Automated vessel-specific coronary artery calcification quantification with deep learning in a large multi-centre registry

Abstract

Aims: Vessel-specific coronary artery calcification (CAC) is additive to global CAC for prognostic assessment. We assessed accuracy and prognostic implications of vessel-specific automated deep learning (DL) CAC analysis on electrocardiogram (ECG) gated and attenuation correction (AC) computed tomography (CT) in a large multi-centre registry.

Methods and results: Vessel-specific CAC was assessed in the left main/left anterior descending (LM/LAD), left circumflex (LCX), and right coronary artery (RCA) using a DL model trained on 3000 gated CT and tested on 2094 gated CT and 5969 non-gated AC CT. Vessel-specific agreement was assessed with linear weighted Cohen's Kappa for CAC zero, 1-100, 101-400, and >400 Agatston units (AU). Risk of major adverse cardiovascular events (MACE) was assessed during 2.4 ± 1.4 years follow-up, with hazard ratios (HR) and 95% confidence intervals (CI). There was strong to excellent agreement between DL and expert ground truth for CAC in LM/LAD, LCX and RCA on gated CT [0.90 (95% CI 0.89 to 0.92); 0.70 (0.68 to 0.73); 0.79 (0.77 to 0.81)] and AC CT [0.78 (0.77 to 0.80); 0.60 (0.58 to 0.62); 0.70 (0.68 to 0.71)]. MACE occurred in 242 (12%) undergoing gated CT and 841(14%) of undergoing AC CT. LM/LAD CAC >400 AU was associated with the highest risk of MACE on gated (HR 12.0, 95% CI 7.96, 18.0, P < 0.001) and AC CT (HR 4.21, 95% CI 3.48, 5.08, P < 0.001).

Conclusion: Vessel-specific CAC assessment with DL can be performed accurately and rapidly on gated CT and AC CT and provides important prognostic information.

Keywords: computed tomography; coronary artery calcification; coronary artery disease; deep learning.

Graphical Abstract

Vessel-specific coronary artery calcium (CAC) from electrocardiogram gated and attenuation correction CT patients in a large multi-centre registry with Deep Learning can be performed accurately, rapidly, and provides important prognostic information for Major Adverse Cardiac Events (MACE).

Figure 1

DL model architecture includes two models in parallel, the heart network produces the cardiac silhouette to avoid any additional or spurious calcification and the multi-vessel CAC convolutional long short term memory (LSTM) network for per vessel quantification. Red—the LM coronary artery, dark blue—the LAD artery, green—the left circumflex artery.

Figure 2

Vessel-specific DL CAC scoring in patients with 0, 1–100, 101–400, and >400 AU CAC assessed on ECG gated CT by expert ground truth annotation and DL. Dark blue—the LAD artery, green—the LCX artery, light blue—the ascending aorta.

Figure 3

Vessel-specific DL CAC scoring in patients with 0, 1–100, 101–400, and >400 AU CAC assessed on CTAC by expert ground truth annotation and DL. Red—the LM coronary artery, dark blue—the LAD artery.

Figure 4

Agreement between DL and expert ground truth CAC for the all-vessel score, LM/LAD, LCX, and RCA for male and female patients undergoing ECG gated CT and AC CT.

Figure 5

Kaplan–Meier curves for the occurrence of MACE in patients with different severity of CAC assessed with DL on ECG gated CT. Total and vessel-specific CAC are assessed in the LM/LAD, LCX, and RCA. *** implies P < 0.001, ** implies P < 0.01, * implies P < 0.05.

Figure 6

Kaplan–Meier curves for the occurrence of MACE in patients with different severity of CAC assessed with DL on AC CT. Total and vessel-specific CAC are assessed in the LM/LAD, LCX, and right coronary artery (RCA). *** implies P < 0.001, ** implies P < 0.01, * implies P < 0.05.

Figure 7

Forrest plots showing univariable (A, B) and multi-variable (C, D) cox proportional HR (cox proportional hazards) for the risk of MACE with vessel-specific DL coronary artery calcium scores from ECG gated CT (A, C) and AC CT (B, D). HR and 95% CI are presented for calcium score groups (1–100, 101–400, and > 400 AU) compared to patients with zero CAC for each of the RCA, LCX, LM/LAD, and all vessels.