Lumen Contours Research Articles

Impact of ultra-high-resolution on cardiac CT assessment of coronary artery stenoses, plaque burden/composition, and pericoronary adipose tissue attenuation

Abstract Background Conventional coronary computed tomography angiography (CCTA) overestimates stenoses within stents and calcified arteries due to "blooming" artifacts. Ultra-high-resolution coronary CT (UHR-CT) may overcome this limitation because of higher spatial resolution. Purpose To investigate differences in coronary arterial lumen volume, stenosis severity, and plaque composition, between CCTA and UHRCT in patients with stents and/or high coronary calcium scores. Methods As part of a prospective pilot study, 14 patients referred for invasive coronary angiography for evaluation of obstructive CAD underwent UHRCT (Canon Precision, 160 x 0.25mm slices) prior to catheterization. Images of identical cardiac phases were reconstructed in two badges for comparison: 1) Ultra-high resolution using a 1,024 x 1,024 matrix, 0.25 mm, and 2) conventional resolution with a 524 x 524 matrix, 0.5 mm. All images were reconstructed using iterative reconstruction algorithm (AIDR 3D) and a sharp kernel (FC05). For the left main and three coronary arteries, an automated contour analysis using a dedicated software (QAngioCT RE 3.2, Medis, Netherlands) was performed to compare conventional CCTA vs. UHR-CT resolution images across common vascular and plaque metrics, including lumen and vessel volume, plaque burden/volume, plaque composition based upon predefined fixed intensity cut-off values of CT attenuation, minimal and maximal plaque thickness, pericoronary adipose tissue attenuation (PCAT), minimal lumen area and diameter, and diameter stenosis. All lumen and vessel contours were generated by the software, without manual editing. The segment model proposed by the SCCT was used as reference and the lesions classified as greater than 25% stenosis by visual assessment were recorded. The quality of the images was subjectively assessed using a 5-point Likert scale. Vessel segments with scores 1 (severe artifacts) and 2 (pronounced artifacts) were not included in the analysis. Results The mean patient age was 68.5 years (±6.8), 2 were female (14.3%), 6 patients had stents (11 stented segments) and the mean Agatston score was 947.6 (±449) among the 8 non-stented patients. A total of 56 vessels, 133 segments, and 57 lesions were analyzed. Eight segments (6%) were excluded from the analysis due to poor quality. Results on a per segment and per lesion basis were summarized in the table. The mean PCAT value was significantly lower for the UHRCT images compared to NR reconstruction in both segment (p<0.001) and lesion based (p=0.002) analyses. Plaque differentiation varied by UHRCT resulted in a lower fibrous plaque volumes but higher necrotic (low attenuation) plaque volumes. Conclusions Ultra-high-resolution CT image analysis results in significantly different coronary artery plaque volume and PCAT assessment vs. conventional resolution CCTA. These findings have implications for evaluating high-risk plaque features and perivascular inflammation in patients.Table 1

Data-driven models for the prediction of coronary atherosclerotic plaque progression/regression

Coronary artery disease is defined by the existence of atherosclerotic plaque on the arterial wall, which can cause blood flow impairment, or plaque rupture, and ultimately lead to myocardial ischemia. Intravascular ultrasound (IVUS) imaging can provide a detailed characterization of lumen and vessel features, and so plaque burden, in coronary vessels. Prediction of the regions in a vascular segment where plaque burden can either increase (progression) or decrease (regression) following a certain therapy, has remained an elusive major milestone in cardiology. Studies like IBIS-4 showed an association between plaque burden regression and high-intensity rosuvastatin therapy over 13 months. Nevertheless, it has not been possible to predict if a patient would respond in a favorable/adverse fashion to such a treatment. This work aims to (i) Develop a framework that processes lumen and vessel cross-sectional contours and extracts geometric descriptors from baseline and follow-up IVUS pullbacks; and to (ii) Develop, train, and validate a machine learning model based on baseline/follow-up IVUS datasets that predicts future percent of atheroma volume changes in coronary vascular segments using only baseline information, i.e. geometric features and clinical data. This is a post hoc analysis, revisiting the IBIS-4 study. We employed 140 arteries, from 81 patients, for which expert delineation of lumen and vessel contours were available at baseline and 13-month follow-up. Contour data from baseline and follow-up pullbacks were co-registered and then processed to extract several frame-wise features, e.g. areas, plaque burden, eccentricity, etc. Each pullback was divided into regions of interest (ROIs), following different criteria. Frame-wise features were condensed into region-wise markers using tools from statistics, signal processing, and information theory. Finally, a stratified 5-fold cross-validation strategy (20 repetitions) was used to train/validate an XGBoost regression models. A feature selection method before the model training was also applied. When the models were trained/validated on ROI defined by the difference between follow-up and baseline plaque burden, the average accuracy and Mathews correlation coefficient were 0.70 and 0.41 respectively. Using a ROI partition criterion based only on the baseline’s plaque burden resulted in averages of 0.60 accuracy and 0.23 Mathews correlation coefficient. An XGBoost model was capable of predicting plaque progression/regression changes in coronary vascular segments of patients treated with rosuvastatin therapy in 13 months. The proposed method, first of its kind, successfully managed to address the problem of stratification of patients at risk of coronary plaque progression, using IVUS images and standard patient clinical data.

