Prospective Deep Learning-based Quantitative Assessment of Coronary Plaque by CT Angiography Compared with Intravascular Ultrasound.

AIMS: Coronary computed tomography angiography provides noninvasive assessment of coronary stenosis severity and flow impairment. Automated artificial intelligence analysis may assist in precise quantification and characterization of coronary atherosclerosis, enabling patient-specific risk determination and management strategies. This multicenter international study compared an automated deep-learning-based method for segmenting coronary atherosclerosis in coronary computed tomography angiography (CCTA) against the reference standard of intravascular ultrasound (IVUS). METHODS AND RESULTS: The study included clinically stable patients with known coronary artery disease from 15 centers in the U.S. and Japan. An artificial intelligence (AI)-enabled plaque analysis service was utilized to quantify and characterize total plaque (TPV), vessel, lumen, calcified plaque (CP), non-calcified plaque (NCP), and low attenuation plaque (LAP) volumes derived from CCTA and compared with IVUS measurements in a blinded, core laboratory-adjudicated fashion. In 237 patients, 432 lesions were assessed; mean lesion length was 24.5 mm. Mean IVUS-TPV was 186.0 mm3. AI-enabled plaque analysis on CCTA showed strong correlation and high accuracy when compared with IVUS; correlation coefficient, slope, and Y intercept for TPV were 0.91, 0.99, and 1.87, respectively; for CP volume 0.91, 1.05, and 5.32, respectively; and for NCP volume 0.87, 0.98, and 15.24, respectively. Bland-Altman analysis demonstrated strong agreement with little bias for these measurements. CONCLUSIONS: Artificial intelligence enabled CCTA quantification and characterization of atherosclerosis demonstrated strong agreement with IVUS reference standard measurements. This tool may prove effective for accurate evaluation of coronary atherosclerotic burden and cardiovascular risk assessment.[ClinicalTrails.gov identifier: NCT05138289].

Long-term prognostic implications of hemodynamic and plaque assessment using coronary CT angiography.

BACKGROUND AND AIMS: Hemodynamic and plaque characteristics can be analyzed using coronary CT angiography (CTA). We aimed to explore long-term prognostic implications of hemodynamic and plaque characteristics using coronary CT angiography (CTA). METHODS: Invasive fractional flow reserve (FFR) and CTA-derived FFR (FFRCT) were undertaken for 136 lesions in 78 vessels and followed-up to 10 years until December 2020. FFRCT, wall shear stress (WSS), change in FFRCT across the lesion (ΔFFRCT), total plaque volume (TPV), percent atheroma volume (PAV), and low-attenuation plaque volume (LAPV) for target lesions [L] and vessels [V] were obtained by independent core laboratories. Their collective influence was evaluated for the clinical endpoints of target vessel failure (TVF) and target lesion failure (TLF). RESULTS: During a median follow-up of 10.1 years, PAV[V] (per 10% increase, HR 2.32 [95% CI 1.11-4.86], p = 0.025), and FFRCT[V] (per 0.1 increase, HR 0.56 [95% CI 0.37-0.84], p = 0.006) were independent predictors of TVF for the per-vessel analysis, and WSS[L] (per 100 dyne/cm2 increase, HR 1.43 [1.09-1.88], p = 0.010), LAPV[L] (per 10 mm3 increase, HR 3.81 [1.16-12.5], p = 0.028), and ΔFFRCT[L] (per 0.1 increase, HR 1.39 [1.02-1.90], p = 0.040) were independent predictors of TLF for the per-lesion analysis after adjustment for clinical and lesion characteristics. The addition of both plaque and hemodynamic predictors improved the predictability for 10-year TVF and TLF of clinical and lesion characteristics (all p < 0.05). CONCLUSIONS: Vessel- and lesion-level hemodynamic characteristics, and vessel-level plaque quantity, and lesion-level plaque compositional characteristics assessed by CTA offer independent and additive long-term prognostic value.

Atlas-ISTN: Joint segmentation, registration and atlas construction with image-and-spatial transformer networks.

