HRP-OG: Online Learning with Generative Feature Replay for Hypertension Risk Prediction in a Nonstationary Environment DOI Creative Commons
Shaofu Lin, Haokang Yan, Shiwei Zhou

et al.

Sensors, Journal Year: 2024, Volume and Issue: 24(15), P. 5033 - 5033

Published: Aug. 3, 2024

Hypertension is a major risk factor for many serious diseases. With the aging population and lifestyle changes, incidence of hypertension continues to rise, imposing significant medical cost burden on patients severely affecting their quality life. Early intervention can greatly reduce prevalence hypertension. Research early warning models based electronic health records (EHRs) an important effective method achieving warning. However, limited by scarcity imbalance multivisit records, nonstationary characteristics features, it difficult predict probability in patient effectively. Therefore, this study proposes online monitoring model (HRP-OG) reinforcement learning generative feature replay. It transforms prediction problem into sequential decision problem, using records. Sensors embedded devices wearables continuously capture real-time physiological data such as blood pressure, heart rate, activity levels, which are integrated EHR. The fit between samples generated generator real visit evaluated maximum likelihood estimation, adversarial discrepancy space incoming incremental data, updated incorporation sensor ensures that adapts dynamically changes condition patients, facilitating timely interventions. In study, publicly available MIMIC-III used validation, experimental results demonstrate compared existing advanced methods, HRP-OG effectively improve accuracy few-shot record environments.

Language: Английский

A Comprehensive Study on Deep Learning Models for the Detection of Ovarian Cancer and Glomerular Kidney Disease using Histopathological Images DOI

S J K Jagadeesh Kumar,

G. Prabu Kanna,

D. Prem Raja

et al.

Archives of Computational Methods in Engineering, Journal Year: 2024, Volume and Issue: unknown

Published: June 1, 2024

Language: Английский

Citations

3
Machine Learning-Based Prediction System for Risk Assessment of Hypertension Using Symptoms Investigations DOI Creative Commons
Simranjit Kaur, Khushboo Bansal, Yogesh Kumar

et al.

International Journal of experimental research and review, Journal Year: 2024, Volume and Issue: 46, P. 139 - 149

Published: Dec. 30, 2024

Hypertension is a common condition of cardiovascular disease that poses significant health challenges among the public on larger scale globally. It important to accurately predict risk hypertension save people and improve overall quality life. Traditionally, detection relies clinical criteria such as blood pressure measurement examination medical history. However, these methods have drawbacks involving potential human error, time consumption, possibility missed diagnoses. The paper aims identify features or symptoms its factors using machine learning algorithms. Apart from this, it utmost importance they play pivotal role in recognizing type for hypertension. To successfully conduct work, dataset 13 attributes, including gender, age, smoking habits, etc, has been used, which further visualized graphically understand pattern them. Later, multiple learning-based techniques applied examined basis standard metrics. Results indicate random forest models outperform existing approaches, achieving an accuracy 87.26% predicting low high-risk Furthermore, classification reports reveal superior precision, recall, F1-score forests compared alternative models. Insights curves confusion matrices provide valuable understanding model performance data sufficiency. Overall, this research highlights impact underscores ongoing efforts translate findings into practical applications.

Language: Английский

Citations

1
HRP-OG: Online Learning with Generative Feature Replay for Hypertension Risk Prediction in a Nonstationary Environment DOI Creative Commons
Shaofu Lin, Haokang Yan, Shiwei Zhou

et al.

Sensors, Journal Year: 2024, Volume and Issue: 24(15), P. 5033 - 5033

Published: Aug. 3, 2024

Hypertension is a major risk factor for many serious diseases. With the aging population and lifestyle changes, incidence of hypertension continues to rise, imposing significant medical cost burden on patients severely affecting their quality life. Early intervention can greatly reduce prevalence hypertension. Research early warning models based electronic health records (EHRs) an important effective method achieving warning. However, limited by scarcity imbalance multivisit records, nonstationary characteristics features, it difficult predict probability in patient effectively. Therefore, this study proposes online monitoring model (HRP-OG) reinforcement learning generative feature replay. It transforms prediction problem into sequential decision problem, using records. Sensors embedded devices wearables continuously capture real-time physiological data such as blood pressure, heart rate, activity levels, which are integrated EHR. The fit between samples generated generator real visit evaluated maximum likelihood estimation, adversarial discrepancy space incoming incremental data, updated incorporation sensor ensures that adapts dynamically changes condition patients, facilitating timely interventions. In study, publicly available MIMIC-III used validation, experimental results demonstrate compared existing advanced methods, HRP-OG effectively improve accuracy few-shot record environments.

Language: Английский

Citations

0