A Decision Support System for Diagnosing Diabetes Using Deep Neural Network.

Area of Science:

• Medical Informatics
• Machine Learning
• Computational Biology

Background:

• Diabetes mellitus is a global health concern and a leading cause of mortality.
• Traditional machine learning models struggle with unbalanced datasets for diabetes prediction.
• Previous artificial neural network approaches lacked sequence information understanding.

Purpose of the Study:

• To develop a deep learning-based decision support system (DSS) for accurate diabetes prediction.
• To utilize bidirectional long/short-term memory (BiLSTM) for enhanced feature extraction and prediction.
• To address the challenges of unbalanced datasets in diabetes prediction.

Main Methods:

• A hybrid model combining data balancing techniques with bidirectional long/short-term memory (BiLSTM) was employed.
• The BiLSTM model was utilized to process patient data for predicting diabetic illness.
• The model was trained and evaluated on a dataset to assess its predictive performance.

Main Results:

• The proposed BiLSTM model achieved a high accuracy of 93.07%.
• The model demonstrated strong performance with 93% precision, 92% recall, and 92% F1-score.
• The trial findings indicate significant improvements over previous methods.

Conclusions:

• The bidirectional long/short-term memory (BiLSTM) model offers superior performance for diabetes detection.
• This deep learning approach provides a promising tool for early and accurate diagnosis of diabetes.
• The study highlights the potential of advanced machine learning in managing chronic diseases.