Article Type
Article
Abstract
Early prediction of gestational diabetes mellitus (GDM) risk is particularly important because it may enable more effective interventions and reduce cumulative maternal and fetal injury. The aim of this study is to develop machine learning (ML) models for early prediction of GDM using widely available variables, facilitating early intervention, and enabling the application of prediction models in settings where more complex tests are not accessible. In this paper, an artificial neural network (ANN) was used to detect the risk of gestational diabetes. A global dataset was used to collect clinical and experimental data from pregnant women and people with diabetes in Iraqi Kurdistan. The dataset used in this study includes records from 3,525 pregnancies. Twelve different deep learning models and their hyperparameters were optimized to achieve early and high prediction performance for GDM. The data augmentation method was used in training to improve the prediction results, as 70% of the data was trained and 30% was tested. Preliminary processing of this data was performed, and missing values were removed. Relevant features were identified, then the ANN model was used and trained with a suitable architecture using the training set. The model was evaluated using a healthy validation set to assess its performance in detecting pregnancy risk. Over-control was conducted to improve model performance. Finally, the improved ANN model was tested to evaluate its ability to predict the risk of gestational diabetes in early pregnancy. The results of this study provide insight into the effectiveness of using ANN for risk detection and contribute to the development of early pregnancy-specific strategies for diabetes, with the system's accuracy reaching 100%.
Keywords
Gestational Diabetes, Pregnancy, Artificial Neural Network (ANN), Deep Learning.
Recommended Citation
Dhahi, Sanaa Hammad
(2023)
"Risk Detection of Gestational Diabetes in Early Pregnancy Depend on Artificial Neural Network,"
Al-Esraa University College Journal for Engineering Sciences: Vol. 5:
Iss.
8, Article 4.
DOI: https://doi.org/10.70080/2790-7732.1045
Included in
Biomedical Engineering and Bioengineering Commons, Chemical Engineering Commons, Civil and Environmental Engineering Commons, Computer Engineering Commons, Materials Science and Engineering Commons, Mechanical Engineering Commons