A5034842922- Radiomics and Machine Learning in Medical Imaging
- Glioma Diagnosis and Treatment
- Medical Imaging Techniques and Applications
- MRI in cancer diagnosis
- Advanced MRI Techniques and Applications
- Meningioma and schwannoma management
- Advanced Neural Network Applications
- Brain Tumor Detection and Classification
- Advanced Neuroimaging Techniques and Applications
- Privacy-Preserving Technologies in Data
- Ferroptosis and cancer prognosis
- Surgical Simulation and Training
- Medical Image Segmentation Techniques
- Medical Imaging and Analysis
- Brain Metastases and Treatment
- AI in cancer detection
- Organ Donation and Transplantation
- Domain Adaptation and Few-Shot Learning
- Advanced Fluorescence Microscopy Techniques
- Mental Health and Psychiatry
- Mathematical Biology Tumor Growth
- Augmented Reality Applications
- Machine Learning and ELM
- Photodynamic Therapy Research Studies
- S100 Proteins and Annexins
Parnassia Groep
2023-2024
Erasmus MC
2015-2024
Erasmus University Rotterdam
2018-2024
Erasmus MC Cancer Institute
2019
Brigham and Women's Hospital
2015
Harvard University
2015
Abstract Background Accurate characterization of glioma is crucial for clinical decision making. A delineation the tumor also desirable in initial stages but time-consuming. Previously, deep learning methods have been developed that can either non-invasively predict genetic or histological features glioma, automatically delineate tumor, not both tasks at same time. Here, we present our method molecular subtype and grade, while simultaneously providing a tumor. Methods We single multi-task...
Abstract Purpose: Patients with 1p/19q codeleted low-grade glioma (LGG) have longer overall survival and better treatment response than patients intact tumors. Therefore, it is relevant to know the status. To investigate whether status can be assessed prior tumor resection, we developed a machine learning algorithm predict of presumed LGG based on preoperative MRI. Experimental Design: Preoperative brain MR images from 284 who had undergone biopsy or resection were used train support vector...
Background The association between contrast enhanced (CE) and non-contrast (NCE) tumor resection survival in patients with glioblastoma relation to molecular subtypes is poorly understood. aim of this study was assess the CE NCE light MGMT promoter methylation newly diagnosed IDH-wildtype glioblastoma. Materials Methods Patients who underwent surgery were eligible. volumes assessed on pre- post-operative MRI scans extent calculated. evaluated using multivariable Cox proportional hazards...
Intraoperative MRI and 5-aminolaevulinic acid guided surgery are useful to maximize the extent of glioblastoma resection. ultrasound is used as a time-and cost-effective alternative, but its value has never been assessed in trial. The goal this randomized controlled trial was assess intraoperative B-mode on resection.In trial, patients 18 years or older with newly diagnosed presumed glioblastoma, deemed totally resectable, presenting at Erasmus MC (Rotterdam, Netherlands) were enrolled (1:1)...
In recent years, deep learning has become the leading method for medical image segmentation. While majority of studies focus on developments network architectures, several have shown that non-architectural factors also play a substantial role in performance improvement. An important factor is preprocessing. However, there no agreement which preprocessing steps work best different applications. The aim this study was to investigate effect model performance. To end, we conducted systematic...
<b>Background</b> To achieve maximal resection with minimal risk of postoperative neurologic morbidity, different neurosurgical adjuncts are being used during low-grade glioma (LGG) surgery. <b>Objectives</b> investigate the effect pre- and intraoperative on extent (EOR) hemispheric LGGs. <b>Methods</b> Medical records were reviewed to identify patients any sex, ≥ 18 years age, who underwent LGG surgery at X Hospital between January 2005 July 2013. Patients divided into eight subgroups based...
The Erasmus Glioma Database (EGD) contains structural magnetic resonance imaging (MRI) scans, genetic and histological features (specifying the WHO 2016 subtype), whole tumor segmentations of patients with glioma. Pre-operative MRI data 774 glioma (281 female, 492 male, 1 unknown, age range 19-86 years) treated at MC between 2008 2018 is available. For all a pre-contrast T1-weighted, post-contrast T2-weighted, T2-weighted FLAIR scan are available, made on variety scanners from four different...
