Machine learning and deep learning predictive models for type 2 diabetes: a systematic review

Diabetology & Metabolic Syndrome, Journal Year: 2021, Volume and Issue: 13(1)

Published: Dec. 1, 2021

Abstract Diabetes Mellitus is a severe, chronic disease that occurs when blood glucose levels rise above certain limits. Over the last years, machine and deep learning techniques have been used to predict diabetes its complications. However, researchers developers still face two main challenges building type 2 predictive models. First, there considerable heterogeneity in previous studies regarding used, making it challenging identify optimal one. Second, lack of transparency about features models, which reduces their interpretability. This systematic review aimed at providing answers challenges. The followed PRISMA methodology primarily, enriched with one proposed by Keele Durham Universities. Ninety were included, model, complementary techniques, dataset, performance parameters reported extracted. Eighteen different types models compared, tree-based algorithms showing top performances. Deep Neural Networks proved suboptimal, despite ability deal big dirty data. Balancing data feature selection helpful increase model’s efficiency. Models trained on tidy datasets achieved almost perfect

Language: Английский

---

A tongue features fusion approach to predicting prediabetes and diabetes with machine learning

Journal of Biomedical Informatics, Journal Year: 2021, Volume and Issue: 115, P. 103693 - 103693

Published: Feb. 2, 2021

Language: Английский

---

Machine learning for diabetes clinical decision support: a review

Advances in Computational Intelligence, Journal Year: 2022, Volume and Issue: 2(2)

Published: April 1, 2022

Language: Английский

---

Machine learning for predicting chronic diseases: a systematic review

Public Health, Journal Year: 2022, Volume and Issue: 205, P. 14 - 25

Published: Feb. 24, 2022

Language: Английский

---

Diabetes risk prediction model based on community follow-up data using machine learning

Preventive Medicine Reports, Journal Year: 2023, Volume and Issue: 35, P. 102358 - 102358

Published: Aug. 19, 2023

Diabetes is a chronic metabolic disease characterized by hyperglycemia, the follow-up management of diabetes patients mostly in community, but relationship between key lifestyle indicators community and risk unclear. In order to explore association life characteristic diabetes, 252,176 records people with from 2016 2023 were obtained Haizhu District, Guangzhou. According data, that affect are determined, optimal feature subset through selection technology accurately assess diabetes. A assessment model based on random forest classifier was designed, which used parameter algorithm comparison, an accuracy 91.24% AUC corresponding ROC curve 97%. improve applicability clinical real life, score card designed tested using original 95.15%, reliability high. The prediction big data mining can be for large-scale screening early warning doctors patient further promoting prevention control strategies, also wearable devices or intelligent biosensors individual self examination, reduce factor levels.

Language: Английский

---

Diabetes mellitus risk prediction in the presence of class imbalance using flexible machine learning methods

BMC Medical Informatics and Decision Making, Journal Year: 2022, Volume and Issue: 22(1)

Published: Feb. 10, 2022

Early detection and prediction of type two diabetes mellitus incidence by baseline measurements could reduce associated complications in the future. The low rate comparison with non-diabetes makes accurate minority class more challenging.Deep neural network (DNN), extremely gradient boosting (XGBoost), random forest (RF) performance is compared predicting Tehran Lipid Glucose Study (TLGS) cohort data. impact changing threshold, cost-sensitive learning, over under-sampling strategies as solutions to imbalance have been improving algorithms performance.DNN highest accuracy diabetes, 54.8%, outperformed XGBoost RF terms AUROC, g-mean, f1-measure original imbalanced Changing threshold based on maximum improved three algorithms. Repeated edited nearest neighbors (RENN) DNN learning tree-based were best tackle issue. RENN increased ROC Precision-Recall AUCs, g-mean from 0.857, 0.603, 0.713, 0.575 0.862, 0.608, 0.773, 0.583, respectively DNN. Weighing 0.667, 0.554 0.776, 0.588 XGBoost, 0.659, 0.543 0.775, 0.566 RF, respectively. Also, AUCs 0.840, 0.578 0.846, 0.591, respectively.G-mean experienced most increase all solutions. efficient strategies, resampling methods are faster handle imbalance. Among sampling had better than others.

