Coronary disease & ACSCohort study

AI-Guided Quantitative Plaque Staging Predicts Long-Term Cardiovascular Outcomes in Patients at Risk for Atherosclerotic CVD

JACC Cardiovasc ImagingPublished

  • Prospective cohort study
  • long-term follow-up

Authors' conclusion

Through 10-year follow-up, AI-QCT plaque staging showed important prognostic value for MACE and showed additional discriminatory value over clinical risk factors, CACS, and manual guideline-recommended CCTA assessment.

Study context

Population
Patients referred for suspected CAD undergoing coronary computed tomography angiography (CCTA) analysed with AI-QCT
Participants
536 patients
Exposure
AI-QCT plaque stages vs stages 0 and 1

Main finding

As reported in the abstract

Major adverse cardiac event (MACE) · median 10.3 years
Each group vs stage 0–1Adjusted hazard ratio (95% CI)P value
Stage 3
3.57(2.12–6.00)
P < 0.001
Noncalcified plaque volume >7.5%
4.37(2.51–7.62)
P < 0.001
Patients with primary outcome
114 patients (21%)

Observational data: associations, not effects.

Further findings

OutcomeValuesnet reclassification improvement (95% CI)P value
Discrimination: clinical risk factors + CACS + AI-QCT vs clinical risk factors + CACS · 10 years · P for AUC comparisonWith AI-QCT AUC 0.82 [95% CI: 0.78-0.87] · Without AI-QCT AUC 0.73 [95% CI: 0.68-0.79]0.21 (0.09–0.38)P < 0.001
Discrimination: AI-QCT vs CAD-RADS 2.0 and manual QCT · 10 years — AI-QCT vs CAD-RADS 2.0 · P for AUC comparison; NRI modestCAD-RADS 2.0 AUC 0.78; 95% CI: 0.73-0.83 · Manual QCT AUC 0.78; 95% CI: 0.73-0.830.09 (−0.02 to 0.29)P = 0.023
Discrimination: AI-QCT vs CAD-RADS 2.0 and manual QCT · 10 years — AI-QCT vs Manual QCT · P for AUC comparison; NRI modest0.04 (−0.05 to 0.27)P = 0.040

Parts of this Brief are withheld: the automatic checks could not confirm them against the abstract.

Original publication

AI-Guided Quantitative Plaque Staging Predicts Long-Term Cardiovascular Outcomes in Patients at Risk for Atherosclerotic CVD. JACC Cardiovasc Imaging. Published 19 July 2023. doi:10.1016/j.jcmg.2023.05.020

Summarised from the published abstract; every number is checked against it