Optimizing hypertension prediction using ensemble learning approaches DOI Creative Commons

Isteaq Kabir Sifat,

Md. Kaderi Kibria

PLoS ONE, Journal Year: 2024, Volume and Issue: 19(12), P. e0315865 - e0315865

Published: Dec. 23, 2024

Hypertension (HTN) prediction is critical for effective preventive healthcare strategies. This study investigates how well ensemble learning techniques work to increase the accuracy of HTN models. Utilizing a dataset 612 participants from Ethiopia, which includes 27 features potentially associated with risk, we aimed enhance predictive performance over traditional single-model methods. A multi-faceted feature selection approach was employed, incorporating Boruta, Lasso Regression, Forward and Backward Selection, Random Forest importance, found 13 common that were considered prediction. Five machine (ML) models such as logistic regression (LR), artificial neural network (ANN), random forest (RF), extreme gradient boosting (XGB), light (LGBM), stacking model trained using selected predict HTN. The models’ on testing set evaluated accuracy, precision, recall, F1-score, area under curve (AUC). Additionally, SHapley Additive exPlanations (SHAP) utilized examine impact individual predictions identify most important risk factors emerged predicting achieving an 96.32%, precision 95.48%, recall 97.51%, F1-score 96.48%, AUC 0.971. SHAP analysis identified weight, drinking habits, history hypertension, salt intake, age, diabetes, BMI, fat intake significant interpretable Our results demonstrate advancements in robustness, highlighting potential pivotal tool analytics. research contributes ongoing efforts optimize models, ultimately supporting early intervention personalized management.

Language: Английский

A prediction study on the occurrence risk of heart disease in older hypertensive patients based on machine learning DOI Creative Commons
Fei Si, Qian Liu, Jing Yu

et al.

BMC Geriatrics, Journal Year: 2025, Volume and Issue: 25(1)

Published: Jan. 11, 2025

Constructing a predictive model for the occurrence of heart disease in elderly hypertensive individuals, aiming to provide early risk identification. A total 934 participants aged 60 and above from China Health Retirement Longitudinal Study with 7-year follow-up (2011-2018) were included. Machine learning methods (logistic regression, XGBoost, DNN) employed build predicting patients. Model performance was comprehensively assessed using discrimination, calibration, clinical decision curves. After older patients, 243 individuals (26.03%) developed disease. Older patients baseline comorbid dyslipidemia, chronic pulmonary diseases, arthritis or rheumatic diseases faced higher future Feature selection significantly improved compared original variable set. The ROC-AUC logistic DNN 0.60 (95% CI: 0.53-0.68), 0.64 0.57-0.71), 0.67 0.60-0.73), respectively, regression achieving optimal calibration. XGBoost demonstrated most noticeable benefit as threshold increased. effectively identifies based on data CHARLS cohort. results suggest that have developing This information could facilitate identification

Language: Английский

Citations

0
Comparative analysis of machine learning approaches for predicting the risk of vaginal laxity DOI Creative Commons

Hongguo Zhao,

Peng Liu, Fei Chen

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Jan. 24, 2025

This study develops predictive models for Chinese female patients with VL utilizing machine learning techniques. The aim is to create an effective model that can assist in clinical diagnosis and treatment of vaginal relaxation, thereby enhancing women's pelvic floor health. In total, 1184 women have been randomly selected categorized into groups using the finger measurement method. Among them, there are 383 cases mild VL, 405 moderate 396 severe VL. Concurrently, healthy without who underwent routine health examinations chosen at random assigned non-VL group. Based on 1580 cases, we established LightGBM, Random Forest, XGBoost, AdaBoost based training dataset 5-fold cross-validation GridSearch, analyzed performance hold-out test dataset. confusion matrix, precision, recall, F1 score, overall accuracy, ROC curve compared. accuracy LightGBM model, RF XGBoost 0.8987, 0.8457, respectively. average AUC 0.976, one 0.9763, 0.9775, 0.928. has more comprehensive reasonable among four prediction models, which accurately distinguish between healthy, as well doctors diagnosing persons' conditions accurately, devising personalized plans, avoiding unnecessary surgeries, reducing psychological stress, improving patient compliance outcomes, thus results.

Language: Английский

Citations

0
Enhancing risk management in hospitals: leveraging artificial intelligence for improved outcomes DOI Creative Commons

Ranieri Guerra

Italian Journal of Medicine, Journal Year: 2024, Volume and Issue: 18(2)

Published: April 15, 2024

In hospital settings, effective risk management is critical to ensuring patient safety, regulatory compliance, and operational effectiveness. Conventional approaches assessment mitigation frequently rely on manual procedures retroactive analysis, which might not be sufficient recognize respond new risks as they arise. This study examines how artificial intelligence (AI) technologies can improve in healthcare facilities, fortifying safety precautions guidelines while improving the standard of care overall. Hospitals proactively identify mitigate risks, optimize resource allocation, clinical outcomes by utilizing AI-driven predictive analytics, natural language processing, machine learning algorithms. The different applications AI are discussed this paper, along with opportunities, problems, suggestions for their use settings.