Deep learning models for semantic segmentation are able to learn powerful representations for pixel-wise predictions, but are sensitive to noise at test time and may lead to implausible topologies. Image registration models on the other hand are able to warp known topologies to target images as a means of segmentation, but typically require large amounts of training data, and have not widely been benchmarked against pixel-wise segmentation models. We propose the Atlas Image-and-Spatial Transformer Network (Atlas-ISTN), a framework that jointly learns segmentation and registration on 2D and 3D image data, and constructs a population-derived atlas in the process. Atlas-ISTN learns to segment multiple structures of interest and to register the constructed atlas labelmap to an intermediate pixel-wise segmentation. Additionally, Atlas-ISTN allows for test time refinement of the model's parameters to optimize the alignment of the atlas labelmap to an intermediate pixel-wise segmentation. This process both mitigates for noise in the target image that can result in spurious pixel-wise predictions, as well as improves upon the one-pass prediction of the model. Benefits of the Atlas-ISTN framework are demonstrated qualitatively and quantitatively on 2D synthetic data and 3D cardiac computed tomography and brain magnetic resonance image data, out-performing both segmentation and registration baseline models. Atlas-ISTN also provides inter-subject correspondence of the structures of interest.

Myocardial Perfusion Simulation for Coronary Artery Disease: A Coupled Patient-Specific Multiscale Model.

Patient-specific models of blood flow are being used clinically to diagnose and plan treatment for coronary artery disease. A remaining challenge is bridging scales from flow in arteries to the micro-circulation supplying the myocardium. Previously proposed models are descriptive rather than predictive and have not been applied to human data. The goal here is to develop a multiscale patient-specific model enabling blood flow simulation from large coronary arteries to myocardial tissue. Patient vasculatures are segmented from coronary computed tomography angiography data and extended from the image-based model down to the arteriole level using a space-filling forest of synthetic trees. Blood flow is modeled by coupling a 1D model of the coronary arteries to a single-compartment Darcy myocardium model. Simulated results on five patients with non-obstructive coronary artery disease compare overall well to [[Formula: see text]O][Formula: see text]O PET exam data for both resting and hyperemic conditions. Results on a patient with severe obstructive disease link coronary artery narrowing with impaired myocardial blood flow, demonstrating the model's ability to predict myocardial regions with perfusion deficit. This is the first report of a computational model for simulating blood flow from the epicardial coronary arteries to the left ventricle myocardium applied to and validated on human data.

Precise measurement of coronary stenosis diameter with CCTA using CT number calibration.

PURPOSE: Coronary x-ray computed tomography angiography (CCTA) continues to develop as a noninvasive method for the assessment of coronary vessel geometry and the identification of physiologically significant lesions. The uncertainty of quantitative lesion diameter measurement due to limited spatial resolution and vessel motion reduces the accuracy of CCTA diagnoses. In this paper, we introduce a new technique called computed tomography (CT)-number-Calibrated Diameter to improve the accuracy of the vessel and stenosis diameter measurements with CCTA. METHODS: A calibration phantom containing cylindrical holes (diameters spanning from 0.8 mm through 4.0 mm) capturing the range of diameters found in human coronary vessels was three-dimensional printed. We also printed a human stenosis phantom with 17 tubular channels having the geometry of lesions derived from patient data. We acquired CT scans of the two phantoms with seven different imaging protocols. Calibration curves relating vessel intraluminal maximum voxel value (maximum CT number of a voxel, described in Hounsfield Units, HU) to true diameter, and full-width-at-half maximum (FWHM) to true diameter were constructed for each CCTA protocol. In addition, we acquired scans with a small constant motion (15 mm/s) and used a motion correction reconstruction (Snapshot Freeze) algorithm to correct motion artifacts. We applied our technique to measure the lesion diameter in the 17 lesions in the stenosis phantom and compared the performance of CT-number-Calibrated Diameter to the ground truth diameter and a FWHM estimate. RESULTS: In all cases, vessel intraluminal maximum voxel value vs diameter was found to have a simple functional form based on the two-dimensional point spread function yielding a constant maximum voxel value region above a cutoff diameter, and a decreasing maximum voxel value vs decreasing diameter below a cutoff diameter. After normalization, focal spot size and reconstruction kernel were the principal determinants of cutoff diameter and the rate of maximum voxel value reduction vs decreasing diameter. The small constant motion had a significant effect on the CT number calibration; however, the motion-correction algorithm returned the maximum voxel value vs diameter curve to that of stationary vessels. The CT number Calibration technique showed better performance than FWHM estimation of diameter, yielding a high accuracy in the tested range (0.8 mm through 2.5 mm). We found a strong linear correlation between the smallest diameter in each of 17 lesions measured by CT-number-Calibrated Diameter (DC ) and ground truth diameter (Dgt ), (DC  = 0.951 × Dgt  + 0.023 mm, r = 0.998 with a slope very close to 1.0 and intercept very close to 0 mm. CONCLUSIONS: Computed tomography-number-Calibrated Diameter is an effective method to enhance the accuracy of the estimate of small vessel diameters and degree of coronary stenosis in CCTA.

