Time filter

Source Type

Rotterdam, Netherlands

We present a method for carotid vessel wall volume quantification from magnetic resonance imaging (MRI). The method combines lumen and outer wall segmentation based on deformable model fitting with a learning-based segmentation correction step. After selecting two initialization points, the vessel wall volume in a region around the bifurcation is automatically determined. The method was trained on eight datasets (16 carotids) from a population-based study in the elderly for which one observer manually annotated both the lumen and outer wall. An evaluation was carried out on a separate set of 19 datasets (38 carotids) from the same study for which two observers made annotations. Wall volume and normalized wall index measurements resulting from the manual annotations were compared to the automatic measurements. Our experiments show that the automatic method performs comparably to the manual measurements. All image data and annotations used in this study together with the measurements are made available through the website http://ergocar.bigr.nl. Source

Van Der Giessen A.G.,Massachusetts General Hospital | Van Der Giessen A.G.,Erasmus University Rotterdam | Toepker M.H.,Massachusetts General Hospital | Donelly P.M.,Massachusetts General Hospital | And 10 more authors.
Investigative Radiology

Purpose: To determine the reproducibility, accuracy, and predictors of accuracy of computed tomography (CT) angiography to detect and characterize coronary atherosclerotic plaque as compared with intravascular ultrasound. Methods: Ten ex vivo human coronary arteries were imaged in a moving phantom by dual-source CT (collimation: 0.6 mm, reconstructed slice thickness: 0.4 mm) and intravascular ultrasound (IVUS). Coregistered cross-sections were assessed at 0.4 mm intervals for the presence and composition of atherosclerotic plaque (noncalcified, mixed, and calcified) on CT and IVUS by independent readers to determine reader agreement and diagnostic accuracy. Quantitative measurements of lumen and plaque area, plaque eccentricity, and intimal thickness on IVUS were used to determine predictors for the detection of noncalcified plaque by CT. Results: Within 1002 coregistered cross-sections, the interobserver agreement to detect plaque on CT was K = 0.48, K = 0.42, and K = 1.00 for noncalcified, mixed, and calcified plaque; respectively. The sensitivity and specificity of CT was 57% out of 84% for noncalcified, 32% of 92% for mixed, and 56% of 93% for calcified plaque when compared with IVUS; respectively. Misclassification occurred in 68% of mixed and 43% of noncalcified plaques. The odds of detecting noncalcified plaque in CT independently increased by 56% (95% CI: 47%-77%, P < 0.0001) with every 0.1 mm increase in maximum intimal thickness as measured by IVUS. Detection rate for noncalcified plaques was poor for plaques <1 mm (36%) but excellent for plaques >1 mm maximal intimal thickness (90%). Conclusion: Reader agreement and diagnostic accuracy for the detection of coronary atherosclerotic plaque vary with plaque composition. Intimal thickness independently predicts detection of noncalcified plaque by CT with excellent sensitivity for >1 mm thick plaques. Copyright © 2010 by Lippincott Williams and Wilkins. Source

Zahnd G.,CNRS Research Center for Image Acquisition and Processing for Health | Zahnd G.,Biomedical Imaging Group Rotterdam | Orkisz M.,CNRS Research Center for Image Acquisition and Processing for Health | Serusclat A.,Louis Pradel Hospital | And 3 more authors.
International Journal of Computer Assisted Radiology and Surgery

