EUVIMEDEUVIMED entdecken
Crossref

A Hybrid Machine Learning and Optimization Framework for Early Cardiovascular Disease Risk Stratification Using Large-Scale Clinical Data

Zainab Marid Alzamili

Journal of Al-Farabi for Engineering Sciences · 2026

DOI: 10.59746/9nsn1y58 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Al-Farabi for Engineering Sciences, 2026. Verfasst von Zainab Marid Alzamili.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Cardiovascular Disease (CVD) is recognized as the major cause of death globally and robust, early risk stratification models are needed. Standard clinical decision rules, including the Framingham Risk Score (FRS), are designed and based on sparse, easily computable, and static variables and do not explain the full complexity of non-linear relationships of variables in a "big data" context such as an Electronic Health Record (EHR). This paper. proposes a novel clinical optimization and risk prediction framework, dubbed: Hybrid Clinical Optimization and Risk Prediction System (HCORPS). The architecture features the use of a dynamic feature selection system based on a meta-heuristic Particle Swarm Optimization (PSO) algorithm on a large-scale clinical data and a deep ensemble machine learning (ME) classifier consisting of Extreme Gradient Boosting and Multi-Layer Perceptrons. HCORPS prunes away unnecessary clinical noise, while maximizing the boundaries for the key hyperparameters, in a cohort of 185,400 patients across various unstructured physiology, lab tests, and demographics. When tested against standard clinical outcomes, HCORPS has an Area Under the Receiver Operating Characteristic Curve (AUC-ROC) of 0.942, which is 18.7% higher than the threshold. False prognostications are decreased by 22.4%. The framework offers a scalable, easy-to-understand approach that can be used to connect predictive analytics with real-time clinical decision support systems using HL7 FHIR standards, but with a reduced computational requirement.

Zitieren

APA

Alzamili, Z. M. (2026). A Hybrid Machine Learning and Optimization Framework for Early Cardiovascular Disease Risk Stratification Using Large-Scale Clinical Data. Journal of Al-Farabi for Engineering Sciences. https://doi.org/10.59746/9nsn1y58

Vancouver

Alzamili ZM. A Hybrid Machine Learning and Optimization Framework for Early Cardiovascular Disease Risk Stratification Using Large-Scale Clinical Data. Journal of Al-Farabi for Engineering Sciences. 2026. doi:10.59746/9nsn1y58

BibTeX
@article{Alzamili2026,
  author = {Alzamili, Zainab Marid},
  title = {A Hybrid Machine Learning and Optimization Framework for Early Cardiovascular Disease Risk Stratification Using Large-Scale Clinical Data},
  journal = {Journal of Al-Farabi for Engineering Sciences},
  year = {2026},
  doi = {10.59746/9nsn1y58}
}
Kostenlos starten →