Prognostication of glioblastoma survival has become more refined due to the molecular reclassification these tumors into isocitrate dehydrogenase (IDH) wild-type and IDH mutant. Since this stratification, however, robust clinical prediction models relevant entire patient population are lacking. This study aimed provide an updated model that predicts individual prognosis in patients with glioblastoma.
In this study, we used the vessel size imaging (VSI) MRI technique to characterize microvasculature features of three subtypes adult-type diffuse glioma lacking enhancement. Thirty-eight patients with confirmed non-enhancing were categorized into subtypes: Oligo (IDH-mut&1p/19q-codeleted), Astro (IDH-mut), and GBM (IDH-wt). The VSI provided quantitative maps cerebral blood volume (CBV), (µCBV), for each patient. Additionally, tissue samples 21 histopathologically analyzed, quantified. Both...
Abstract Background We aimed to describe the microvascular features of three types adult-type diffuse glioma by comparing dynamic susceptibility contrast (DSC) perfusion magnetic resonance imaging (MRI) with intraoperative high-frame-rate ultrafast Doppler ultrasound. Methods Case series seven patients primary brain tumours underwent both DSC MRI and intra-operative From ultrasound images, three-dimensional vessel segmentation was obtained tumour vascular bed. Relative cerebral blood volume...
Amide proton transfer (APT) weighted chemical exchange saturation (CEST) imaging is increasingly used to investigate high-grade, enhancing brain tumours. Non-enhancing glioma currently less studied, but shows heterogeneous pathophysiology with subtypes having equally poor prognosis as glioma. Here, we the use of CEST MRI best differentiate non-enhancing from healthy tissue and image tumour heterogeneity.A 3D pulsed sequence was applied at 3 Tesla whole coverage 31 off-resonance frequencies...
The authors conducted a study to determine whether cognitive functioning of patients with presumed low-grade glioma is associated white matter (WM) tract changes.
Abstract Background Nonenhancing glioma typically have a favorable outcome, but approximately 19–44% highly aggressive course due to glioblastoma genetic profile. The aim of this retrospective study is use physiological MRI parameters both perfusion and diffusion distinguish the molecular profiles without enhancement at presentation. Methods Ninety-nine patients with nonenhancing were included, in whom status (including 1p/19q codeletion IDH mutation) preoperative (T2w/FLAIR, dynamic...
Tumor growth models have the potential to model and predict spatiotemporal evolution of glioma in individual patients. Infiltration cells is known be faster along white matter tracts, therefore structural magnetic resonance imaging (MRI) diffusion tensor (DTI) can used inform model. However, applying evaluating real patient data challenging. In this work, we propose formulate problem tumor as a ranking problem, opposed segmentation use average precision (AP) performance metric. This enables...
Introduction O6-methylguanine-methyltransferase (MGMT) promotor methylation and isocitrate dehydrogenase (IDH) mutation status are important prognostic factors for patients with glioblastoma. There conflicting reports about a differential topographical distribution of glioblastoma vs. without MGMT methylation, possibly caused by molecular heterogeneity in populations. We initiated this study to re-evaluate the light updated WHO 2016 classification. Methods Pre-operative T2-weighted/FLAIR...
When finetuning a convolutional neural network (CNN) on data from new domain, catastrophic forgetting will reduce performance the original training data. Elastic Weight Consolidation (EWC) is recent technique to prevent this, which we evaluated while and re-training CNN segment glioma two different datasets. The was trained public BraTS dataset finetuned an in-house with non-enhancing low-grade glioma. EWC found decrease in this case, but also restrict adaptation domain.
Whether the Swiss psychiatrist Carl Jung (1875–1961) became psychotic after his mid-thirties is much debated. His recently published Black Books, a seven-volume journal, reveal new insights into this debate. Based on phenomenological analysis of self-reports in these books and other writings, we here identify several types anomalous perceptual experiences: hypnagogic-hypnopompic experiences, hyperphantasia, hallucinations, personifications, sensed presence. We argue that experiences were not...
Accurate characterization of glioma is crucial for clinical decision making. A delineation the tumor also desirable in initial stages but a time-consuming task. Leveraging latest GPU capabilities, we developed single multi-task convolutional neural network that uses full 3D, structural, pre-operative MRI scans to can predict IDH mutation status, 1p/19q co-deletion and grade tumor, while simultaneously segmenting tumor. We trained our method using largest, most diverse patient cohort date...