Organized thrombus is a frequent underlying feature in culprit lesion morphology in non-ST-elevation myocardial infarction. A study using optical coherence tomography and magnetic resonance imaging

The concept that the culprit lesion in non-ST segment elevation myocardial infarction (NSTEMI) is caused by sudden plaque rupture with acute thrombus formation has recently been challenged. While angiography is an old gold-standard for culprit identification it merely visualizes the lumen contour. Optical coherence tomography (OCT) provides a detailed view of culprit features. Combined with myocardial edema on cardiac magnetic resonance (CMR), indicating acute ischemia and thus culprit location, we aimed to characterize culprit lesions using OCT. Patients with NSTEMI referred for angiography were prospectively enrolled. OCT was performed on angiographic stenoses ≥50% and on operator-suspected culprit lesions. Hierarchical OCT-culprit identifiers were defined in case of multiple unstable lesions, including OCT-defined thrombus age. An OCT-based definition of an organizing thrombus as corresponding to histological early healing stage was introduced. Lesions were classified as OCT-culprit or non-culprit, and characteristics compared. CMR was performed in a subset of patients. We included 65 patients with 97 lesions, of which 49 patients (75%) had 53 (54%) OCT-culprit lesions. The most common OCT-culprit identifiers were the presence of acute (66%) and organizing thrombus (19%). Plaque rupture was visible in 45% of OCT-culprit lesions. CMR performed in 38 patients revealed myocardial oedema in the corresponding territories of 67% of acute thrombi and 50% of organizing thrombi. A culprit lesion was identified by OCT in 75% patients with NSTEMI. Acute thrombus was the most frequent feature followed by organizing thrombus. Applying specific OCT-criteria to identify the culprit could prove valuable in ambiguous cases.

Biomechanical characterization of tissue types in murine dissecting aneurysms based on histology and 4D ultrasound-derived strain

Abdominal aortic aneurysm disease is the local enlargement of the aorta, typically in the infrarenal section, causing up to 200,000 deaths/year. In vivo information to characterize the individual elastic properties of the aneurysm wall in terms of rupture risk is lacking. We used a method that combines 4D ultrasound and direct deformation estimation to compute in vivo 3D Green-Lagrange strain in murine angiotensin II-induced dissecting aortic aneurysms, a commonly used mouse model. After euthanasia, histological staining of cross-sectional sections along the aorta was performed in areas where in vivo strains had previously been measured. The histological sections were segmented into intact and fragmented elastin, thrombus with and without red blood cells, and outer vessel wall including the adventitia. Meshes were then created from the individual contours based on the histological segmentations. The isolated contours of the outer wall and lumen from both imaging modalities were registered individually using a coherent point drift algorithm. 2D finite element models were generated from the meshes, and the displacements from the registration were used as displacement boundaries of the lumen and wall contours. Based on the resulting deformed contours, the strains recorded were grouped according to segmented tissue regions. Strains were highest in areas containing intact elastin without thrombus attachment. Strains in areas with intact elastin and thrombus attachment, as well as areas with disrupted elastin, were significantly lower. Strains in thrombus regions with red blood cells were significantly higher compared to thrombus regions without. We then compared this analysis to statistical distribution indices and found that the results of each aligned, elucidating the relationship between vessel strain and structural changes. This work demonstrates the possibility of advancing in vivo assessments to a microstructural level ultimately improving patient outcomes.

