Polymeric endovascular strut and lumen detection algorithm for intracoronary optical coherence tomography images

An All-in-One Tool for 2D Atherosclerotic Disease Assessment and 3D Coronary Artery Reconstruction

Diagnosis of coronary artery disease is mainly based on invasive imaging modalities such as X-ray angiography, intravascular ultrasound (IVUS) and optical coherence tomography (OCT). Computed tomography coronary angiography (CTCA) is also used as a non-invasive imaging alternative. In this work, we present a novel and unique tool for 3D coronary artery reconstruction and plaque characterization using the abovementioned imaging modalities or their combination. In particular, image processing and deep learning algorithms were employed and validated for the lumen and adventitia borders and plaque characterization at the IVUS and OCT frames. Strut detection is also achieved from the OCT images. Quantitative analysis of the X-ray angiography enables the 3D reconstruction of the lumen geometry and arterial centerline extraction. The fusion of the generated centerline with the results of the OCT or IVUS analysis enables hybrid coronary artery 3D reconstruction, including the plaques and the stent geometry. CTCA image processing using a 3D level set approach allows the reconstruction of the coronary arterial tree, the calcified and non-calcified plaques as well as the detection of the stent location. The modules of the tool were evaluated for efficiency with over 90% agreement of the 3D models with the manual annotations, while a usability assessment using external evaluators demonstrated high usability resulting in a mean System Usability Scale (SUS) score equal to 0.89, classifying the tool as “excellent”.

A State-Of-The-Art Review on Coronary Artery Border Segmentation Algorithms for Intravascular Ultrasound (IVUS) Images

Intravascular Ultrasound images (IVUS) is a useful guide for medical practitioners to identify the vascular status of coronary arteries in human beings. IVUS is a unique intracoronary imaging modality that is used as an adjunct to angioplasty to view vessel structures using a catheter with high resolutions. Segmentation of IVUS images has always remained a challenging task due to various impediments, for example, similar tissue components, vessel structures, and artifacts imposed during the acquisition process. Many researchers have applied various techniques to develop standard methods of image interpretation, however, the ultimate goal is still elusive to most researchers. This challenge was presented at the MICCAI- Computing and Visualization for (Intra)Vascular Imaging (CVII) workshop in 2011. This paper presents a major review of recently reported work in the field, with a detailed analysis of various segmentation techniques applied in IVUS, and highlights the directions for future research. The findings recommend a reference database with a larger number of samples acquired at varied transducer frequencies with special consideration towards complex lesions, suitable validation metrics, and ground-truth definition as a standard against which to compare new and current algorithms.

Automated Coronary Optical Coherence Tomography Feature Extraction with Application to Three-Dimensional Reconstruction

Coronary optical coherence tomography (OCT) is an intravascular, near-infrared light-based imaging modality capable of reaching axial resolutions of 10-20 µm. This resolution allows for accurate determination of high-risk plaque features, such as thin cap fibroatheroma; however, visualization of morphological features alone still provides unreliable positive predictive capability for plaque progression or future major adverse cardiovascular events (MACE). Biomechanical simulation could assist in this prediction, but this requires extracting morphological features from intravascular imaging to construct accurate three-dimensional (3D) simulations of patients' arteries. Extracting these features is a laborious process, often carried out manually by trained experts. To address this challenge, numerous techniques have emerged to automate these processes while simultaneously overcoming difficulties associated with OCT imaging, such as its limited penetration depth. This systematic review summarizes advances in automated segmentation techniques from the past five years (2016-2021) with a focus on their application to the 3D reconstruction of vessels and their subsequent simulation. We discuss four categories based on the feature being processed, namely: coronary lumen; artery layers; plaque characteristics and subtypes; and stents. Areas for future innovation are also discussed as well as their potential for future translation.

Automatic lumen segmentation using uniqueness of vascular connected region for intravascular optical coherence tomography

We present an automatic lumen segmentation method using uniqueness of connected region for intravascular optical coherence tomography (IVOCT), which can effectively remove the effect on lumen segmentation caused by blood artifacts. Utilizing the uniqueness of vascular wall on A-lines, we detect the A-lines shared by multiple connected regions, identify connected regions generated by blood artifacts using traversal comparison of connected regions' location, shared ratio and area ratio and then remove all artifacts. We compare these three methods by 216 challenging images with severe blood artifacts selected from clinical 1076 IVOCT images. The metrics of the proposed method are evaluated including Dice index, Jaccard index and accuracy of 94.57%, 90.12%, 98.02%. Compared with automatic lumen segmentation based on the previous morphological feature method and widely used dynamic programming method, the metrics of the proposed method are significantly enhanced, especially in challenging images with severe blood artifacts.