Language: Английский

---

Predicting the Risk of Incident Type 2 Diabetes Mellitus in Chinese Elderly Using Machine Learning Techniques

Journal of Personalized Medicine, Journal Year: 2022, Volume and Issue: 12(6), P. 905 - 905

Published: May 31, 2022

Early identification of individuals at high risk diabetes is crucial for implementing early intervention strategies. However, algorithms specific to elderly Chinese adults are lacking. The aim this study build effective prediction models based on machine learning (ML) the type 2 mellitus (T2DM) in elderly. A retrospective cohort was conducted using health screening data older than 65 years Wuhan, China from 2018 2020. With a strict filtration, 127,031 records eligible participants were utilized. Overall, 8298 diagnosed with incident T2DM during 2-year follow-up (2019-2020). dataset randomly split into training set (n = 101,625) and test 25,406). We developed four ML algorithms: logistic regression (LR), decision tree (DT), random forest (RF), extreme gradient boosting (XGBoost). Using LASSO regression, 21 features selected. Random under-sampling (RUS) applied address class imbalance, Shapley Additive Explanations (SHAP) used calculate visualize feature importance. Model performance evaluated by area under receiver operating characteristic curve (AUC), sensitivity, specificity, accuracy. XGBoost model achieved best (AUC 0.7805, sensitivity 0.6452, specificity 0.7577, accuracy 0.7503). Fasting plasma glucose (FPG), education, exercise, gender, waist circumference (WC) top five important predictors. This showed that can be screen phrase, which has strong potential intelligent prevention control diabetes. key could also useful developing targeted interventions.

Language: Английский

---

An investigation of machine learning algorithms and data augmentation techniques for diabetes diagnosis using class imbalanced BRFSS dataset

Healthcare Analytics, Journal Year: 2023, Volume and Issue: 5, P. 100297 - 100297

Published: Dec. 30, 2023

Diabetes is a prevalent chronic condition that poses significant challenges to early diagnosis and identifying at-risk individuals. Machine learning plays crucial role in diabetes detection by leveraging its ability process large volumes of data identify complex patterns. However, imbalanced data, where the number diabetic cases substantially smaller than non-diabetic cases, complicates identification individuals with using machine algorithms. This study focuses on predicting whether person at risk diabetes, considering individual's health socio-economic conditions while mitigating posed data. We employ several augmentation techniques, such as oversampling (Synthetic Minority Over Sampling for Nominal Data, i.e.SMOTE-N), undersampling (Edited Nearest Neighbor, i.e. ENN), hybrid sampling techniques (SMOTE-Tomek SMOTE-ENN) training before applying algorithms minimize impact Our sheds light significance carefully utilizing without any leakage enhance effectiveness Moreover, it offers complete structure healthcare practitioners, from obtaining prediction, enabling them make informed decisions.

Language: Английский

---

Smart healthcare disease diagnosis and patient management: Innovation, improvement and skill development

Machine Learning with Applications, Journal Year: 2020, Volume and Issue: 3, P. 100011 - 100011

Published: Dec. 18, 2020

Data mining (DM) is an instrument of pattern detection and retrieval knowledge from a large quantity data. Many robust early services other health-related technologies have developed clinical diagnostic evidence in both the DM healthcare sectors. Artificial Intelligence (AI) commonly used research health care Classification or predictive analytics key part AI machine learning (ML). Present analyses new models founded on ML methods demonstrate promise area scientific research. Healthcare professionals need accurate predictions outcomes various illnesses that patients suffer from. In addition, timing another significant aspect affects choices for precise predictions. this regard, authors reviewed numerous publications terms method, algorithms, performance. This review paper summarized documentation examined accordance with approaches, styles, activities, processes. The assessment techniques selected papers are discussed appraisal findings presented to conclude article. statistical remedies been scientifically uncertainty between has now clarified. study related reveals prediction existing forecasting differs even if same dataset used. Predictive also essential, approaches be improved.

Language: Английский

---

Automated detection and classification of diabetes disease based on Bangladesh demographic and health survey data, 2011 using machine learning approach

Diabetes & Metabolic Syndrome Clinical Research & Reviews, Journal Year: 2020, Volume and Issue: 14(3), P. 217 - 219

Published: March 11, 2020

Language: Английский

---