Predicting Coronary Artery Calcium Severity based on Non-Contrast Cardiac CT images using Deep Learning

TL;DR

This study tackles efficient risk stratification of coronary artery calcium (CAC) by predicting CAC severity from non-contrast, ECG-gated CT images using a customised convolutional neural network (CNN) to classify into six categories: $0$, $0-10$, $10-100$, $101-400$, $401-1000$, and $>$1000. Using 68 scans with ground-truth labels from semiautomatic CAC scoring, the authors trained and evaluated the CNN with five-fold cross-validation, achieving an overall accuracy of 96.5% and a Cohen's κ of $0.962$, with high per-class performance and only 32 misclassifications (mostly biased toward higher categories). The results demonstrate the feasibility of finer-grained CAC risk stratification with CNNs, potentially reducing radiologist workload and interobserver variability, though limitations include a small, single-site dataset and artefact-driven misclassifications; future work should explore ROI segmentation and multi-site data to enhance generalisability. Overall, the work supports adopting expanded CAC category schemes in clinical practice when powered by robust deep learning models.

Abstract

Cardiovascular disease causes high rates of mortality worldwide. Coronary artery calcium (CAC) scoring is a powerful tool to stratify the risk of atherosclerotic cardiovascular disease. Current scoring practices require time-intensive semiautomatic analysis of cardiac computed tomography by radiologists and trained radiographers. The purpose of this study is to develop a deep learning convolutional neural networks (CNN) model to classify the calcium score in cardiac, non-contrast computed tomography images into one of six clinical categories. A total of 68 patient scans were retrospectively obtained together with their respective reported semiautomatic calcium score using an ECG-gated GE Discovery 570 Cardiac SPECT/CT camera. The dataset was divided into training, validation and test sets. Using the semiautomatic CAC score as the reference label, the model demonstrated high performance on a six-class CAC scoring categorisation task. Of the scans analysed, the model misclassified 32 cases, tending towards overestimating the CAC in 26 out of 32 misclassifications. Overall, the model showed high agreement (Cohen's kappa of 0.962), an overall accuracy of 96.5% and high generalisability. The results suggest that the model outputs were accurate and consistent with current semiautomatic practice, with good generalisability to test data. The model demonstrates the viability of a CNN model to stratify the calcium score into an expanded set of six clinical categories.

Figures (4)

• Figure 1: Sample Non-contrast cardiac CT images used in this study.
• Figure 2: Customised CNN architecture for predicting severity of calcium in CTCA images.
• Figure 3: Confusion matrix showing true vs predicted labels for 6-class classification.
• Figure 4: Qualitative results showing misclassified scans with their true and predicted class label.