E-wave asymmetry elucidates diastolic ventricular stiffness-relaxation coupling: model-based prediction with in vivo validation

Load, chamber stiffness, and relaxation are the three established determinants of global diastolic function (DF). Coupling of systolic stiffness and isovolumic relaxation has been hypothesized; however, diastolic stiffness-relaxation coupling (DSRC) remains unknown. The parametrized diastolic filling (PDF) formalism, a validated DF model incorporates DSRC. PDF model-predicted DSRC was validated by analysis of 159 Doppler E-waves from a published data set (22 healthy volunteers undergoing bicycle exercise). E-waves at varying (46-120 bpm) heart rates (HR) demonstrated variation in acceleration time (AT), deceleration time (DT), and E-wave peak velocity. AT, DT, and E_peak were converted into PDF parameters: stiffness ([Formula: see text]), relaxation ([Formula: see text]), and load (x_o) using published numerical methods. Univariate linear regression showed that over a twofold increase in HR, AT, and DT decrease ([Formula: see text] = -0.44; P < 0.001 and r = -0.42; P < 0.001, respectively), while, DT/AT remains constant (r = -0.04; P = 0.67). Similarly, [Formula: see text] increases with HR (r = 0.55; P < 0.001), while [Formula: see text] has no significant correlation with HR (r = 0.08; P = 0.32). However, the dimensionless DSRC parameter ψ = c²/4k shows no significant correlation with HR (r = -0.03; P = 0.7). Furthermore, ψ is uniquely determined by DT/AT rather than AT or DT independently. Constancy of ψ in spite of a twofold increase in HR establishes that stiffness (k) and relaxation (c) are coupled and manifest via a HR-invariant parameter of E-wave asymmetry and should not be considered independent of each other. The manifestation of DSRC through E-wave asymmetry via ψ underscores the value of DT/AT as a physiological, mechanism-derived index of DF.NEW & NOTEWORTHY: Although diastolic stiffness and relaxation are considered independent chamber properties, the cardio-hemic inertial oscillation that generates E-waves obeys Newton's law. E-waves vary with heart rate requiring simultaneous change in stiffness and relaxation. By retrospective analysis of human heart-rate varying transmitral Doppler-data, we show that diastolic stiffness and relaxation are coupled and that the coupling manifests through E-wave asymmetry, quantified through a parametrized diastolic filling model-derived dimensionless parameter, which only depends on deceleration time and acceleration time, readily obtainable via standard echocardiography.

Privileged Modality Distillation for Vessel Border Detection in Intracoronary Imaging

Intracoronary imaging is a crucial imaging technology in coronary disease diagnosis as it visualizes the internal tissue morphologies of coronary arteries. Vessel border detection in intracoronary images (VBDI) is desired because it can help the succeeding procedures of computer-aided disease diagnosis. However, existing VDBI methods suffer from the challenge of vessel-environment variability (i.e. high intra- and inter-subject diversity of vessels and their surrounding tissues appeared in images). This challenge leads to the ineffectiveness in the vessel region representation for hand-crafted features, in the receptive field extraction for deeply-represented features, as well as performance suppression derived from clinical data limitation. To solve this challenge, we propose a novel privileged modality distillation (PMD) framework for VBDI. PMD transforms the single-input-single-task (SIST) learning problem in the single-mode VBDI to a multiple-input-multiple-task (MIMT) problem by using the privileged image modality to help the learning model in the target modality. This learns the enriched high-level knowledge with similar semantics and generalizes PMD on diversity-increased low-level image features for improving the model adaptation to diverse vessel environments. Moreover, PMD refines MIMT to SIST by distilling the learned knowledge from multiple to one modality. This eliminates the reliance on privileged modality in the test phase, and thus enables the applicability to each of different intracoronary modalities. A structure-deformable neural network is proposed as an elaborately-designed implementation of PMD. It expands a conventional SIST network structure to the MIMT structure, and then recovers it to the final SIST structure. The PMD is validated on intravascular ultrasound imaging and optical coherence tomography imaging. One modality is the target, and the other one can be considered as the privileged modality owing to their semantic relatedness. The experiments show that our PMD is effective in VBDI (e.g. the Dice index is larger than 0.95), as well as superior to six state-of-the-art VBDI methods.

A novel automated lumen segmentation and classification algorithm for detection of irregular protrusion after stents deployment

BACKGROUND

Clinically, irregular protrusions and blockages after stent deployment can lead to significant adverse outcomes such as thrombotic reocclusion or restenosis. In this study, we propose a novel fully automated method for irregular lumen segmentation and normal/abnormal lumen classification.

METHODS

The proposed method consists of a lumen segmentation, feature extraction, and lumen classification. In total, 92 features were extracted to classify normal/abnormal lumen. The lumen classification method is a combination of supervised learning algorithm and feature selection that is a partition-membership filter method.

RESULTS

As the results, our proposed lumen segmentation method obtained the average of dice similarity coefficient (DSC) and the accuracy of proposed features and the random forest (RF) for normal/abnormal lumen classification as 97.6% and 98.2%, respectively.

CONCLUSIONS

Therefore, we can lead to better understanding of the overall vascular status and help to determine cardiovascular diagnosis.