TeTrIS: Template Transformer Networks for Image Segmentation With Shape Priors.

In this paper, we introduce and compare different approaches for incorporating shape prior information into neural network-based image segmentation. Specifically, we introduce the concept of template transformer networks, where a shape template is deformed to match the underlying structure of interest through an end-to-end trained spatial transformer network. This has the advantage of explicitly enforcing shape priors, and this is free of discretization artifacts by providing a soft partial volume segmentation. We also introduce a simple yet effective way of incorporating priors in the state-of-the-art pixel-wise binary classification methods such as fully convolutional networks and U-net. Here, the template shape is given as an additional input channel, incorporating this information significantly reduces false positives. We report results on synthetic data and sub-voxel segmentation of coronary lumen structures in cardiac computed tomography showing the benefit of incorporating priors in neural network-based image segmentation.

Attention gated networks: Learning to leverage salient regions in medical images.

We propose a novel attention gate (AG) model for medical image analysis that automatically learns to focus on target structures of varying shapes and sizes. Models trained with AGs implicitly learn to suppress irrelevant regions in an input image while highlighting salient features useful for a specific task. This enables us to eliminate the necessity of using explicit external tissue/organ localisation modules when using convolutional neural networks (CNNs). AGs can be easily integrated into standard CNN models such as VGG or U-Net architectures with minimal computational overhead while increasing the model sensitivity and prediction accuracy. The proposed AG models are evaluated on a variety of tasks, including medical image classification and segmentation. For classification, we demonstrate the use case of AGs in scan plane detection for fetal ultrasound screening. We show that the proposed attention mechanism can provide efficient object localisation while improving the overall prediction performance by reducing false positives. For segmentation, the proposed architecture is evaluated on two large 3D CT abdominal datasets with manual annotations for multiple organs. Experimental results show that AG models consistently improve the prediction performance of the base architectures across different datasets and training sizes while preserving computational efficiency. Moreover, AGs guide the model activations to be focused around salient regions, which provides better insights into how model predictions are made. The source code for the proposed AG models is publicly available.

Lumen boundaries extracted from coronary computed tomography angiography on computed fractional flow reserve (FFRCT): validation with optical coherence tomography.

AIMS: The aim of this study was to evaluate the accuracy of minimum lumen area (MLA) by coronary computed tomography angiography (cCTA) and its impact on fractional flow reserve (FFRCT). METHODS AND RESULTS: Fifty-seven patients (118 lesions, 72 vessels) who underwent cCTA and optical coherence tomography (OCT) were enrolled. OCT and cCTA were co-registered and MLAs were measured with both modalities. FFROCT was calculated using OCT-updated models with cCTA-based lumen geometry replaced by OCT-derived geometry. Lesions were grouped by Agatston score (AS) and minimum lumen diameter (MLD) using the OCT catheter and guidewire size (1.0 mm) as a threshold. For all lesions, the average absolute difference between cCTA and OCT MLA was 0.621±0.571 mm2. Pearson correlation coefficients between cCTA and OCT MLAs in lesions with low-intermediate and high AS were 0.873 and 0.787, respectively (both p<0.0001). Irrespective of AS score, excellent correlations were observed for MLA (r=0.839, p<0.0001) and FFR comparisons (r=0.918, p<0.0001) in lesions with MLD ≥1.0 mm but not for lesions with MLD <1.0 mm. CONCLUSIONS: The spatial resolution of cCTA or calcification does not practically limit the accuracy of lumen boundary identification by cCTA or FFRCT calculations for MLD ≥1.0 mm. The accuracy of cCTA MLA could not be adequately assessed for lesions with MLD <1.0 mm.

