Abstract. Simple algorithms for segmenting healthy lung parenchyma in CT are unable to deal with high density tissue common in pulmonary diseases. To overcome this problem, we propose a multi-stage learningbased approach that combines anatomical information to predict an initialization of a statistical shape model of the lungs. The initialization first detects the carina of the trachea, and uses this to detect a set of automatically selected stable landmarks on regions near the lung (e.g., ribs, spine). These landmarks are used to align the shape model, which is then refined through boundary detection to obtain fine-grained segmentation. Robustness is obtained through hierarchical use of discriminative classifiers that are trained on a range of manually annotated data of diseased and healthy lungs. We demonstrate fast detection (35s per volume on average) and segmentation of 2 mm accuracy on challenging data.
We demonstrate single-breath-hold 3-D CINE imaging in volunteers and three example patient cases, which features fast reconstruction and allows reformatting to arbitrary orientations.
Abstract. We present a novel generic segmentation system for the fully automatic multi-organ segmentation from CT medical images. Thereby we combine the advantages of learning-based approaches on point cloudbased shape representation, such a speed, robustness, point correspondences, with those of PDE-optimization-based level set approaches, such as high accuracy and the straightforward prevention of segment overlaps. In a benchmark on 10-100 annotated datasets for the liver, the lungs, and the kidneys we show that the proposed system yields segmentation accuracies of 1.17-2.89mm average surface errors. Thereby the level set segmentation (which is initialized by the learning-based segmentations) contributes with an 20%-40% increase in accuracy.
Iterative reconstruction can substantially improve high-resolution dynamic CE-MRA image quality, most notably in small to mid-size vasculature. Dynamic CE-MRA with iterative reconstruction could become an alternative to conventional static 3D CE-MRA, thus simplifying the clinical workflow. Magn Reson Med 77:833-840, 2017. © 2016 International Society for Magnetic Resonance in Medicine.
A novel reconstruction using both spatial and spectral regularization allows obtaining accurate FF and [Formula: see text] maps for prospectively highly accelerated acquisitions.
Background: Obesity has become an epidemic in China with its increased prevalence, especially in the male population. Disparities in fat distribution rather than increasing body mass index (BMI) confer the risk of different diseases, including cardiac abnormalities. Therefore, early detection of cardiac abnormalities is important for treatment to reverse the progression to heart failure. Nowadays, strain analysis based on cardiac magnetic resonance (CMR) imaging has been established to assess myocardial function in diverse cardiac diseases. We aimed to assess the relationship between fat distribution and subclinical diastolic dysfunction in obese Chinese men assessed by deformation registration algorithm (DRA)-based myocardial strain rate (SR) analysis. Methods: A total of 115 male participants with different BMI underwent CMR scanning using a 1.5T MAGNETOM Aera (Siemens Healthcare, Erlangen, Germany) and computed tomography (CT) scan. All the participants were enrolled from September 2017 to April 2018. They were classified into 3 groups according to their BMIs with 23 and 27.5 kg/m 2 being the cutoff values. A Trufi-Strain prototype software (version 2.0, Siemens Healthcare, Erlangen, Germany) was used to quantify SR in both early and late diastole from CMR cine images. Ratios of early and late SRs were calculated. Areas of epicardial and pericardial adipose tissue (EAT and PAT) were measured on a single 4-chamber-view slice of cine images. Volumes of visceral and subcutaneous adipose tissue (VAT and SAT) were acquired semi-automatically from CT images using the dedicated software Cardiac Risk2.0 (Siemens Healthcare). Waist and hip circumferences were manually measured (WC and HC). Analysis of variance or nonparametric tests, along with correlation and stepwise multivariate regression analysis models, was applied for statistical analysis. Results: Peak late diastolic SRs were higher in obese men compared with their lean counterparts [−36.25±10.46 vs. −29.46±8.17, 66.97±18.58 vs. 45.62 (42.44, 55.96), and 56.81±15.07 vs. 41.40±6.41 for radial, circumferential, and longitudinal SRs, respectively; P<0.05]. All SR ratios in the obese subgroups were lower than those of lean men (3.12±1.14 vs. 4.63±1.24, 2.12±0.58 vs. 2.96±0.62 and 1.63±0.50 vs. 2.20±0.63 for radial, circumferential, and longitudinal directions, respectively; P<0.05). EAT was a significant predictor of diastolic function assessed by radial and circumferential SR ratios (β=−0.439 and −0.337 respectively; all P<0.001), while VAT was a significant predictor of circumferential and longitudinal SR ratios (β=−0.216 and −0.355, respectively, P<0.05). Conclusions: Decreased LV diastolic function assessed by DRA-based SR analysis in obesity is associated with fat distribution. Furthermore, EAT and VAT might be better predictors of a decrease of diastolic function in obese Chinese men than BMI.
We present a software, called CoroEval, for the evaluation of 3D coronary vessel reconstructions from clinical data. It runs on multiple operating systems and is designed to be independent of the imaging modality used. At this point, its purpose is the comparison of reconstruction algorithms or acquisition protocols, not the clinical diagnosis. Implemented metrics are vessel sharpness and diameter. All measurements are taken from the raw intensity data to be independent of display windowing functions. The user can either import a vessel centreline segmentation from other software, or perform a manual segmentation in CoroEval. An automated segmentation correction algorithm is provided to improve non-perfect centrelines. With default settings, measurements are taken at 1 mm intervals along the vessel centreline and from 10 different angles at each measurement point. This allows for outlier detection and noise-robust measurements without the burden and subjectivity a manual measurement process would incur. Graphical measurement results can be directly exported to vector or bitmap graphics for integration into scientific publications. Centreline and lumen segmentations can be exported as point clouds and in various mesh formats. We evaluated the diameter measurement process using three phantom datasets. An average deviation of 0.03 ± 0.03 mm was found. The software is available in binary and source code form at http://www5.cs.fau.de/CoroEval/.
Cardiovascular magnetic resonance imaging is the gold standard for cardiac function assessment. Quantification of clinical results (CR) requires precise segmentation. Clinicians statistically compare CRs to ensure reproducibility. Convolutional Neural Network developers compare their results via metrics. Aim: Introducing software capable of automatic multilevel comparison. A multilevel analysis covering segmentations and CRs builds on a generic software backend. Metrics and CRs are calculated with geometric accuracy. Segmentations and CRs are connected to track errors and their effects. An interactive GUI makes the software accessible to different users. The software’s multilevel comparison was tested on a use case based on cardiac function assessment. The software shows good reader agreement in CRs and segmentation metrics (Dice > 90%). Decomposing differences by cardiac position revealed excellent agreement in midventricular slices: > 90% but poorer segmentations in apical (> 71%) and basal slices (> 74%). Further decomposition by contour type locates the largest millilitre differences in the basal right cavity (> 3 ml). Visual inspection shows these differences being caused by different basal slice choices. The software illuminated reader differences on several levels. Producing spreadsheets and figures concerning metric values and CR differences was automated. A multilevel reader comparison is feasible and extendable to other cardiac structures in the future.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.