Development and Validation of a Hypertension Risk Prediction Model Based on Particle Swarm Optimization-Support...

Area of Science:

• Gerontology
• Computational Medicine
• Biomedical Engineering

Background:

• Hypertension is a widespread health concern, particularly in the elderly population, leading to various complications.
• Accurate and timely hypertension detection is vital for effective patient management.
• Conventional diagnostic methods may lack precision and efficiency, necessitating advanced computational approaches.

Purpose of the Study:

• To evaluate the efficacy of machine learning models, specifically Support Vector Machine (SVM) and Particle Swarm Optimization-Support Vector Machine (PSO-SVM), for enhanced hypertension detection in elderly individuals.
• To compare the performance of PSO-SVM against standard SVM in identifying hypertension.

Main Methods:

• Data from 1460 hypertensive and 1416 non-hypertensive individuals (aged 65+) were analyzed.
• Support Vector Machine (SVM) and Particle Swarm Optimization-Support Vector Machine (PSO-SVM) models were developed and validated.
• Model performance was assessed using metrics including sensitivity, specificity, accuracy, F1 score, and Area Under the Curve (AUC) of the ROC curve.

Main Results:

• The PSO-SVM model demonstrated superior performance compared to the standard SVM.
• Key performance improvements for PSO-SVM included sensitivity (93.9%), F1 score (0.838), and AUC-ROC (0.871).

Conclusions:

• The PSO-SVM model proves effective for complex classification tasks, particularly in the domain of hypertension detection.
• This advanced computational method provides a robust foundation for the early diagnosis and treatment of hypertension in older adults.