Machine Learning Models for Cardiovascular Disease Events Prediction
Problem
Cardiovascular diseases cause high mortality worldwide, and early prediction of CVD-related death within ten years could enable prevention.
Why it matters
CVDs are among the most serious disorders leading to high global mortality rates.
What the AI did
Six machine learning models (LR, SVM, RF, NB, XGB, AdaBoost) were trained on clinical risk factors and biochemical data from the LURIC cohort to predict CVD-caused death within ten years.
Results
2943 patients analyzed (484 dead due to CVD); Logistic Regression found most reliable; metric values not stated in source.
Impact
Comparative analysis on a real patient cohort identified the best-performing prediction model.
Record
| Venue | Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society |
|---|---|
| Published | 2022 |
| Work type | conference-paper |
| Authors | Konstantina Tsarapatsani, Antonis I. Sakellarios, Vasileios C. Pezoulas, Vassilis Tsakanikas, Marcus Edi Kleber, Winfried März, Lampros K. Michalis, Dimitrios I. Fotiadis |
| Citations | 22 (OpenAlex, at import time) |
| References | 25 |
| Model or method | Logistic Regression, SVM, Random Forest, Naïve Bayes, XGBoost, AdaBoost |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1109/embc48229.2022.9871121 |
| Open access | closed |
| Keywords | machine learning, cardiovascular disease, prediction, LURIC cohort, logistic regression |
Importance rating
major — millions of lives, an industry or a field
Abstract states CVDs are among the most serious disorders leading to high mortality rates worldwide; early 10-year death prediction could enable prevention. Cohort of 2943 patients, LR best at 72.20% accuracy.