Precisión diagnóstica de la TC-RFF con un nuevo algoritmo de subpíxel grueso a fino en la detección de isquemia específica de lesión: un estudio multicéntrico prospectivo

Introduction and objectivesA new computed tomography-derived fractional flow reserve (CT-FFR) technique with a “coarse-to-fine subpixel” algorithm has been developed to generate precise lumen contours. The aim of this study was to assess the diagnostic performance of this new CT-FFR algorithm for discriminating lesion-specific ischemia using wire-based FFR ≤ 0.80 as the reference standard in patients with coronary artery disease. MethodsThis prospective, multicenter study screened 330 patients undergoing coronary CT angiography (CCTA) and invasive FFR (median interval 2 days) from 6 tertiary hospitals. CT-FFR was evaluated in a blinded fashion with a “coarse-to-fine subpixel” algorithm for lumen contour. ResultsBetween March 2019 and May 2020, we included 316 patients with 324 vessels. There was a good correlation between CT-FFR and invasive FFR (r=0.76, P<.001). The diagnostic sensitivity, specificity, and accuracy on a per-vessel level were 95.3%, 89.8%, and 92.0% for CT-FFR, and 96.4%, 26.4%, and 53.1% for CCTA>50% stenosis, respectively. CT-FFR showed improved discrimination of ischemia compared with CCTA alone overall (AUC, 0.95 vs 0.74, P<.001) and in intermediate (AUC, 0.96 vs 0.62, P<.001) and “gray zone” lesions (AUC, 0.88 vs 0.61, P<.001). The diagnostic specificity, accuracy, and AUC for CT-FFR (71.9%, 82.8%, and 0.84) outperformed CCTA (9.4%, 48.3%, and 0.66) in patients or in vessels with severe calcification (all P<.05). ConclusionsCT-FFR with a new “coarse-to-fine subpixel” algorithm showed high performance in identifying hemodynamically significant stenosis. The diagnostic performance of CT-FFR was superior to that of CCTA in intermediate lesions, “gray zone” lesions, and severely calcified lesions.Clinical Trial Register: NCT04731285Full English text available from:www.revespcardiol.org/en

Fast Computational Approaches to Derive Fractional Pressure Ratio in Patients with Extracranial or Intracranial Symptomatic Stenosis

We aimed to evaluate the performance of fast and straightforward μQFR computation in cerebrovascular stenosis lesions. A total of 30 patients with symptomatic stenosis lesions of 50-70% luminal stenosis and underwent FPR assessment at our hospital were included in this study. μQFR was applied to the interrogated vessel. An artificial intelligence algorithm was proposed for automatic delineation of lumen contours of cerebrovascular stenosis. We used invasive fractional pressure ratios (FPRs) as a reference standard. Pearson's correlation coefficient (r) was used to assess the correlation strength between μQFR and FPR, while Bland-Altman plots were used to evaluate the agreement between μQFR and FPR. An analysis of receiver-operating characteristics (ROC) was used to evaluate the performance of μQFR. Our result displayed strong positive correlations (r = 0.92; p<0.001) between μQFR and pressure wire FPR. An excellent agreement was observed between μQFR and FPR with a mean difference of 0.01 + 0.08; p=0.263; range from -0.16 to 0.14. The overall accuracy for identifying an FPR of <0.7 was 92% (95% CI: 85% to 100%). The area under the receiver-operating characteristic curve was higher for μQFR (0.92 [95% CI: 0.81 to 0.98]) than diameter stenosis (DS) (0.88 [95% CI: 0.75 to 0.95]. The positive likelihood ratio was 3.9 for μQFR with a negative likelihood ratio of 0, respectively. The μQFR computation has a strong correlation and agrees with the FPR calculated from the pressure wire. Therefore, μQFR may provide an essential therapeutic aid in patients with symptomatic stenosis lesions.

Automatic 3D reconstruction of an anatomically correct upper airway from endoscopic long range OCT images.

