Diabetes is gradually becoming a global challenge owing to the gradual increase in the number of cases of Type 2 diabetes mellitus (T2DM). T2DM is characterized as a state of hyperglycaemia due to abnormal control of insulin levels that eventually affects metabolism. This study aimed to review articles that implement machine learning methods to identify suitable risk factors for prediabetes.
The study adopted the preferred reporting items for systematic review (PRISMA) protocol and research questions were formulated by the identification of synonyms and related terms "predictors and prediabetes and machine learning" from PubMed and Google scholar. Both observational and interventional original articles that were published between 2018 and 2023 were included in this study. Eligibility for inclusion was determined by scanning the article title, abstract, and study methodology section.
Inhaltsverzeichnis (Table of Contents)
- 1 Introduction
- 1.1 Situation of Diabetes Mellitus
- 1.2 Introduction to Machine Learning
- 2 Materials and methods
- 2.1 Data Extraction and Analysis
- 3 Results
- 4 Conclusions
Zielsetzung und Themenschwerpunkte (Objectives and Key Themes)
This study aimed to review articles employing machine learning methods to identify suitable risk factors for prediabetes, a critical stage preceding Type 2 diabetes mellitus (T2DM). The review focuses on analyzing the effectiveness of various machine learning models in predicting prediabetes and identifying key risk factors from existing research.
- Prevalence and global burden of diabetes and prediabetes
- Application of machine learning in prediabetes risk prediction
- Analysis of machine learning model performance in predicting prediabetes
- Identification of key risk factors for prediabetes
- Need for generalized prediabetes screening tools across diverse populations
Zusammenfassung der Kapitel (Chapter Summaries)
1 Introduction: This chapter sets the stage by highlighting the escalating global health crisis posed by diabetes, particularly Type 2 diabetes mellitus (T2DM). It emphasizes the significant mortality rates associated with diabetes worldwide, with a focus on the disproportionate impact on developing countries and the projected rise in prevalence, especially in low-income nations due to lifestyle changes. The chapter also addresses the urgency of early diagnosis, particularly for Type 1 diabetes mellitus (T1DM), to prevent life-threatening complications. Specific regional data from Sub-Saharan Africa and Kenya illustrate the concerning trends and the need for improved diagnostic strategies. The chapter then introduces machine learning as a powerful tool for analyzing data and predicting outcomes, contrasting supervised and unsupervised learning methods and their applications in disease diagnosis and risk assessment. The importance of understanding prediabetes as a precursor to T2DM and its prevalence in various populations, including the USA, China, and Kenya, is stressed, emphasizing the gap in knowledge and the need for improved prediction tools.
2 Materials and methods: This chapter details the methodology employed in the systematic review, adhering to the PRISMA protocol. It explains the search strategy, including the databases used (PubMed and Google Scholar), search terms, inclusion/exclusion criteria (publication dates, language, study types), and the process of selecting relevant articles. A detailed flowchart visually represents the various stages of the literature search, highlighting the number of articles identified, screened, and ultimately included in the review. The chapter meticulously outlines the steps taken to ensure the accuracy and reliability of the selected studies. This rigorous approach is fundamental to the review's validity and provides transparency in the selection process.
3 Results: This section summarizes the findings from the four selected articles focusing on prediabetes prediction using machine learning. The studies were conducted in diverse geographical locations (Hong Kong, rural India, Korea, and the USA). The chapter details the machine learning algorithms utilized (support vector machine, extreme gradient boosting, and random forest) and their performance metrics (AUC-ROC and F1-Score), highlighting the generally good performance achieved across the reviewed studies. This section also discusses how prediabetes was defined in the included studies (primarily using American guidelines). Although the studies indicate promising results regarding machine learning's ability to predict prediabetes, the chapter emphasizes the need for further research to develop more universally applicable tools that account for racial and ethnic variations.
Schlüsselwörter (Keywords)
Prediabetes, Risk Prediction, Machine Learning, Decision Support Systems, Type 2 Diabetes Mellitus (T2DM), Global Health, Systematic Review, Algorithm Performance, Risk Factors
Frequently asked questions about the Language Preview
What is the Language Preview document about?
This document is a language preview containing the title, table of contents, objectives and key themes, chapter summaries, and keywords related to a study analyzing the use of machine learning for prediabetes risk prediction.
What is included in the Table of Contents?
The table of contents outlines the structure of the study, including sections on the Introduction (covering the situation of diabetes mellitus and an introduction to machine learning), Materials and Methods (data extraction and analysis), Results, and Conclusions.
What are the key objectives and themes of the study?
The study aims to review articles that use machine learning to identify risk factors for prediabetes. Key themes include the prevalence of diabetes and prediabetes, the application of machine learning for risk prediction, the performance of different models, the identification of key risk factors, and the need for generalized screening tools across diverse populations.
What does the Introduction chapter summarize?
The Introduction highlights the global health crisis of diabetes, focusing on T2DM and the urgency of early diagnosis, especially for T1DM. It also introduces machine learning as a valuable tool for analyzing data and predicting outcomes, emphasizing the significance of prediabetes as a precursor to T2DM and the need for better prediction tools.
What is described in the Materials and Methods chapter?
This chapter details the systematic review methodology used, following the PRISMA protocol. It covers the search strategy, including databases (PubMed and Google Scholar), search terms, inclusion/exclusion criteria, and the selection process for relevant articles.
What is the focus of the Results chapter?
The Results chapter summarizes the findings from four selected articles, conducted in various geographical locations, focusing on prediabetes prediction using machine learning algorithms like support vector machine, extreme gradient boosting, and random forest. It also mentions how prediabetes was defined and the need for more universally applicable tools.
What keywords are associated with the study?
The keywords include: Prediabetes, Risk Prediction, Machine Learning, Decision Support Systems, Type 2 Diabetes Mellitus (T2DM), Global Health, Systematic Review, Algorithm Performance, and Risk Factors.
Where were the studies within the results based?
Hong Kong, rural India, Korea, and the USA
- Quote paper
- Amos Olwendo (Author), 2025, A survey of Machine Learning Models for Prediabetes Screening, Munich, GRIN Verlag, https://www.hausarbeiten.de/document/1567635
