Diabetes prediction using machine learning and explainable AI techniques

Healthcare Technology Letters, Год журнала: 2022, Номер 10(1-2), С. 1 - 10

Опубликована: Дек. 14, 2022

Globally, diabetes affects 537 million people, making it the deadliest and most common non-communicable disease. Many factors can cause a person to get affected by diabetes, like excessive body weight, abnormal cholesterol level, family history, physical inactivity, bad food habit etc. Increased urination is one of symptoms this People with for long time several complications heart disorder, kidney disease, nerve damage, diabetic retinopathy But its risk be reduced if predicted early. In paper, an automatic prediction system has been developed using private dataset female patients in Bangladesh various machine learning techniques. The authors used Pima Indian collected additional samples from 203 individuals local textile factory Bangladesh. Feature selection algorithm mutual information applied work. A semi-supervised model extreme gradient boosting utilized predict insulin features dataset. SMOTE ADASYN approaches have employed manage class imbalance problem. classification methods, that is, decision tree, SVM, Random Forest, Logistic Regression, KNN, ensemble techniques, determine which produces best results. After training on testing all models, proposed provided result XGBoost classifier approach 81% accuracy, 0.81 F1 coefficient AUC 0.84. Furthermore, domain adaptation method implemented demonstrate versatility system. explainable AI LIME SHAP frameworks understand how predicts final Finally, website framework Android smartphone application input instantaneously. Bangladeshi programming codes are available at following link: https://github.com/tansin-nabil/Diabetes-Prediction-Using-Machine-Learning.

Язык: Английский

---

Advancements and Prospects of Machine Learning in Medical Diagnostics: Unveiling the Future of Diagnostic Precision

Archives of Computational Methods in Engineering, Год журнала: 2024, Номер unknown

Опубликована: Июнь 26, 2024

Язык: Английский

---

Using artificial intelligence to rapidly identify microplastics pollution and predict microplastics environmental behaviors

Journal of Hazardous Materials, Год журнала: 2024, Номер 474, С. 134865 - 134865

Опубликована: Июнь 12, 2024

Язык: Английский

---

Synergistic Feature Selection and Distributed Classification Framework for High-Dimensional Medical Data Analysis

MethodsX, Год журнала: 2025, Номер 14, С. 103219 - 103219

Опубликована: Фев. 13, 2025

Feature selection and classification efficiency accuracy are key to improving decision-making regarding medical data analysis. Since the datasets large complex, they give rise certain problematic issues such as computational complexity, limited memory space, a lesser number of correct classifications. In order overcome these drawbacks, new integrated algorithm is presented here: Synergistic Kruskal-RFE Selector Distributed Multi-Kernel Classification Framework (SKR-DMKCF). The innovative architecture SKR-DMKCF results in reduction dimensionality while preserving useful characteristics image utilizing recursive feature elimination multi-kernel distributed environment. Detailed evaluations were performed on four broad established our performance advantage. average ratio was 89 % for proposed method, SKR-DMKCF, which can outperform all methods by achieving best 85.3 %, precision 81.5 recall 84.7 %. On calculations, it seen that usage 25 compared existing speed-up time significant improvement well assure scalability resource-limited environments.•Innovative efficient datasets.•Distributed superior efficiency.

Язык: Английский

---

Prediction of compressive and flexural strength of coal gangue-based geopolymer using machine learning method

Materials Today Communications, Год журнала: 2025, Номер 44, С. 112076 - 112076

Опубликована: Март 1, 2025

Язык: Английский

---

A review of radiomics and genomics applications in cancers: the way towards precision medicine

Radiation Oncology, Год журнала: 2022, Номер 17(1)

Опубликована: Дек. 30, 2022

Abstract The application of radiogenomics in oncology has great prospects precision medicine. Radiogenomics combines large volumes radiomic features from medical digital images, genetic data high-throughput sequencing, and clinical-epidemiological into mathematical modelling. amalgamation radiomics genomics provides an approach to better study the molecular mechanism tumour pathogenesis, as well new evidence-supporting strategies identify characteristics cancer patients, make clinical decisions by predicting prognosis, improve development individualized treatment guidance. In this review, we summarized recent research on applications solid cancers presented challenges impeding adoption practice. More standard guidelines are required normalize reproducible convincible analyses develop it a mature field.