Endoscopic airway optical coherence tomography (OCT) is a non-invasive and high resolution imaging modality for the diagnosis and analysis of airway-related diseases. During OCT imaging of the upper airway, in order to reliably characterize its 3D structure, there is a need to automatically detect the airway lumen contour, correct rotational distortion and perform 3D airway reconstruction. Based on a long-range endoscopic OCT imaging system equipped with a magnetic tracker, we present a fully automatic framework to reconstruct the 3D upper airway model with correct bending anatomy. Our method includes an automatic segmentation method for the upper airway based on dynamic programming algorithm, an automatic initial rotation angle error correction method for the detected 2D airway lumen contour, and an anatomic bending method combined with the centerline detected from the magnetically tracked imaging probe. The proposed automatic reconstruction framework is validated on experimental datasets acquired from two healthy adults. The result shows that the proposed framework allows the full automation of 3D airway reconstruction from OCT images and thus reveals its potential to improve analysis efficiency of endoscopic OCT images.

Diagnostic accuracy of CT-FFR with a new coarse-to-fine subpixel algorithm in detecting lesion-specific ischemia: a prospective multicenter study

Introduction and objectivesA new computed tomography-derived fractional flow reserve (CT-FFR) technique with a “coarse-to-fine subpixel” algorithm has been developed to generate precise lumen contours. The aim of this study was to assess the diagnostic performance of this new CT-FFR algorithm for discriminating lesion-specific ischemia using wire-based FFR ≤ 0.80 as the reference standard in patients with coronary artery disease. MethodsThis prospective, multicenter study screened 330 patients undergoing coronary CT angiography (CCTA) and invasive FFR (median interval 2 days) from 6 tertiary hospitals. CT-FFR was evaluated in a blinded fashion with a “coarse-to-fine subpixel” algorithm for lumen contour. ResultsBetween March 2019 and May 2020, we included 316 patients with 324 vessels. There was a good correlation between CT-FFR and invasive FFR (r=0.76, P<.001). The diagnostic sensitivity, specificity, and accuracy on a per-vessel level were 95.3%, 89.8%, and 92.0% for CT-FFR, and 96.4%, 26.4%, and 53.1% for CCTA>50% stenosis, respectively. CT-FFR showed improved discrimination of ischemia compared with CCTA alone overall (AUC, 0.95 vs 0.74, P<.001) and in intermediate (AUC, 0.96 vs 0.62, P<.001) and “gray zone” lesions (AUC, 0.88 vs 0.61, P<.001). The diagnostic specificity, accuracy, and AUC for CT-FFR (71.9%, 82.8%, and 0.84) outperformed CCTA (9.4%, 48.3%, and 0.66) in patients or in vessels with severe calcification (all P<.05). ConclusionsCT-FFR with a new “coarse-to-fine subpixel” algorithm showed high performance in identifying hemodynamically significant stenosis. The diagnostic performance of CT-FFR was superior to that of CCTA in intermediate lesions, “gray zone” lesions, and severely calcified lesions.Clinical Trial Register: NCT04731285

Extrapolation of CT aortic annulus derived diameter/CT aortic annulus area from 2D echocardiography aortic annulus diameter in TAVR: not a reliable tool

Abstract Funding Acknowledgements Type of funding sources: None. Background Computed tomography (CT) aortic annulus area, CT area derived diameter and two-dimensional echocardiography (2D ECHO) diameters are used for valvular sizing in transcutaneous aortic valve replacement (TAVR). According to our previous study, 2D ECHO annulus diameter was smaller in comparison to CT derived diameter (22.6±2.9 vs 25.0±5.5mm, p = 0.013), but both CT aortic annulus area derived diameter/CT aortic annulus and 2D ECHO diameter predicted correct valve size selection with the same accuracy of 79–80%. However, CT requires radiation and iodine contrast administration, which might be the limitation in patients with chronic kidney disease or history of iodine allergy. Purpose To determine if 2D ECHO diameter can be used for reliable estimation of CT area derived diameter/CT aortic annulus area. Methods 139 consecutive patients with favourable outcome of Sapien XT/Sapien S3 TAVR (no more than mild aortic regurgitation, only one valve implanted) were included in the study. 2D ECHO diameter was measured from mid oesophageal three-chamber view (110–135°) just below the anatomical aortic valve annulus. CT aortic annulus was obtained as double oblique plane by manual multiplanar reconstruction of CT dataset, defined by three lowest attachment points of valvular leaflets. Aortic annulus area was measured by manual tracking of lumen contours, area derived diameter as 2*√(area/π). Linear regression analysis between pre-implant CT area derived diameter and 2D ECHO diameter was performed; extrapolated CT annulus area diameter and CT aortic annulus area [π*(d/2)2] were calculated from 2D ECHO diameter. Concordance of valve size selection based on extrapolated CT area derived diameter vs original CT area derived diameter and extrapolated CT aortic annulus area vs original CT aortic annulus area was determined on patients where CT aortic annulus area/annulus derived diameter were falling outside the overlap of the manufacturer recommended ranges for neighbouring valve sizes. Results Significant relationship between CT aera derived diameter and 2D ECHO diameter was found (r=0.720, p&amp;lt;0.001, Fig 1,2). Extrapolated CT area derived diameter would have led to inappropriate size selection in 33/100 more cases (undersizing in 16%, oversizing in 17%), similar to extrapolated aortic annulus area, where both undersizing and oversizing was observed in 18/106 (17%) more cases in comparison to original CT based parameters. Conclusion Extrapolation of CT aortic annulus area/area derived diameter from 2D ECHO diameter is not a reliable tool in valvular sizing for balloon expandable Sapien valve. In patients with chronic kidney disease or history of iodine allergy, three-dimensional (3D) ECHO aortic annulus assessment remains the only viable alternative tool.

