of randomized controlled trials (RCTs) but are considered for EVT clinical practice is uncertain. We aimed to systematically identify LVO underrepresented RCTs who might benefit from EVT. Following the premises (i) without reperfusion after represent a non-treated control group and (ii) level affects treatment benefit, we...
Abstract The circle of Willis (CoW) is a network cerebral arteries with significant inter-individual anatomical variations. Deep learning has been used to characterize and quantify the status CoW in various applications for diagnosis treatment cerebrovascular disease. In medical imaging, performance deep models limited by diversity size training datasets. To address data scarcity, generative adversarial networks (GANs) have applied generate synthetic vessel neuroimaging data. However,...
Abstract Introduction Aneurysmal subarachnoid hemorrhage (aSAH) is a life-threatening condition with significant variability in patients’ outcomes. Radiographic scores used to assess the extent of SAH or other potentially outcome-relevant pathologies are limited by interrater and do not utilize all available information from imaging. Image segmentation plays an important role extracting relevant images enabling precise identification delineation objects regions interest. Thus, offers...