EDOGBANYA, SHADRACH UGBEDEOJO (2026) DEVELOPMENT OF AN AI-BASED HYPERTENSION RISK PREDICTION AND ADVISORY SYSTEM USING XGBOOST AND GENERATIVE AI. Other thesis, Godfrey Okoye University, Enugu.
|
Text
EDOGBANYA SHADRACH UGBEDEOJO FINAL PROJECT32.pdf Download (2MB) | Preview |
Abstract
Hypertension, or high blood pressure, is a very common health condition in the world. One problem with hypertension is that many people don't know they have it as it can develop without any symptoms. People can therefore begin to function normally without knowing that their health is in danger until they develop complications like stroke, heart disease, or kidney trouble. In recent years a number of mobile and web-based health apps have been created that help people to manage their health records. These can be helpful in keeping track of the data, but many of them only collect data and do not communicate to users their risk of developing high blood pressure in the future. They also offer some assistance with regard to preventive health decisions. This project was specifically concerned with the development of HyperSense, a web-based decision support system to predict the risk of hypertension for the user and offer personalized health advice based on the user's result. This process analyzes various factors associated with hypertension and is done using the XGBoost machine learning algorithm, which is widely used for analyzing the system. These factors are analyzed to calculate a person's risk. Google Gemini was integrated into the platform to make it more functional and interesting. This enables the user to get recommendations based on their own risk profile, rather than only getting general health advice. The application was created to be responsive, allowing it to be easily accessed on a variety of devices. Once the model was developed and it was tested using the validation data. The results showed that the model was able to make predictions with a high level of accuracy. It has an accuracy of 98.24% and sensitivity of 98.35%, demonstrating its ability to accurately detect people who might be at risk for hypertension. This research shows that machine learning and generative AI can be leveraged to develop meaningful healthcare solutions. HyperSense can help guide early intervention and make more positive lifestyle choices, with both risk prediction and personalized recommendations. These systems can potentially help to promote health literacy and better control of hypertension in the general population.
| Item Type: | Thesis (Other) |
|---|---|
| Subjects: | Q Science > Q Science (General) |
| Divisions: | Faculty of Engineering, Science and Mathematics > School of Electronics and Computer Science |
| Depositing User: | MICHAEL MADUBUKO |
| Date Deposited: | 22 Jul 2026 10:41 |
| Last Modified: | 22 Jul 2026 10:41 |
| URI: | http://eprints.gouni.edu.ng/id/eprint/5987 |
Actions (login required)
 |
View Item |