Language: Английский

Citations

3
Prediction of myofascial pelvic pain syndrome based on random forest model DOI Creative Commons
Hang Yu,

Hongguo Zhao,

Dongxia Liu

et al.

Heliyon, Journal Year: 2024, Volume and Issue: 10(11), P. e31928 - e31928

Published: May 27, 2024

The objective is to construct a random forest model for predicting the occurrence of Myofascial pelvic pain syndrome (MPPS) and compare its performance with logistic regression demonstrate superiority model.

Language: Английский

Citations

2
A comparative analysis of machine learning algorithms with tree-structured parzen estimator for liver disease prediction DOI Creative Commons
Rakibul Islam, Azrin Sultana, Md. Nuruzzaman Tuhin

et al.

Healthcare Analytics, Journal Year: 2024, Volume and Issue: 6, P. 100358 - 100358

Published: Aug. 16, 2024

The liver is one of the most essential organs in body, which helps with metabolism and keeping body healthy. Successful treatments better patient outcomes depend on early correct Liver Disease (LD) diagnosis identification. This study proposes a system for predicting LD by combining techniques Machine Learning (ML) algorithms that include Decision Tree, Random Forest, Extra Tree Classifier (ETC), LightGBM, Adaboost, Tree-Structured Parzen Estimator (TPE) method hyperparameter tuning. No previous literature research has utilized ML TPE to predict LD. For this research, Indian Patients' Dataset 583 instances 11 attributes was used. In pre-processing data, such as upsampling have been address class imbalance problem. Normalization employed scale dataset, feature selection applied choose important features. proposed model analyzed compared using 10-fold cross-validation process, various evaluation metrics including accuracy, precision, recall, F1-score. achieved best level accuracy while employing ETC approach, recorded 95.8%.

Language: Английский

Citations

2
Design of upper limb muscle strength assessment system based on surface electromyography signals and joint motion DOI Creative Commons
Siqi Wang, Lai Wei, Yudong Zhang

et al.

Frontiers in Neurology, Journal Year: 2024, Volume and Issue: 15

Published: Dec. 13, 2024

Purpose This study aims to develop a assessment system for evaluating shoulder joint muscle strength in patients with varying degrees of upper limb injuries post-stroke, using surface electromyographic (sEMG) signals and motion data. Methods The includes modules acquiring electromyography (EMG) EMG from the anterior, middle, posterior deltoid muscles were collected, filtered, denoised extract time-domain features. Concurrently, data captured MPU6050 sensor processed feature extraction. extracted features sEMG analyzed three algorithms: Random Forest (RF), Backpropagation Neural Network (BPNN), Support Vector Machines (SVM), predict through regression models. Model performance was evaluated Root Mean Squared Error ( RMSE ), R-Square R 2 Absolute MAE Bias MBE identify most accurate prediction algorithm. Results effectively collected Among models tested, Regression (SVR) model achieved highest accuracy an 0.8059, 0.2873, 0.2155, 0.0071. 0.7997, 0.3039, 0.2405, 0.0090. BPNN 0.7542, 0.3173, 0.2306, 0.0783. Conclusion SVR demonstrated superior predicting strength. RF model, its importance capabilities, provides valuable insights that can assist therapists process.

Language: Английский

Citations

0
Artificial Intelligence and Aging DOI

Rodrigo Edgar Palacios Leyva,

Luis Enrique Sucar Succar,

Héctor Hugo Avilés Arriaga

et al.

Published: Jan. 1, 2024

Language: Английский

Citations

0
Optimizing hypertension prediction using ensemble learning approaches DOI Creative Commons

Isteaq Kabir Sifat,

Md. Kaderi Kibria

PLoS ONE, Journal Year: 2024, Volume and Issue: 19(12), P. e0315865 - e0315865

Published: Dec. 23, 2024

Hypertension (HTN) prediction is critical for effective preventive healthcare strategies. This study investigates how well ensemble learning techniques work to increase the accuracy of HTN models. Utilizing a dataset 612 participants from Ethiopia, which includes 27 features potentially associated with risk, we aimed enhance predictive performance over traditional single-model methods. A multi-faceted feature selection approach was employed, incorporating Boruta, Lasso Regression, Forward and Backward Selection, Random Forest importance, found 13 common that were considered prediction. Five machine (ML) models such as logistic regression (LR), artificial neural network (ANN), random forest (RF), extreme gradient boosting (XGB), light (LGBM), stacking model trained using selected predict HTN. The models’ on testing set evaluated accuracy, precision, recall, F1-score, area under curve (AUC). Additionally, SHapley Additive exPlanations (SHAP) utilized examine impact individual predictions identify most important risk factors emerged predicting achieving an 96.32%, precision 95.48%, recall 97.51%, F1-score 96.48%, AUC 0.971. SHAP analysis identified weight, drinking habits, history hypertension, salt intake, age, diabetes, BMI, fat intake significant interpretable Our results demonstrate advancements in robustness, highlighting potential pivotal tool analytics. research contributes ongoing efforts optimize models, ultimately supporting early intervention personalized management.

Language: Английский

Citations

0