No abstract

No abstract

Significance: Detection and characterization of coronary atherosclerotic plaques often need reviews of a large number of optical coherence tomography (OCT) imaging slices to make a clinical decision. However, it is a challenge to manually review all the slices and consider the interrelationship between adjacent slices. Approach: Inspired by the recent success of deep convolutional network on the classification of medical images, we proposed a ResNet-3D network for classification of coronary plaque calcification in OCT pullbacks. The ResNet-3D network was initialized with a trained ResNet-50 network and a three-dimensional convolution filter filled with zeros padding and non-zeros padding with a convolutional filter. To retrain ResNet-50, we used a dataset of ∼4860 OCT images, derived by 18 entire pullbacks from different patients. In addition, we investigated a two-phase training method to address the data imbalance. For an improved performance, we evaluated different input sizes for the ResNet-3D network, such as 3, 5, and 7 OCT slices. Furthermore, we integrated all ResNet-3D results by majority voting. Results: A comparative analysis proved the effectiveness of the proposed ResNet-3D networks against ResNet-2D network in the OCT dataset. The classification performance (F1-scores ¼ 94% for non-zeros padding and F1-score ¼ 96% for zeros padding) demonstrated the potential of convolutional neural networks (CNNs) in classifying plaque calcification. Conclusions: This work may provide a foundation for further work in extending the CNN to voxel segmentation, which may lead to a supportive diagnostic tool for assessment of coronary plaque vulnerability.

There is a need to develop a validated algorithm for plaque characterization which can help to facilitate the standardization of optical coherence tomography (OCT) image interpretation of plaque morphology, and improve the efficiency and accuracy in the application of OCT imaging for the quantitative assessment of plaque vulnerability. In this study, a machine learning algorithm was implemented for characterization of atherosclerotic plaque components by intravascular OCT using ex vivo carotid plaque tissue samples. A total of 31 patients underwent carotid endarterectomy and the ex vivo carotid plaques were imaged with OCT. Optical parameter, texture features and relative position of pixels were extracted within the region of interest and then used to quantify the tissue characterization of plaque components. The potential of individual and combined feature set to discriminate tissue components was quantified using sensitivity, specificity, accuracy. The results show there was a lower classification accuracy in the calcified tissue than the fibrous tissue and lipid tissue. The pixel-wise classification accuracy obtained by the developed method, to characterize the fibrous, calcified and lipid tissue by comparing with histology, were 80.0, 62.0, and 83.1, respectively. The developed algorithm was capable of characterizing plaque components with an excellent accuracy using the combined feature set.

Intravascular optical coherence tomography (IVOCT) has been successfully utilized for in vivo diagnostics of coronary plaques. However, classification of atherosclerotic tissues is mainly performed manually by experienced experts, which is time-consuming and subjective. To overcome these limitations, an automatic method of segmentation and classification of IVOCT images is developed in this paper. The method is capable of detecting the plaque contour between the fibrous tissues and other components. Subsequently, the method classifies the tissues based on their texture features described by Fourier transform and discrete wavelet transform. The experimental results of 103 images show that an overall classification accuracy of over 80% in the indicator of depth and span angle is achieved in comparison to manual results. The validation suggests that this method is objective, accurate, and automatic without any manual intervention. The proposed method is able to demonstrate the artery wall morphology successfully, which is valuable for the research of atherosclerotic disease.