AI Article Synopsis

  • This study focuses on early prediction of Gestational Diabetes Mellitus (GDM) risk using machine learning (ML) models, which can help in effective interventions for both mother and fetus, especially in areas without advanced medical testing.
  • A dataset of 1,611 pregnancies was analyzed, optimizing twelve ML models to enhance prediction accuracy, with key variable selection methods implemented for better results.
  • The top models demonstrated high sensitivity and specificity for GDM prediction, achieving a balance of accuracy and the number of input variables needed, indicating potential for use in regular prenatal care.

Article Abstract

Background: Early prediction of Gestational Diabetes Mellitus (GDM) risk is of particular importance as it may enable more efficacious interventions and reduce cumulative injury to mother and fetus. The aim of this study is to develop machine learning (ML) models, for the early prediction of GDM using widely available variables, facilitating early intervention, and making possible to apply the prediction models in places where there is no access to more complex examinations.

Methods: The dataset used in this study includes registries from 1,611 pregnancies. Twelve different ML models and their hyperparameters were optimized to achieve early and high prediction performance of GDM. A data augmentation method was used in training to improve prediction results. Three methods were used to select the most relevant variables for GDM prediction. After training, the models ranked with the highest Area under the Receiver Operating Characteristic Curve (AUCROC), were assessed on the validation set. Models with the best results were assessed in the test set as a measure of generalization performance.

Results: Our method allows identifying many possible models for various levels of sensitivity and specificity. Four models achieved a high sensitivity of 0.82, a specificity in the range 0.72-0.74, accuracy between 0.73-0.75, and AUCROC of 0.81. These models required between 7 and 12 input variables. Another possible choice could be a model with sensitivity of 0.89 that requires just 5 variables reaching an accuracy of 0.65, a specificity of 0.62, and AUCROC of 0.82.

Conclusions: The principal findings of our study are: Early prediction of GDM within early stages of pregnancy using regular examinations/exams; the development and optimization of twelve different ML models and their hyperparameters to achieve the highest prediction performance; a novel data augmentation method is proposed to allow reaching excellent GDM prediction results with various models.

Download full-text PDF

Source
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC10288662PMC
http://dx.doi.org/10.1186/s12884-023-05766-4DOI Listing

Publication Analysis

Top Keywords

early prediction
12
models
11
prediction
9
machine learning
8
learning models
8
gestational diabetes
8
prediction gdm
8
prediction models
8
twelve models
8
models hyperparameters
8

Similar Publications

Background: The triglyceride glucose-body mass index (TyG-BMI) is considered to be a reliable surrogate marker of insulin resistance (IR). However, limited evidence exists regarding its association with the severity of coronary artery disease (CAD), particularly in hypertensive patients with different glucose metabolic states, including those with H-type hypertension. This study aimed to investigate the relationship between TyG-BMI and CAD severity across different glucose metabolism conditions.

View Article and Find Full Text PDF

Background: Artcure diffusional patch (ADP) is a novel transdermal therapeutic system that started to be used in the last decade for lumbar disc herniation (LDH). Previous studies have reported early results of the therapy. In this study, we aimed to evaluate the medium- to long-term functional outcomes of this treatment in LDH patients and examine factors predicting the need for surgery after treatment.

View Article and Find Full Text PDF

Interpreting positive celiac serology in children with new-onset type 1 diabetes.

J Pediatr Endocrinol Metab

January 2025

Division of Gastroenterology, Hepatology, & Nutrition, The Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, USA.

Objectives: The association of celiac disease (CD) in type 1 diabetes mellitus (T1DM) is well-established, yet variation exists in screening practices. This study measures the accuracy of early screening with tissue transglutaminase Immunoglobulin A (TTG-IgA) and endomysial antibody (EMA) in newly diagnosed T1DM.

Methods: This is a retrospective study of children with T1DM between 2013 and 2019 with early CD screening and follow-up.

View Article and Find Full Text PDF

Background: Ultrasonographic optic nerve sheath diameter (ONSD) is a satisfactory noninvasive intracranial pressure (ICP) monitoring test. Our aim was to evaluate ONSD as an objective screening tool to predict and diagnose ICP changes early in sepsis-associated encephalopathy (SAE).

Methods: Our prospective observational study was conducted on patients with sepsis, and after intensive care unit (ICU) admission, the time to diagnose SAE was recorded, and patients were divided into a non-SAE group including conscious patients with sepsis and a SAE group including patients with sepsis with acute onset of disturbed conscious level.

View Article and Find Full Text PDF

Transplant-associated thrombotic microangiopathy (TA-TMA) is an increasingly recognized complication in hematopoietic cell transplantation (HCT). Given the rarity of prospective pediatric studies on TA-TMA, this study aimed to evaluate the incidence, survival outcomes, and risk factors for predicting early the development of TA-TMA in a pediatric population following allogeneic HCT. We conducted a prospective analysis of 173 pediatric patients to evaluate the incidence, survival outcome, and risk factors of TA-TMA.

View Article and Find Full Text PDF

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!