Generation of Patient-Specific Cardiac Vascular Networks: A Hybrid Image-Based and Synthetic Geometric Model.

OBJECTIVE: In this paper, we propose an algorithm for the generation of a patient-specific cardiac vascular network starting from segmented epicardial vessels down to the arterioles. METHOD: We extend a tree generation method based on satisfaction of functional principles, named constrained constructive optimization, to account for multiple, competing vascular trees. The algorithm simulates angiogenesis under vascular volume minimization with flow-related and geometrical constraints adapting the simultaneous tree growths to patient priors. The generated trees fill the entire left ventricle myocardium up to the arterioles. RESULTS: From actual vascular tree models segmented from CT images, we generated networks with 6000 terminal segments for six patients. These networks contain between 33 and 62 synthetic trees. All vascular models match morphometry properties previously described. CONCLUSION AND SIGNIFICANCE: Image-based models derived from CT angiography are being used clinically to simulate blood flow in the coronary arteries of individual patients to aid in the diagnosis of disease and planning treatments. However, image resolution limits vessel segmentation to larger epicardial arteries. The generated model can be used to simulate the blood flow and derived quantities from the aorta into the myocardium. This is an important step for diagnosis and treatment planning of coronary artery disease.

3D reconstruction techniques of human coronary bifurcations for shear stress computations.

BACKGROUND: Heterogeneity in plaque composition in human coronary artery bifurcations is associated with blood flow induced shear stress. Shear stress is generally determined by combing 3D lumen data and computational fluid dynamics (CFD). We investigated two new procedures to generate 3D lumen reconstructions of coronary artery bifurcations for shear stress computations. METHODS: We imaged 10 patients with multislice computer tomography (MSCT) and intravascular ultrasound (IVUS). The 3D reconstruction of the main branch was based on the fusion of MSCT and IVUS. The proximal part of side branch was reconstructed using IVUS data or MSCT data, resulting in two different reconstructions of the bifurcation region. The distal part of the side branch was based on MSCT data alone. The reconstructed lumen was combined with CFD to determine the shear stress. Low and high shear stress regions were defined and shear stress patterns in the bifurcation regions were investigated. RESULTS: The 3D coronary bifurcations were successfully generated with both reconstruction procedures. The geometrical features of the bifurcation region for the two reconstruction procedures did not reveal appreciable differences. The shear stress maps showed a qualitative agreement, and the low and high shear stress regions were similar in size and average shear stress values were identical. The low and high shear stress regions showed an overlap of approximately 75%. CONCLUSION: Reconstruction of the side branch with MSCT data alone is an adequate technique to study shear stress and wall thickness in the bifurcation region. The reconstruction procedure can be applied to further investigate the effect of shear stress on atherosclerosis in coronary bifurcations.

Automatic segmentation, detection and quantification of coronary artery stenoses on CTA.

Accurate detection and quantification of coronary artery stenoses is an essential requirement for treatment planning of patients with suspected coronary artery disease. We present a method to automatically detect and quantify coronary artery stenoses in computed tomography coronary angiography. First, centerlines are extracted using a two-point minimum cost path approach and a subsequent refinement step. The resulting centerlines are used as an initialization for lumen segmentation, performed using graph cuts. Then, the expected diameter of the healthy lumen is estimated by applying robust kernel regression to the coronary artery lumen diameter profile. Finally, stenoses are detected and quantified by computing the difference between estimated and expected diameter profiles. We evaluated our method using the data provided in the Coronary Artery Stenoses Detection and Quantification Evaluation Framework. Using 30 testing datasets, the method achieved a detection sensitivity of 29% and a positive predictive value (PPV) of 24% as compared to quantitative coronary angiography (QCA), and a sensitivity of 21% and a PPV of 23% as compared manual assessment based on consensus reading of CTA by 3 observers. The stenoses degree was estimated with an absolute average difference of 31%, a root mean square difference of 39.3% when compared to QCA, and a weighted kappa value of 0.29 when compared to CTA. A Dice of 68 and 65% was reported for lumen segmentation of healthy and diseased vessel segments respectively. According to the ranking of the evaluation framework, our method finished fourth for stenosis detection, second for stenosis quantification and second for lumen segmentation.

