Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
June 4, 2026Applied SciencesOpen Access

Machine learning models fail to outperform clinician baseline accuracy for CAD prognosis.

View Full Paper
Ask AI
Bookmark
Share

Why the study?

Prior machine learning studies relying on multimodal clinical, imaging, and genetic data do not provide early screening or prognosis, leaving the feasibility of using lifestyle and medical history variables alone unclear.

Can machine learning models using lifestyle and medical history data accurately predict CAD prognosis compared to clinicians?

Population

571 participants with and without CAD

Comparison

Multiple ML models using lifestyle and medical history variables vs clinician baseline

Design

Cohort study with 10-fold cross-validation

Key result

Machine learning models using lifestyle and medical history variables achieved up to 76% accuracy for CAD prognosis, compared to a clinician baseline of 78.8%.

Authors

ASAgorastos-Dimitrios SamarasIAIoannis D. ApostolopoulosEPElpiniki Papageorgiou

Discussion

Loading...

Member takes

Overview

ML models approach clinician accuracy for CAD prognosis; leaves open their role as adjunct screening tools pending validation.

Key Points

  • The aim is to evaluate how well lifestyle and medical history data can predict coronary artery disease prognosis using machine learning models.
  • Cohort of 571 participants with and without CAD was analyzed.
  • Multiple machine learning models were evaluated including Random Forest and XGBoost using 10-fold cross-validation.
  • An explainability analysis was conducted for the best-performing model using SHAP.
  • Predictive performance across models ranged from 72% to 76% accuracy, with the best model achieving 76%.
  • The clinician baseline performance was 78.8%, indicating that ML models can approach clinical diagnostic accuracy.
  • SHAP analysis provided insights into the features influencing the model predictions.

Study Design

Type

Observational (n=571)

Structured PICO

Can machine learning models using lifestyle and medical history data accurately predict CAD prognosis compared to clinicians?

P
Population
571 participants with and without CAD evaluated to determine if CAD prognosis can be achieved using lifestyle and medical history variables alone.
E
Exposure
Machine learning models (Random Forest, CatBoost, AdaBoost, XGBoost, TabPFN, and k-Nearest Neighbors) using lifestyle and medical history variables
C
Comparator
Clinician baseline
O
Outcome
Diagnostic accuracy for CAD prognosis

Main Result

Absolute Event Rate: 76% vs 78.8%

Machine learning models using only lifestyle and medical history data achieve moderate accuracy (up to 76%) for CAD prognosis, slightly lower than clinician baseline, highlighting their potential as complementary early screening tools.

Cite This Study

Samaras et al. (2026) conducted an observational in Coronary artery disease (CAD) (n=571). Machine learning models using lifestyle and medical history variables vs. Clinician baseline was evaluated on Predictive accuracy for CAD prognosis. Machine learning models using lifestyle and medical history variables achieved up to 76% accuracy for CAD prognosis, compared to a clinician baseline of 78.8%.

synapsesocial.com/papers/6a2117a4d499ed480b1707e7https://doi.org/10.3390/app16115444
View Full Paper
Ask AI
Bookmark
Share