Back to Search
Start Over
A prediction model of blood pressure for telemedicine.
- Source :
- Health Informatics Journal; Sep2018, Vol. 24 Issue 3, p227-244, 18p
- Publication Year :
- 2018
-
Abstract
- This paper presents a new study based on a machine learning technique, specifically an artificial neural network, for predicting systolic blood pressure through the correlation of variables (age, BMI, exercise level, alcohol consumption level, smoking status, stress level, and salt intake level). The study was carried out using a database containing a variety of variables/factors. Each database of raw data was split into two parts: one part for training the neural network and the remaining part for testing the performance of the network. Two neural network algorithms, back-propagation and radial basis function, were used to construct and validate the prediction system. According to the experiment, the accuracy of our predictions of systolic blood pressure values exceeded 90%. Our experimental results show that artificial neural networks are suitable for modeling and predicting systolic blood pressure. This new method of predicting systolic blood pressure helps to give an early warning to adults, who may not get regular blood pressure measurements that their blood pressure might be at an unhealthy level. Also, because an isolated measurement of blood pressure is not always very accurate due to daily fluctuations, our predictor can provide the predicted value as another figure for medical staff to refer to. [ABSTRACT FROM AUTHOR]
- Subjects :
- BLOOD pressure
TELEMEDICINE
MACHINE learning
ARTIFICIAL neural networks
SYSTOLIC blood pressure
ALCOHOL drinking
SMOKING
PSYCHOLOGICAL stress
MATHEMATICAL models
AGE distribution
ALCOHOLISM
BLOOD pressure measurement
HYPERTENSION
MEDICAL informatics
SEX distribution
BODY mass index
EXERCISE intensity
Subjects
Details
- Language :
- English
- ISSN :
- 14604582
- Volume :
- 24
- Issue :
- 3
- Database :
- Complementary Index
- Journal :
- Health Informatics Journal
- Publication Type :
- Academic Journal
- Accession number :
- 131074727
- Full Text :
- https://doi.org/10.1177/1460458216663025