Registration of 3D+t coronary CTA and monoplane 2D+t X-ray angiography.

A method for registering preoperative 3D+t coronary CTA with intraoperative monoplane 2D+t X-ray angiography images is proposed to improve image guidance during minimally invasive coronary interventions. The method uses a patient-specific dynamic coronary model, which is derived from the CTA scan by centerline extraction and motion estimation. The dynamic coronary model is registered with the 2D+t X-ray sequence, considering multiple X-ray time points concurrently, while taking breathing induced motion into account. Evaluation was performed on 26 datasets of 17 patients by comparing projected model centerlines with manually annotated centerlines in the X-ray images. The proposed 3D+t/2D+t registration method performed better than a 3D/2D registration method with respect to the accuracy and especially the robustness of the registration. Registration with a median error of 1.47 mm was achieved.

Vessel specific coronary artery calcium scoring: an automatic system.

RATIONALE AND OBJECTIVES: The aim of this study was to automatically detect and quantify calcium lesions for the whole heart as well as per coronary artery on non-contrast-enhanced cardiac computed tomographic images. MATERIALS AND METHODS: Imaging data from 366 patients were randomly selected from patients who underwent computed tomographic calcium scoring assessments between July 2004 and May 2009 at Erasmum MC, Rotterdam. These data included data sets with 1.5-mm and 3.0-mm slice spacing reconstructions and were acquired using four different scanners. The scores of manual observers, who annotated the data using commercially available software, served as ground truth. An automatic method for detecting and quantifying calcifications for each of the four main coronary arteries and the whole heart was trained on 209 data sets and tested on 157 data sets. Statistical testing included determining Pearson's correlation coefficients and Bland-Altman analysis to compare performance between the system and ground truth. Wilcoxon's signed-rank test was used to compare the interobserver variability to the system's performance. RESULTS: Automatic detection of calcified objects was achieved with sensitivity of 81.2% per calcified object in the 1.5-mm data set and sensitivity of 86.6% per calcified object in the 3.0-mm data set. The system made an average of 2.5 errors per patient in the 1.5-mm data set and 2.2 errors in the 3.0-mm data set. Pearson's correlation coefficients of 0.97 (P < .001) for both 1.5-mm and 3.0-mm scans with respect to the calcium volume score of the whole heart were found. The average R values over Agatston, mass, and volume scores for each of the arteries (left circumflex coronary artery, right coronary artery, and left main and left anterior descending coronary arteries) were 0.93, 0.96, and 0.99, respectively, for the 1.5-mm scans. Similarly, for 3.0-mm scans, R values were 0.94, 0.94, and 0.99, respectively. Risk category assignment was correct in 95% and 89% of the data sets in the 1.5-mm and 3-mm scans. CONCLUSIONS: An automatic vessel-specific coronary artery calcium scoring system was developed, and its feasibility for calcium scoring in individual vessels and risk category classification has been demonstrated.

Semiautomatic carotid lumen segmentation for quantification of lumen geometry in multispectral MRI.

