Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
October 26, 2022Frontiers in Cardiovascular MedicineOpen Access

Machine learning models, such as random forest (AUC 0.92; 95% CI 0.85-0.99), demonstrated higher prognostic performance for predicting major adverse cardiovascular events than conventional scores.

View Full Paper
Ask AI
Bookmark
Share

Why the study?

The study aimed to compare the prognostic performance of conventional scoring systems with machine learning models on CCTA for predicting MACE and identifying key contributing factors.

Do machine learning models using CCTA scans improve prognostic prediction for MACE compared to conventional scoring systems in patients undergoing CCTA?

Population

416 patients referred for CCTA

Comparison

Seven machine learning models vs six conventional scoring systems

Design

Observational cohort study

Follow-up

20.5 ± 7.9 months

Key result

Machine learning models, such as random forest (AUC 0.92; 95% CI 0.85-0.99), demonstrated higher prognostic performance for predicting major adverse cardiovascular events than conventional scores.

Authors

SGSeyyed Mojtaba GhorashiAFAmir FazeliBHBehnam Hedayat

Discussion

Loading...

Member takes

Overview

ML models may refine MACE risk stratification on CCTA; hypothesis-generating and requires prospective validation before practice change.

Study Design

Type

Observational (n=416)

Structured PICO

Do machine learning models using CCTA scans improve prognostic prediction for MACE compared to conventional scoring systems in patients undergoing CCTA?

P
Population
416 patients (mean age 60, 46.2% female) referred for CCTA, followed for a mean of 20.5 months to assess MACE prediction.
E
Exposure
Machine learning models (7 models designed) using CCTA scans
C
Comparator
Conventional scoring systems (6 scores calculated)
O
Outcome
Composite of all-cause mortality, non-fatal myocardial infarction, late coronary revascularization, and hospitalization for unstable angina or heart failurecomposite

Main Result

Effect estimate: AUC 0.92 (95% CI 0.85-0.99)

Machine learning models applied to CCTA scans provide better prognostic prediction for MACE than conventional scoring systems, with anatomical features being the most important predictors.

Cite This Study

Ghorashi et al. (2022) conducted an observational in Patients referred for coronary computed tomography angiography (CCTA) (n=416). Machine learning models vs. Conventional scoring systems was evaluated on Major adverse cardiovascular events (all-cause mortality, non-fatal myocardial infarction, late coronary revascularization, and hospitalization for unstable angina or heart failure) (AUC 0.92, 95% CI 0.85-0.99). Machine learning models, such as random forest (AUC 0.92; 95% CI 0.85-0.99), demonstrated higher prognostic performance for predicting major adverse cardiovascular events than conventional scores.

synapsesocial.com/papers/6a23d1cd55bd20cf6fa63dd9https://doi.org/10.3389/fcvm.2022.994483
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1The obesity paradox in acute coronary syndrome: a meta-analysis2014 · 278 citations
  2. 2Multicollinearity in Regression Analyses Conducted in Epidemiologic Studies2016 · 1,317 citations
  3. 3National Kidney Foundation Practice Guidelines for Chronic Kidney Disease: Evaluation, Classification, and Stratification2003 · 4,651 citations
  4. 4Clinical risk factors and atherosclerotic plaque extent to define risk for major events in patients without obstructive coronary artery disease: the long-term coronary computed tomography angiography CONFIRM registry2020 · 84 citations
  5. 5Differences in Prevalence, Extent, Severity, and Prognosis of Coronary Artery Disease Among Patients With and Without Diabetes Undergoing Coronary Computed Tomography Angiography2012 · 179 citations