Abstract 14241: Feasibility and Safety of Non-Contrast Optical Coherence Tomography Imaging Using Hydroxyethyl Starch in Coronary Arteries

Background/Objectives: Contemporary high-resolution intracoronary optical coherence tomography (OCT) requires injection of flushing media for image acquisition. There is a growing need for an alternative flushing media other than iodine contrast to reduce the risk of acute kidney injury. To address this issue, we investigated the safety and feasibility of hydroxyethyl starch (pentastarch) as a flushing media for OCT image acquisition and compared with the contrast media flushed images. Methods: We prospectively enrolled 43 patients with 70 coronary lesions (46 stented and 24 pre-stented or non-culprit lesions). Paired OCT images were obtained by manual injection of iodine contrast (average 12 mL), followed by pentastarch (average 15 mL). A total of 162 runs (81 pairs) were analyzed and each OCT pullback was assessed on a frame-by-frame basis by automated customized lumen contour and stent strut segmentation algorithm. Results: Image quality as assessed by proportion of clear image segments were comparable between the contrast media and pentastarch flushed images (93.3% vs. 93.7%; p=0.501). Pixel-based flushing capability showed no significant difference between the paired image segments (median(IQR), 0.951(0.947-0.953) vs. 0.951(0.949-0.952), p=0.264; Figure). Quantitative measurements of minimal stent area (R 2 =0.977, p&lt;0.001), minimal lumen area (R 2 =0.983, p&lt;0.001) correlated well between the paired OCT runs. There were no clinically relevant complications associated with OCT imaging with pentastarch such as anaphylaxis, life-threatening arrhythmia, or acute renal deterioration. Conclusions: Intracoronary OCT using pentastarch is safe and technically feasible without affecting the optical signal intensity and overall image quality. Non-contrast imaging with pentastarch significantly reduces the obligatory iodine contrast load, therefore, could be considered as a promising strategy in high-risk patients for renal impairment.

A TWO-STAGE EXTRACTION NETWORK FOR THE OPTICAL COHERENCE TOMOGRAPHY CONTOUR

Optical coherence tomography (OCT) with high resolution was found lucrative for coronary heart disease diagnosis and treatment such as evaluating the degree of vascular stenosis and guiding stent implantation. Lumen contour is an essential information for assessing the shape and stenosis of blood vessels. The automatic and accurate extraction of lumen contour can significantly reduce the burden of doctors and improve the efficiency of diagnosis and treatment of coronary heart disease. The existing contour segmentation methods have poor performance in extracting the contour of the image with a missing blood vessel wall. Therefore, this paper proposes a two-stage contour extraction method based on deep learning. First, the method uses the basic contour extraction network to complete the feature image extraction and coarse contour segmentation, then inputs the feature image into the attentional class feature (ACF) module, and uses the coarse segmentation results to guide the generation of the new feature map for the refined contour segmentation. Finally, the fully connected conditional random field (CRF) module is used for post-processing yielding a more refined and smooth segmentation boundary. The experiments show that compared with the other methods, the lumen contour segmentation method proposed in this paper can adequately resolve the problem of partial vessel wall absence in OCT images and improve the accuracy of lumen contour extraction.

