Decision Support System in Healthcare for Predicting Blood Pressure Disorders

Blood pressure problems including hypertension and hypotension are considered common among the elderly population, especially hypertension. Recently it started being common among younger adults due to many factors related to unhealthy lifestyles, stress, or genetic factors. Besides the fact of consi...

Full description

Saved in:
Bibliographic Details
Main Authors: Owess, Marwa Mustafa, Owda, Amani Yousef, Owda, Majdi
Format: Conference Proceeding
Language:English
Subjects:
blood pressure
classification
Decision support system
Decision support systems
Hypertension
Hypotension
machine learning
Predictive models
Pressure measurement
Sociology
Statistics
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Blood pressure problems including hypertension and hypotension are considered common among the elderly population, especially hypertension. Recently it started being common among younger adults due to many factors related to unhealthy lifestyles, stress, or genetic factors. Besides the fact of considering hypertension as a chronic disease, it is also considered a primary or contributing cause of complicated risky health issues such as strokes, heart diseases, and chronic kidney failure. In many cases, hypertensive patients may not be aware of their problem because they don't experience any symptoms or warning signs. For this reason, it is essential to build a decision support system to identify individuals at high risk of blood pressure problems including raised blood pressure and low blood pressure. This paper proposes a decision support model for predicting blood pressure disorders by using input variables such as sex, age, body mass index (BMI), cholesterol level, heart rate, and glucose level. The decision model helps in early warning of the potential risk of hypertension or hypotension. As a result, people under potential risk are advised to measure their blood pressure regularly and take the needed precautions or medications to avoid or control this health issue. The proposed decision support model is based on using supervised machine learning classification algorithms, mainly Random Forest, Decision Tree, and XGBoost. The experimental results show that the model achieved the best performance when implemented using a Random Forest classifier with a 10-fold cross-validation method, with an accuracy of 85.81%.
ISSN:2831-3399
DOI:10.1109/ICIT58056.2023.10226098