Deep Learning Coronary Artery Calcium Scores from SPECT/CT Attenuation Maps Improve Prediction of Major Adverse Cardiac Events

Abstract

Low-dose ungated CT attenuation correction (CTAC) scans are commonly obtained with SPECT/CT myocardial perfusion imaging. Despite the characteristically low image quality of CTAC, deep learning (DL) can potentially quantify coronary artery calcium (CAC) from these scans in an automatic manner. We evaluated CAC quantification derived with a DL model, including correlation with expert annotations and associations with major adverse cardiovascular events (MACE). Methods: We trained a convolutional long short-term memory DL model to automatically quantify CAC on CTAC scans using 6,608 studies (2 centers) and evaluated the model in an external cohort of patients without known coronary artery disease (n = 2,271) obtained in a separate center. We assessed agreement between DL and expert annotated CAC scores. We also assessed associations between MACE (death, revascularization, myocardial infarction, or unstable angina) and CAC categories (0, 1-100, 101-400, or >400) for scores manually derived by experienced readers and scores obtained fully automatically by DL using multivariable Cox models (adjusted for age, sex, past medical history, perfusion, and ejection fraction) and net reclassification index. Results: In the external testing population, DL CAC was 0 in 908 patients (40.0%), 1-100 in 596 (26.2%), 100-400 in 354 (15.6%), and >400 in 413 (18.2%). Agreement in CAC category by DL CAC and expert annotation was excellent (linear weighted κ, 0.80), but DL CAC was obtained automatically in less than 2 s compared with about 2.5 min for expert CAC. DL CAC category was an independent risk factor for MACE with hazard ratios in comparison to a CAC of zero: CAC of 1-100 (2.20; 95% CI, 1.54-3.14; P < 0.001), CAC of 101-400 (4.58; 95% CI, 3.23-6.48; P < 0.001), and CAC of more than 400 (5.92; 95% CI, 4.27-8.22; P < 0.001). Overall, the net reclassification index was 0.494 for DL CAC, which was similar to expert annotated CAC (0.503). Conclusion: DL CAC from SPECT/CT attenuation maps agrees well with expert CAC annotations and provides a similar risk stratification but can be obtained automatically. DL CAC scores improved classification of a significant proportion of patients as compared with SPECT myocardial perfusion alone.

Keywords: artificial intelligence; cardiology; coronary artery calcification; deep learning; myocardial perfusion imaging; risk stratification.

Graphical abstract

FIGURE 1.

Outline of model architecture. ConvLSTM includes network trained to segment CAC, as well as second network for segmentation of heart, which limits CAC scoring. Softmax argmax function normalizes output of network to expected probabilities. Model identifies coronary calcium (red) and noncoronary calcium (green) within heart mask.

FIGURE 2.

Examples of expert scores compared with DL CAC scores. Model identifies coronary calcium (red) and noncoronary calcium (green). In case 1, expert and DL annotations identified similar left circumflex CAC as well as ascending aorta calcium. No CAC was identified by either expert or DL scoring in case 2. In case 3, expert and DL annotations identified similar right coronary artery CAC as well as mitral annular calcification. BMI = body mass index.

FIGURE 3.

Concordance matrix between DL and expert CAC categories in external testing population.

FIGURE 4.

Kaplan–Meier survival curves for MACE. Increasing CAC category was associated with increasing risk of MACE for DL and expert annotated CAC scores on SPECT/CT attenuation maps.

FIGURE 5.

Results of net-reclassification analysis. We assessed addition of CAC categories to full multivariable model outlined in Table 2.