PREFACE: A SPECIAL SELECTION ON EMERGING TECHNIQUES FOR BIOMECHANICS–PART II

PREFACE: A SPECIAL SELECTION ON EMERGING TECHNIQUES FOR BIOMECHANICS–PART II

Human vs. Machine vs. Core Lab in Lumen and Vessel Contour Segmentation With Intravascular Ultrasound

Human vs. Machine vs. Core Lab in Lumen and Vessel Contour Segmentation With Intravascular Ultrasound

Deep Learning-Based Automated Detection of Arterial Vessel Wall and Plaque on Magnetic Resonance Vessel Wall Images.

PurposeTo develop and evaluate an automatic segmentation method of arterial vessel walls and plaques, which is beneficial for facilitating the arterial morphological quantification in magnetic resonance vessel wall imaging (MRVWI).MethodsMRVWI images acquired from 124 patients with atherosclerotic plaques were included. A convolutional neural network-based deep learning model, namely VWISegNet, was used to extract the features from MRVWI images and calculate the category of each pixel to facilitate the segmentation of vessel wall. Two-dimensional (2D) cross-sectional slices reconstructed from all plaques and 7 main arterial segments of 115 patients were used to build and optimize the deep learning model. The model performance was evaluated on the remaining nine-patient test set using the Dice similarity coefficient (DSC) and average surface distance (ASD).ResultsThe proposed automatic segmentation method demonstrated satisfactory agreement with the manual method, with DSCs of 93.8% for lumen contours and 86.0% for outer wall contours, which were higher than those obtained from the traditional U-Net, Attention U-Net, and Inception U-Net on the same nine-subject test set. And all the ASD values were less than 0.198 mm. The Bland–Altman plots and scatter plots also showed that there was a good agreement between the methods. All intraclass correlation coefficient values between the automatic method and manual method were greater than 0.780, and greater than that between two manual reads.ConclusionThe proposed deep learning-based automatic segmentation method achieved good consistency with the manual methods in the segmentation of arterial vessel wall and plaque and is even more accurate than manual results, hence improved the convenience of arterial morphological quantification.

Human vs. machine vs. core lab for the assessment of coronary atherosclerosis with lumen and vessel contour segmentation with intravascular ultrasound.

A machine learning (ML) algorithm for automatic segmentation of intravascular ultrasound was previously validated. It has the potential to improve efficiency, accuracy and precision of coronary vessel segmentation compared to manual segmentation by interventional cardiology experts. The aim of this study is to compare the performance of human readers to the machine and against the readings from a Core Laboratory. This is a post-hoc, cross-sectional analysis of the IBIS-4 study. Forty frames were randomly selected and analyzed by 10 readers of varying expertise two separate times, 1week apart. Their measurements of lumen, vessel, plaque areas, and plaque burden were performed in an offline software. Among humans, the intra-observer variability was not statistically significant. For the total 80 frames, inter-observer variability between human readers, the ML algorithm and Core Laboratory for lumen area, vessel area, plaque area and plaque burden were not statistically different. For lumen area, however, relative differences between the human readers and the Core Lab ranged from 0.26 to 12.61%. For vessel area, they ranged from 1.25 to 9.54%. Efficiency between the ML algorithm and the readers differed notably. Humans spent 47min on average to complete the analyses, while the ML algorithm took on average less than 1min. The overall lumen, vessel and plaque means analyzed by humans and the proposed ML algorithm are similar to those of the Core Lab. Machines, however, are more time efficient. It is warranted to consider use of the ML algorithm in clinical practice.

Quantitative assessment of microstructural evolution of intracranial aneurysm wall by vessel wall imaging.