Objectives: The aim of this work is to present and evaluate a novel segmentation method for localizing the contours of the intima-media complex in the carotid artery wall through longitudinal ultrasound B-mode imaging. Themethod is used to investigate the association between atherosclerosis risk factors and the cyclic variation of the intima-media thickness during the heart beat. Methods: The framework introduced is based on two main features. The first is a simultaneous extraction of both the lumen-intima and the media-adventitia interfaces, using the combination of an original shape-adapted filter bank and a specific dynamic programming scheme. The second is an innovative spatial transformation that eases the extraction of skewed and curved contours, and exploits the result from the previous image as a priori information, when processing the current image. The intima-media thickness is automatically derived from the estimated contours for each time step during the cardiac cycle. Our method was evaluated in vivo on 57 healthy volunteers and 25 patients at high cardiovascular risk. Reference contours were generated for each subject by averaging the tracings performed by three experienced observers. Results: Segmentation errors were 29 ± 27 μm for the lumen-intima interface, 42 ± 38 μm for the media-adventitia interface, and 22 ± 16 μm for the intima-media thickness. This uncertainty was similar to inter- and intra-observer variability. Furthermore, the amplitude of the temporal variation in thickness of the intima-media layers during the cardiac cycle was significantly higher in at-risk patients compared to healthy volunteers (79 ± 36 vs. 64 ± 26 μm, p = 0.032). Conclusion: The method proposed may provide a relevant diagnostic aid for atherosclerosis screening in clinical studies. © CARS 2013. Source

Stepanova T.,Erasmus MC | Smal I.,Biomedical Imaging Group Rotterdam | Van Haren J.,Erasmus MC | Akinci U.,Erasmus MC | And 11 more authors.
Current Biology

In Chinese hamster ovary cells, microtubules originate at the microtubule organizing center (MTOC) and grow persistently toward the cell edge, where they undergo catastrophe [1]. In axons, microtubule dynamics must be regulated differently because microtubules grow parallel to the plasma membrane and there is no MTOC. GFP-tagged microtubule plus end tracking proteins (+TIPs) mark the ends of growing neuronal microtubules [2]. Their fluorescent "comet-like" pattern reflects turnover of +TIP binding sites [3, 4]. Using GFP-tagged +TIPs and fluorescence-based segmentation and tracking tools, we show that axonal microtubules grow with a constant average velocity and that they undergo catastrophes at random positions, yet in a programmed fashion. Using protein depletion approaches, we find that the +TIPs CLIP-115 and CLIP-170 affect average microtubule growth rate and growth distance in neurons but not the duration of a microtubule growth event. In N1E-115 neuroblastoma cells, we find that EB1, the core +TIP [5], regulates microtubule growth rate, growth distance, and duration, consistent with in vitro data [6]. Combined, our data suggest that CLIPs influence the axonal microtubule/tubulin ratio, whereas EB1 stimulates microtubule growth and structural transitions at microtubule ends, thereby regulating microtubule catastrophes and the turnover of +TIP binding sites. © 2010 Elsevier Ltd. All rights reserved. Source

Sun W.,Biomedical Imaging Group Rotterdam | Niessen W.J.,Biomedical Imaging Group Rotterdam | Niessen W.J.,Technical University of Delft | Van Stralen M.,University Utrecht | Klein S.,Biomedical Imaging Group Rotterdam
IEEE Transactions on Image Processing

Multiresolution strategies are commonly used in the nonrigid registration to avoid local minima in the optimization space. Generally, a step-by-step hierarchical approach is adopted, in which the registration starts on a level with reduced complexity (downsampled images, global transformations), then continuing to levels with increased complexity, until the finest level is reached. In this paper, we propose two alternative multiresolution strategies for both the data and transformation models, in which different resolution levels are considered simultaneously instead of subsequently. Through combining the different strategies for data and transformation, we systematically define 3×3 multiresolution schemes, including both existing and novel methods. Experiments on 10 pairs of computed tomography lung data sets showed that the best performing strategy resulted in a reduction of the upper quartile of the mean target registration error from 2 to 1.5 mm, compared with the conventionally hierarchical multiresolution method, while achieving smoother deformations. Experiments with intersubject registration of 18 3D T1-weighted MRI brain scans confirmed that simultaneous multiresolution strategies produce more accurate registration results (median of mean overlap increased from 0.55 to 0.57) and smoother deformation fields than the traditionally hierarchical method. Evaluation of robustness indicated that the largest differences in accuracy between methods are observed for structures with a relatively large initial misalignment. © 1992-2012 IEEE. Source

Discover hidden collaborations