A5065796620- Medical Image Segmentation Techniques
- Medical Imaging Techniques and Applications
- Radiomics and Machine Learning in Medical Imaging
- Advanced Radiotherapy Techniques
- Cardiac Imaging and Diagnostics
- Medical Imaging and Analysis
- Esophageal Cancer Research and Treatment
- Coronary Interventions and Diagnostics
- Lung Cancer Diagnosis and Treatment
- Advanced MRI Techniques and Applications
- MRI in cancer diagnosis
- Cerebrovascular and Carotid Artery Diseases
- Advanced X-ray and CT Imaging
- Prostate Cancer Diagnosis and Treatment
- Advanced Neuroimaging Techniques and Applications
- Cardiovascular Disease and Adiposity
- Advanced Neural Network Applications
- Brain Tumor Detection and Classification
- Advanced Image Processing Techniques
- Colorectal Cancer Surgical Treatments
- COVID-19 diagnosis using AI
- AI in cancer detection
- Bioinformatics and Genomic Networks
- Photoacoustic and Ultrasonic Imaging
- 3D Shape Modeling and Analysis
Centre for Medical Systems Biology
2022-2024
Leiden University Medical Center
2016-2023
Purpose To develop and validate a robust accurate registration pipeline for automatic contour propagation online adaptive Intensity‐Modulated Proton Therapy (IMPT) of prostate cancer using elastix software deep learning. Methods A three‐dimensional (3D) Convolutional Neural Network was trained bladder segmentation the computed tomography (CT) scans. The alongside scan is jointly optimized to add explicit knowledge about underlying anatomy algorithm. We included three datasets from different...
We propose a supervised nonrigid image registration method, trained using artificial displacement vector fields (DVF), for which we and compare three network architectures. The DVFs allow training in fully voxel-wise dense manner, but without the cost usually associated with creation of densely labeled data. scheme to artificially generate DVFs, chest CT augment these simulated respiratory motion. proposed architectures are embedded multi-stage approach, increase capture range networks order...
Manual or automatic delineation of the esophageal tumor in CT images is known to be very challenging. This due low contrast between and adjacent tissues, anatomical variation esophagus, as well occasional presence foreign bodies (e.g. feeding tubes). Physicians therefore usually exploit additional knowledge such endoscopic findings, clinical history, imaging modalities like PET scans. Achieving his information time-consuming, while results are error-prone might lead non-deterministic...
Due to the intricate relationship between pelvic organs and vital structures, such as vessels nerves, anatomy is often considered be complex comprehend. In oncological surgery, a trade-off has made complete tumor resection preserving function by preventing damage nerves. Damage autonomic nerves causes undesirable post-operative side-effects fecal urinal incontinence, well sexual dysfunction in up 80 percent of cases. Since these are not visible pre-operative MRI scans or during avoiding...
In this paper we propose a supervised method to predict registration misalignment using convolutional neural networks (CNNs). This task is casted classification problem with multiple classes of misalignment: "correct" 0-3 mm, "poor" 3-6 mm and "wrong" over 6 mm. Rather than direct prediction, hierarchical approach, where the prediction gradually refined from coarse fine. Our solution based on Long Short-Term Memory (LSTM), predictions three resolutions image pair, leveraging intrinsic...
Medical image registration and segmentation are two of the most frequent tasks in medical analysis. As these complementary correlated, it would be beneficial to apply them simultaneously a joint manner. In this paper, we formulate as problem via Multi-Task Learning (MTL) setting, allowing leverage their strengths mitigate weaknesses through sharing information. We propose merge not only on loss level, but architectural level well. studied approach context adaptive image-guided radiotherapy...
Abstract Aims Coronary computed tomography angiography (CCTA) is inferior to intravascular imaging in detecting plaque morphology and quantifying burden. We aim to, for the first time, train a deep-learning (DL) methodology accurate quantification characterization CCTA using near-infrared spectroscopy–intravascular ultrasound (NIRS–IVUS). Methods results Seventy patients were prospectively recruited who underwent NIRS–IVUS imaging. Corresponding cross sections matched an in-house developed...
Coronary computed tomography angiography (CCTA) has been proposed as an alternative to intravascular imaging for assessing plaque pathology.
Abstract Purpose The assessment of vulnerable plaque characteristics and distribution is important to stratify cardiovascular risk in a patient. Computed tomography angiography (CTA) offers promising alternative invasive imaging but limited by the fact that range Hounsfield units (HU) lipid-rich areas overlaps with HU fibrotic tissue calcified plaques contrast within contrast-filled lumen. This paper investigate whether can be detected more accurately on cross-sectional CTA images using deep...
This study aimed to investigate the impact of calcific (Ca) on efficacy coronary computed angiography (CTA) in evaluating plaque burden (PB) and composition with near-infrared spectroscopy-intravascular ultrasound (NIRS-IVUS) serving as reference standard.
Positron Emission Tomography (PET) is an imaging method that can assess physiological function rather than structural disturbances by measuring cerebral perfusion or glucose consumption. However, this technique relies on injection of radioactive tracers and expensive. On the contrary, Arterial Spin Labeling (ASL) MRI a non-invasive, non-radioactive, relatively cheap for brain hemodynamic measurements, which allows quantification to some extent. In paper we propose convolutional neural...