This study aimed to evaluate new quantitative parameters of aneurysm wall enhancement (AWE) on magnetic resonance vessel wall imaging (VWI) in differentiating between the stable and evolving unruptured intracranial aneurysms (UIAs). Thirty-eight consecutive patients with UIAs (27 stable and 11 evolving) underwent VWI with contrast-enhanced 3D T1 volume isotropic turbo spin echo acquisition. The voxel-based enhancement maps were created using pre- and post-contrast images. The aneurysmal lumen with signal suppression by black-blood method was segmented. Then, one voxel outer and inner layers of the lumen contour were automatically segmented. The shape features of the aneurysms and AWE of the two layers were compared between stable and evolving groups. The shape features, including aneurysm volume, surface, and compacity were significantly different between the stable and evolving groups (P = 0.024, 0.028, and 0.033, respectively). Stable and evolving groups also differed significantly in the AWE at the union of outer and inner layers of the aneurysm wall (P = 0.0082) but not in that of the outer or inner layer alone. Multivariate logistic regression analysis revealed significant differences in aneurysm volume, surface, and AWE at the union of outer and inner layers between the two groups (P = 0.0029, 0.0092, and 0.0033, respectively). Receiver operating characteristics curve analysis revealed that the area under the curve of the logistic regression model was 0.89. Quantitative combined analysis of aneurysm shape features and AWE of the union of outer and inner layers were effective for differentiating between stable and evolving UIAs.

Fully automated lumen and vessel contour segmentation in intravascular ultrasound datasets.

Segmentation of lumen and vessel contours in intravascular ultrasound (IVUS) pullbacks is an arduous and time-consuming task, which demands adequately trained human resources. In the present study, we propose a machine learning approach to automatically extract lumen and vessel boundaries from IVUS datasets. The proposed approach relies on the concatenation of a deep neural network to deliver a preliminary segmentation, followed by a Gaussian process (GP) regressor to construct the final lumen and vessel contours. A multi-frame convolutional neural network (MFCNN) exploits adjacency information present in longitudinally neighboring IVUS frames, while the GP regression method filters high-dimensional noise, delivering a consistent representation of the contours. Overall, 160 IVUS pullbacks (63 patients) from the IBIS-4 study (Integrated Biomarkers and Imaging Study-4, Trial NCT00962416), were used in the present work. The MFCNN algorithm was trained with 100 IVUS pullbacks (8427 manually segmented frames), was validated with 30 IVUS pullbacks (2583 manually segmented frames) and was blindly tested with 30 IVUS pullbacks (2425 manually segmented frames). Image and contour metrics were used to characterize model performance by comparing ground truth (GT) and machine learning (ML) contours. Median values (interquartile range, IQR) of the Jaccard index for lumen and vessel were 0.913, [0.882,0.935] and 0.940, [0.917,0.957], respectively. Median values (IQR) of the Hausdorff distance for lumen and vessel were 0.196mm, [0.146,0.275]mm and 0.163mm, [0.122,0.234]mm, respectively. Also, the mean value of lumen area predictions, and limits of agreement were -0.19mm2, [1.1,-1.5]mm2, while the mean value and limits of agreement of plaque burden were 0.0022, [0.082,-0.078]. The results obtained with the model developed in this work allow us to conclude that the proposed machine learning approach delivers accurate segmentations in terms of image metrics, contour metrics and clinically relevant variables, enabling its use in clinical routine by mitigating the costs involved in the manual management of IVUS datasets.

Convolutional networks for the segmentation of intravascular ultrasound images: Evaluation on a multicenter dataset

Background and objectiveThe delineation of the lumen contour and external elastic lamina (EEL) in intravascular ultrasound (IVUS) images is crucial for the quantitative analysis of coronary atherosclerotic plaques. However, the presence of ultrasonic shadows and anatomical structures (such as bifurcations and calcified plaque) complicates the automatic delineation of the lumen contour and EEL. The purpose of this paper is to evaluate the IVUS segmentation performances of different convolutional networks and the impact factors on a large-scale multiple-center dataset. MethodsA total of 6516 cross-sectional images from 175 IVUS pullbacks acquired in different centers by different IVUS imaging catheters were screened from a corelab to evaluate the segmentation methods. The IVUS images included bifurcation, side branch ostia, and various image artifacts to reflect the general image characteristics in routine clinical acquisition. We compared three generic fully convolutional networks (FCNs) and two FCNs specifically designed for the segmentation of IVUS images and explored the factors impacting the segmentation performance, including the training images and the input of consecutive images to the models. The performance of the FCNs was evaluated by using the Dice similarity coefficient (DSC), the Jaccard index (JI), the Hausdorff distance (HD), linear regression and Bland-Altman analysis. ResultsThe 4-cascaded RefineNet and DeepLabv3+ outperformed U-net and IVUS-net in the segmentation of the lumen contour and EEL on IVUS images. DeepLabv3+ had the best segmentation performance, with DSCs of 0.927 and 0.944, JIs of 0.911 and 0.933, and HDs of 0.336 mm and 0.367 mm for delineation of the lumen and EEL, respectively. Excellent agreement between DeepLabv3+ and the manual delineation was found in the quantification of the coronary plaque area (r = 0.98). ConclusionsThe convolutional network architecture is effective in the automatic segmentation of IVUS images. It might contribute to the clinical application of quantitative IVUS analysis in real-world as well as the efficient assessment of coronary atherosclerosis.