Язык: Английский

---

Application of Machine Learning for Cardiovascular Disease Risk Prediction

Computational Intelligence and Neuroscience, Год журнала: 2023, Номер 2023(1)

Опубликована: Янв. 1, 2023

Cardiovascular diseases (CVDs) are a common cause of heart failure globally. The need to explore possible ways tackle the disease necessitated this study. study designed machine learning model for cardiovascular risk prediction in accordance with dataset that contains 11 features which may be used forecast disease. from Kaggle on includes approximately 70,000 patient records were determine outcome. Compared UCI dataset, has many more training and validation records. Models created using neural networks, random forests, Bayesian C5.0, QUEST compared dataset. On testing data sets, results acquired high accuracy (99.1 percent), is significantly superior previous methods. Ahead‐of‐time detection diagnosis cardiac disease, as well better treatment outcomes, strong possibilities suggested model. Additionally, it help patients manage their illness or life forms order increase chances recovery/survival. result showed greater promising signs machine‐learning algorithms can indeed assist early identification improvement

Язык: Английский

---

A Diabetes Monitoring System and Health-Medical Service Composition Model in Cloud Environment

IEEE Access, Год журнала: 2023, Номер 11, С. 32804 - 32819

Опубликована: Янв. 1, 2023

Diabetes is a common chronic illness or absence of sugar in the blood. The early detection this disease decreases serious risk factor. Nowadays, Machine Learning based cloud environment acts as vital role detection. people who belong to rural areas are not getting proper health care treatments. So, research work proposed an automated eHealth system for detecting diabetes earlier stage decrease mortality rate and provides treatment facilities peoples. Extreme (ELM) type Artificial Neural Network (ANN) that has lot potential solving classification challenges. This consisting several activities like feature normalization, selection classification. We have employed principal component analysis (PCA) extreme learning machine Finally, computing-based with three numbers virtual machines (vCPU-4, vCPU-8, vCPU-16), used diabetes. efficacy model been evaluated PIMA dataset both standalone environments achieved 90.57 % accuracy, 82.24 sensitivity, 73.23 specificity, 75.03 F-1 score vCPU-16. experimental results define superior other state-of-art models better accuracy less number features.

Язык: Английский

---

A Comparison of Machine Learning Techniques for the Detection of Type-2 Diabetes Mellitus: Experiences from Bangladesh

Information, Год журнала: 2023, Номер 14(7), С. 376 - 376

Опубликована: Июль 2, 2023

Diabetes is a chronic disease caused by persistently high blood sugar level, causing other diseases, including cardiovascular, kidney, eye, and nerve damage. Prompt detection plays vital role in reducing the risk severity associated with diabetes, identifying key factors can help individuals become more mindful of their lifestyles. In this study, we conducted questionnaire-based survey utilizing standard diabetes variables to examine prevalence Bangladesh. To enable prompt compared different machine learning techniques proposed an ensemble-based framework that incorporated algorithms such as decision tree, random forest, extreme gradient boost algorithms. order address class imbalance within dataset, initially applied synthetic minority oversampling technique (SMOTE) (ROS) techniques. We evaluated performance various classifiers, tree (DT), logistic regression (LR), support vector (SVM), (GB), (XGBoost), forest (RF), ensemble (ET), on our datasets. Our experimental results showed ET outperformed classifiers; further enhance its effectiveness, fine-tuned hyperparameters ET. Using statistical techniques, also ranked features identified age, thirst, family are significant prove instrumental patients. This method has great potential for clinicians effectively identify at facilitating timely intervention care.

Язык: Английский

---

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

Preventive Medicine Reports, Год журнала: 2023, Номер 35, С. 102358 - 102358

Опубликована: Авг. 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.

Язык: Английский

---