Automated stent coverage analysis in intravascular OCT (IVOCT) image volumes using a support vector machine and mesh growing

Absence of vascular-stent tissue coverage by IVOCT is a biomarker for potential stent-related thrombosis. We developed highly-automated algorithms to classify covered and uncovered struts and quantitatively evaluate stent apposition. We trained a machine learning model on 7,125 images, and included an active learning, relabeling step to improve noisy labels. We obtained uncovered strut classification sensitivity/specificity (94%/90%) comparable to analyst inter-and-intra-observer variability and AUC (0.97), and tissue coverage thickness measurement arguably better than the commercial product. By comparing classification models from regular and relabeled data sets, we observed robustness of the support vector machine to noisy data. A graph-based algorithm detected clusters of uncovered struts thought to pose a greater risk than isolated uncovered struts. The software enables highly-automated, objective, repeatable, comprehensive stent analysis.

A Mechanical Approach for Smooth Surface Fitting to Delineate Vessel Walls in Optical Coherence Tomography Images

Automated analysis of vascular imaging techniques is limited by the inability to precisely determine arterial borders. Intravascular optical coherence tomography (OCT) offers unprecedented detail of artery wall structure and composition, but does not provide consistent visibility of the outer border of the vessel due to the limited penetration depth. Existing interpolation and surface fitting methods prove insufficient to accurately fill the gaps between the irregularly spaced and sometimes unreliably identified visible segments of the vessel outer border. This paper describes an intuitive, efficient, and flexible new method of 3D surface fitting and smoothing suitable for this task. An anisotropic linear-elastic mesh is fit to irregularly spaced and uncertain data points corresponding to visible segments of vessel borders, enabling the fully automated delineation of the entire inner and outer borders of diseased vessels in OCT images for the first time. In a clinical dataset, the proposed smooth surface fitting approach had great agreement when compared with human annotations: areas differed by just 11 ± 11% (0.93 ± 0.84 mm²), with a coefficient of determination of 0.89. Overlapping and non-overlapping area ratios were 0.91 and 0.18, respectively, with a sensitivity of 90.8 and specificity of 99.0. This spring mesh method of contour fitting significantly outperformed all alternative surface fitting and interpolation approaches tested. The application of this promising proposed method is expected to enhance clinical intervention and translational research using OCT.

Automatic segmentation of optical coherence tomography pullbacks of coronary arteries treated with bioresorbable vascular scaffolds: Application to hemodynamics modeling

Background / Objectives

Automatic algorithms for stent struts segmentation in optical coherence tomography (OCT) images of coronary arteries have been developed over the years, particularly with application on metallic stents. The aim of this study is three-fold: (1) to develop and to validate a segmentation algorithm for the detection of both lumen contours and polymeric bioresorbable scaffold struts from 8-bit OCT images, (2) to develop a method for automatic OCT pullback quality assessment, and (3) to demonstrate the applicability of the segmentation algorithm for the creation of patient-specific stented coronary artery for local hemodynamics analysis.

Methods

The proposed OCT segmentation algorithm comprises four steps: (1) image pre-processing, (2) lumen segmentation, (3) stent struts segmentation, (4) strut-based lumen correction. This segmentation process is then followed by an automatic OCT pullback image quality assessment. This method classifies the OCT pullback image quality as ‘good’ or ‘poor’ based on the number of regions detected by the stent segmentation. The segmentation algorithm was validated against manual segmentation of 1150 images obtained from 23 in vivo OCT pullbacks.

Results

When considering the entire set of OCT pullbacks, lumen segmentation showed results comparable with manual segmentation and with previous studies (sensitivity ~97%, specificity ~99%), while stent segmentation showed poorer results compared to manual segmentation (sensitivity ~63%, precision ~83%). The OCT pullback quality assessment algorithm classified 7 pullbacks as ‘poor’ quality cases. When considering only the ‘good’ classified cases, the performance indexes of the scaffold segmentation were higher (sensitivity >76%, precision >86%).

Conclusions

This study proposes a segmentation algorithm for the detection of lumen contours and stent struts in low quality OCT images of patients treated with polymeric bioresorbable scaffolds. The segmentation results were successfully used for the reconstruction of one coronary artery model that included a bioresorbable scaffold geometry for computational fluid dynamics analysis.

Position Paper Computational Cardiology

Computational cardiology is the scientific field devoted to the development of methodologies that enhance our mechanistic understanding, diagnosis and treatment of cardiovascular disease. In this regard, the field embraces the extraordinary pace of discovery in imaging, computational modeling, and cardiovascular informatics at the intersection of atherogenesis and vascular biology. This paper highlights existing methods, practices, and computational models and proposes new strategies to support a multidisciplinary effort in this space. We focus on the means by that to leverage and coalesce these multiple disciplines to advance translational science and computational cardiology. Analyzing the scientific trends and understanding the current needs we present our perspective for the future of cardiovascular treatment.