Optical Coherence Tomography-Derived Changes in Plaque Structural Stress Over the Cardiac Cycle: A New Method for Plaque Biomechanical Assessment.

Introduction: Cyclic plaque structural stress has been hypothesized as a mechanism for plaque fatigue and eventually plaque rupture. A novel approach to derive cyclic plaque stress in vivo from optical coherence tomography (OCT) is hereby developed.Materials and Methods: All intermediate lesions from a previous OCT study were enrolled. OCT cross-sections at representative positions within each lesion were selected for plaque stress analysis. Detailed plaque morphology, including plaque composition, lumen and internal elastic lamina contours, were automatically delineated. OCT-derived vessel and plaque morphology were included in a 2-dimensional finite element analysis, loaded with patient-specific intracoronary pressure tracing data, to calculate the changes in plaque structural stress (ΔPSS) on vessel wall over the cardiac cycle.Results: A total of 50 lesions from 41 vessels were analyzed. A significant ΔPSS gradient was observed across the plaque, being maximal at the proximal shoulder (45.7 [32.3, 78.6] kPa), intermediate at minimal lumen area (MLA) (39.0 [30.8, 69.1] kPa) and minimal at the distal shoulder (35.1 [28.2, 72.3] kPa; p = 0.046). The presence of lipidic plaques were observed in 82% of the diseased segments. Larger relative lumen deformation and ΔPSS were observed in diseased segments, compared with normal segments (percent diameter change: 8.2 ± 4.2% vs. 6.3 ± 2.3%, p = 0.04; ΔPSS: 59.3 ± 48.2 kPa vs. 27.5 ± 8.2 kPa, p < 0.001). ΔPSS was positively correlated with plaque burden (r = 0.37, p < 0.001) and negatively correlated with fibrous cap thickness (r = −0.25, p = 0.004).Conclusions: ΔPSS provides a feasible method for assessing plaque biomechanics in vivo from OCT images, consistent with previous biomechanical and clinical studies based on different methodologies. Larger ΔPSS at proximal shoulder and MLA indicates the critical sites for future biomechanical assessment.

Simultaneous assessment of coronary stenosis relevance with automated computed tomography angiography and intravascular ultrasound analyses and fractional flow reserve

Objectives To assess the diagnostic performance of computed tomography angiography (CTA) and intravascular ultrasound (IVUS) derived minimum lumen areas (MLA) from the same lesions that correspond to an FFR ≤0.80. Methods and results A total of 24 patients (33 arteries) were collected retrospectively according to the following inclusion criteria: presence of a CTA diagnostic followed by an IVUS and FFR percutaneous coronary procedures. CTA and IVUS lumen contours were automatically performed using previously validated methods. The correlation between CTA and IVUS for the MLA was r = 0.45. In terms of MLA, the mean difference between CTA and IVUS was 0.81 mm2. Of note, a much smaller CTA-derived MLA (2.10 mm2) was found to be related to significant FFR lesions compared to that of the MLA derived from IVUS (3.19 mm2). The area under the curve, accuracy, sensitivity and specificity for this CTA-derived MLA were 0.80, 0.76, 0.50 and 0.87, respectively, while these values for IVUS-derived MLA were 0.87, 0.85, 0.80 and 0.87. Conclusions Computed tomography angiography and intravascular ultrasound-derived minimum lumen areas have moderate diagnostic efficiency, albeit slightly better for IVUS, in identifying hemodynamically severe coronary stenoses. The utility of MLA, automatically derived from either CTA or IVUS as an alternative to FFR to guide the decision to revascularize, should be tested clinically.