Classification of calcium in intravascular OCT images for the purpose of intervention planning

Artificial Intelligence in Cardiovascular Atherosclerosis Imaging

At present, artificial intelligence (AI) has already been applied in cardiovascular imaging (e.g., image segmentation, automated measurements, and eventually, automated diagnosis) and it has been propelled to the forefront of cardiovascular medical imaging research. In this review, we presented the current status of artificial intelligence applied to image analysis of coronary atherosclerotic plaques, covering multiple areas from plaque component analysis (e.g., identification of plaque properties, identification of vulnerable plaque, detection of myocardial function, and risk prediction) to risk prediction. Additionally, we discuss the current evidence, strengths, limitations, and future directions for AI in cardiac imaging of atherosclerotic plaques, as well as lessons that can be learned from other areas. The continuous development of computer science and technology may further promote the development of this field.

The Prognostic Value of a Validated and Automated Intravascular Ultrasound-Derived Calcium Score

Background

Coronary calcification has been linked to cardiovascular events. We developed and validated an algorithm to automatically quantify coronary calcifications on intravascular ultrasound (IVUS). We aimed to assess the prognostic value of an IVUS-calcium score (ICS) on patient-oriented composite endpoint (POCE).

Methods

We included patients that underwent coronary angiography plus pre-procedural IVUS imaging. The ICS was calculated per patient. The primary endpoint was a composite of all-cause mortality, stroke, myocardial infarction, and revascularization (POCE).

Results

In a cohort of 408 patients, median ICS was 85. Both an ICS ≥ 85 and a 100 unit increase in ICS increased the risk of POCE at 6-year follow-up (adjusted hazard ratio (aHR) 1.51, 95%CI 1.05–2.17, p value = 0.026, and aHR 1.21, 95%CI 1.04–1.41, p value = 0.014, respectively).

Conclusions

The ICS, calculated by a validated automated algorithm derived from routine IVUS pullbacks, was strongly associated with the long-term risk of POCE.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s12265-021-10103-1.

Application and Evaluation of Highly Automated Software for Comprehensive Stent Analysis in Intravascular Optical Coherence Tomography

Intravascular optical coherence tomography (IVOCT) is used to assess stent tissue coverage and malapposition in stent evaluation trials. We developed the OCT Image Visualization and Analysis Toolkit for Stent (OCTivat-Stent), for highly automated analysis of IVOCT pullbacks. Algorithms automatically detected the guidewire, lumen boundary, and stent struts; determined the presence of tissue coverage for each strut; and estimated the stent contour for comparison of stent and lumen area. Strut-level tissue thickness, tissue coverage area, and malapposition area were automatically quantified. The software was used to analyze 292 stent pullbacks. The concordance-correlation-coefficients of automatically measured stent and lumen areas and independent manual measurements were 0.97 and 0.99, respectively. Eleven percent of struts were missed by the software and some artifacts were miscalled as struts giving 1% false-positive strut detection. Eighty-two percent of uncovered struts and 99% of covered struts were labeled correctly, as compared to manual analysis. Using the highly automated software, analysis was harmonized, leading to a reduction of inter-observer variability by 30%. With software assistance, analysis time for a full stent analysis was reduced to less than 30 minutes. Application of this software to stent evaluation trials should enable faster, more reliable analysis with improved statistical power for comparing designs.

Automated classification of dense calcium tissues in gray-scale intravascular ultrasound images using a deep belief network

Background

IVUS is widely used to quantitatively assess coronary artery disease. The purpose of this study was to automatically characterize dense calcium (DC) tissue in the gray scale intravascular ultrasound (IVUS) images using the image textural features.

Methods

A total of 316 Gy-scale IVUS and corresponding virtual histology images from 26 patients with acute coronary syndrome who underwent IVUS along with X-ray angiography between October 2009 to September 2014 were retrospectively acquired and analyzed. One expert performed all procedures and assessed their IVUS scans. After image acquisition, the DC candidate and corresponding acoustic shadow regions were automatically determined. Then, nine image-base feature groups were extracted from the DC candidates. In order to reduce the dimensionalities, principal component analysis (PCA) was performed, and selected feature sets were utilized as an input for a deep belief network. Classification results were validated using 10-fold cross validation.

Results

The dimensionality of the feature map was efficiently reduced by 50% (from 66 to 33) without any performance decrease using PCA method. Sensitivity, specificity, and accuracy of the proposed method were 92.8 ± 0.1%, 85.1 ± 0.1%, and 88.4 ± 0.1%, respectively (p < 0.05). We found that the window size could largely influence the characterization results, and selected the 5 × 5 size as the best condition. We also validated the performance superiority of the proposed method with traditional classification methods.

Conclusions

These experimental results suggest that the proposed method has significant clinical applicability for IVUS-based cardiovascular diagnosis.

Automated plaque characterization using deep learning on coronary intravascular optical coherence tomographic images

Accurate identification of coronary plaque is very important for cardiologists when treating patients with advanced atherosclerosis. We developed fully-automated semantic segmentation of plaque in intravascular OCT images. We trained/tested a deep learning model on a folded, large, manually annotated clinical dataset. The sensitivities/specificities were 87.4%/89.5% and 85.1%/94.2% for pixel-wise classification of lipidous and calcified plaque, respectively. Automated clinical lesion metrics, potentially useful for treatment planning and research, compared favorably (<4%) with those derived from ground-truth labels. When we converted the results to A-line classification, they were significantly better (p < 0.05) than those obtained previously by using deep learning classifications of A-lines.

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.