Quantitative information about the geometry of the carotid artery bifurcation is relevant for investigating the onset and progression of atherosclerotic disease. This paper proposes an automatic approach for quantifying the carotid bifurcation angle, carotid area ratio, carotid bulb size and the vessel tortuosity from multispectral MRI. First, the internal and external carotid centerlines are determined by finding a minimum cost path between user-defined seed points where the local costs are based on medialness and intensity. The minimum cost path algorithm is iteratively applied after curved multi-planar reformatting to refine the centerline. Second, the carotid lumen is segmented using a topology preserving geodesic active contour which is initialized by the extracted centerlines and steered by the MR intensities. Third, the bifurcation angle and vessel tortuosity are automatically extracted from the segmented lumen. The methods for centerline tracking and lumen segmentation are evaluated by comparing their accuracy to the inter- and intra-observer variability on 48 datasets (96 carotid arteries) acquired as part of a longitudinal population study. The evaluation reveals that 94 of 96 carotid arteries are segmented successfully. The distance between the tracked centerlines and the reference standard (0.33 mm) is similar to the inter-observer variation (0.32 mm). The lumen segmentation accuracy (average DSC=0.89, average mean absolute surface distance=0.31 mm) is close to the inter-observer variation (average dice=0.92, average mean surface distance=0.23 mm). The correlation coefficient of manually and automaticly derived bifurcation angle, carotid proximal area ratio, carotid proximal bulb size and vessel totuosity quantifications are close to the correlation of these measures between observers. This demonstrates that the automated method can be used for replacing manual centerline annotation and manual contour drawing for lumen segmentation in MRIs data prior to quantifying the carotid bifurcation geometry.

Regression-based cardiac motion prediction from single-phase CTA.

State of the art cardiac computed tomography (CT) enables the acquisition of imaging data of the heart over the entire cardiac cycle at concurrent high spatial and temporal resolution. However, in clinical practice, acquisition is increasingly limited to 3-D images. Estimating the shape of the cardiac structures throughout the entire cardiac cycle from a 3-D image is therefore useful in applications such as the alignment of preoperative computed tomography angiography (CTA) to intra-operative X-ray images for improved guidance in coronary interventions. We hypothesize that the motion of the heart is partially explained by its shape and therefore investigate the use of three regression methods for motion estimation from single-phase shape information. Quantitative evaluation on 150 4-D CTA images showed a small, but statistically significant, increase in the accuracy of the predicted shape sequences when using any of the regression methods, compared to shape-independent motion prediction by application of the mean motion. The best results were achieved using principal component regression resulting in point-to-point errors of 2.3±0.5 mm, compared to values of 2.7±0.6 mm for shape-independent motion estimation. Finally, we showed that this significant difference withstands small variations in important parameter settings of the landmarking procedure.

Automated bone removal in CT angiography: comparison of methods based on single energy and dual energy scans.

PURPOSE: To evaluate dual energy based methods for bone removal in computed tomography angiography (CTA) images and compare these with single energy based methods that use an additional, nonenhanced, CT scan. METHODS: Four different bone removal methods were applied to CT scans of an anthropomorphic thorax phantom, acquired with a second generation dual source CT scanner. The methods differed by the way information on the presence of bone was obtained (either by using an additional, nonenhanced scan or by scanning with two tube voltages at the same time) and by the way the bone was removed from the CTA images (either by masking or subtracting the bone). The phantom contained parts which mimic vessels of various diameters in direct contact with bone. Both a quantitative and qualitative analysis of image quality after bone removal was performed. Image quality was quantified by the contrast-to-noise ratio (CNR) normalized to the square root of the dose (CNRD). At locations where vessels touch bone, the quality of the bone removal and the vessel preservation were visually assessed. The dual energy based methods were assessed with and without the addition of a 0.4 mm tin filter to the high voltage x-ray tube filtration. For each bone removal method, the dose required to obtain a certain CNR after bone removal was compared with the dose of a reference scan with the same CNR but without automated bone removal. The CNRD value of the reference scan was maximized by choosing the lowest tube voltage available. RESULTS: All methods removed the bone completely. CNRD values were higher for the masking based methods than for the subtraction based methods. Single energy based methods had a higher CNRD value than the corresponding dual energy based methods. For the subtraction based dual energy method, tin filtration improved the CNRD value with approximately 50%. For the masking based dual energy method, it was easier to differentiate between iodine and bone when tin filtration was applied. The CNRD value decreased only with 4% in that case. Compared to the dual scan based methods, the dual energy based methods had the advantage that only a single scan was made without the need of image registration. This might be easier to implement in clinical practice. Vessel preservation was better with bone subtraction than with bone masking. Smaller vessels were completely occluded by the bone mask. None of the bone removal methods was dose neutral. CONCLUSIONS: In general, dual scan based methods that use the lowest tube voltage available, have a higher CNR than the dual energy based approaches at the same dose level. Tin filtration improves the ability to differentiate between iodine and bone for the dual energy based masking method. In clinical practice, the advantages of the dual energy masking method might outweigh its disadvantage of a slightly higher dose penalty compared to the conventional dual scan masking method.

