Prediction of primary Hypertension in Primary Health Care Settings in Coastal Karnataka Using Artificial Neural Network.

Abstract

Background: Hypertension, characterized by chronically elevated blood pressure, poses a significant global health burden. Its prevalence, a critical public health concern, necessitates ac-curate prediction models for timely intervention and management.

Aim: The proposed approach leverages the capability of an Artificial Neural Network to capture complex patterns and non-linear relationships within the time series data, allowing for the devel-opment of a robust forecasting model to predict Hypertension. The study population consisted of known hypertensives. In this study, historical time series data related to Hypertension, including patient demographics, lifestyle factors, and medical records, were collected from a Rural primary health center associated with the medical college in coastal Karnataka, India, which is employed to train and validate the model.

Methods: The performance of the Artificial Neural Network (ANN) is evaluated using metrics such as MAE (Mean Absolute Error) and RMSE (Root Mean Squared Error) on a separate test dataset. This research explores the potential of ANN in time series forecasting of Hypertension.

Result: ANN performs well for this data and has been chosen as the best algorithm for this data set, as it has the lowest MAP (0.20) and MAE (0.45) and the highest R-Square (0.89), making it the most accurate and reliable model for the given data. If these algorithms prove beneficial, they can be used in the primary prevention of Hypertension. Individuals, institutions, and even govern-ment bodies can use it to save treatment costs and lives.

Conclusion: The ANN model demonstrated reasonably accurate predictions despite the lower overall fit. It has shown the potential to be used as a primary healthcare tool by helping physicians predict and warn about the dangers of elevated blood pressure to patients. These algorithms, de-ployed using a web application, will enable people to evaluate themselves in the comfort of their homes. This would make us inch closer to the WHO's broader goal of making health a universal right, irrespective of a person's place of residence.