Comparison of shape regression methods under landmark position uncertainty.

Despite the growing interest in regression based shape estimation, no study has yet systematically compared different regression methods for shape estimation. We aimed to fill this gap by comparing linear regression methods with a special focus on shapes with landmark position uncertainties. We investigate two scenarios: In the first, the uncertainty of the landmark positions was similar in the training and test dataset, whereas in the second the uncertainty of the training and test data were different. Both scenarios were tested on simulated data and on statistical models of the left ventricle estimating the end-systolic shape from end-diastole with landmark uncertainties derived from the segmentation process, and of the femur estimating the 3D shape from one projection with landmark uncertainties derived from the imaging setup. Results show that in the first scenario linear regression methods tend to perform similar. In the second scenario including estimates of the test shape landmark uncertainty in the regression improved results.

Detection and quantification of coronary atherosclerotic plaque by 64-slice multidetector CT: a systematic head-to-head comparison with intravascular ultrasound.

OBJECTIVE: We evaluated the ability of 64-slice multidetector computed tomography (MDCT)-derived plaque parameters to detect and quantify coronary atherosclerosis, using intravascular ultrasound (IVUS) as the reference standard. METHODS: In 32 patients, IVUS and 64-MDCT was performed. The MDCT and IVUS datasets of 44 coronary arteries were co-registered using a newly developed fusion technique and quantitative parameters were derived from both imaging modalities. The threshold of >0.5 mm of maximum wall thickness was used to establish plaque presence on MDCT and IVUS. RESULTS: We analyzed 1364 coregistered 1-mm coronary cross-sections and 255 segments of 5-mm length. Compared with IVUS, 64-MDCT enabled correct detection in 957 of 1109 cross-sections containing plaque (sensitivity 86%). In 180 of 255 cross-sections atherosclerosis was correctly excluded (specificity 71%). On the segmental level, MDCT detected 213 of 220 segments with any atherosclerotic plaque (sensitivity 96%), whereas the presence of any plaque was correctly ruled out in 28 of 32 segments (specificity 88%). Interobserver agreement for the detection of atherosclerotic cross-sections was moderate (Cohen's kappa coefficient K=0.51), but excellent for the atherosclerotic segments (K=1.0). Pearson's correlation coefficient for vessel plaque volumes measured by MDCT and IVUS was r=0.91 (p<0.001). Bland-Altman analysis showed a slight non-significant underestimation of any plaque volume by MDCT (p=0.5), with a trend to underestimate noncalcified and overestimate mixed/calcified plaque volumes (p=0.22 and p=0.87 respectively). CONCLUSION: MDCT is able to detect and quantify atherosclerotic plaque. Further improvement in CT resolution is necessary for more reliable assessment of very small and distal coronary plaques.

Robust shape regression for supervised vessel segmentation and its application to coronary segmentation in CTA.

This paper presents a vessel segmentation method which learns the geometry and appearance of vessels in medical images from annotated data and uses this knowledge to segment vessels in unseen images. Vessels are segmented in a coarse-to-fine fashion. First, the vessel boundaries are estimated with multivariate linear regression using image intensities sampled in a region of interest around an initialization curve. Subsequently, the position of the vessel boundary is refined with a robust nonlinear regression technique using intensity profiles sampled across the boundary of the rough segmentation and using information about plausible cross-sectional vessel shapes. The method was evaluated by quantitatively comparing segmentation results to manual annotations of 229 coronary arteries. On average the difference between the automatically obtained segmentations and manual contours was smaller than the inter-observer variability, which is an indicator that the method outperforms manual annotation. The method was also evaluated by using it for centerline refinement on 24 publicly available datasets of the Rotterdam Coronary Artery Evaluation Framework. Centerlines are extracted with an existing method and refined with the proposed method. This combination is currently ranked second out of 10 evaluated interactive centerline extraction methods. An additional qualitative expert evaluation in which 250 automatic segmentations were compared to manual segmentations showed that the automatically obtained contours were rated on average better than manual contours.