Leveraging a 7-Layer Long Short-Term Memory Model for Early Detection and Prevention of Diabetes in Oman: An...

Area of Science:

• Artificial Intelligence in Healthcare
• Medical Informatics
• Public Health Technology

Background:

• Rising prevalence of Type 2 diabetes in Oman, projected to affect 23.8% by 2050.
• Increasing impact of factors like obesity and genetic predispositions on diabetes rates.
• Need for enhanced public health awareness and preventative measures for diabetes management.

Purpose of the Study:

• To develop and evaluate a 7-layer Long Short-Term Memory (LSTM) model for early diabetes detection in Oman.
• To leverage Artificial Intelligence (AI) in healthcare to address the growing diabetes epidemic.
• To improve diagnostic capabilities and support preventative strategies within Oman's healthcare system.

Main Methods:

• Development of a 7-layer Long Short-Term Memory (LSTM) neural network architecture.
• Utilizing LSTM networks to analyze factors contributing to diabetes, such as obesity and genetics.
• Performance evaluation using key metrics including accuracy, sensitivity, specificity, recall, F1 score, and AUC.

Main Results:

• The LSTM model achieved a high accuracy of 99.40% in diabetes detection.
• Demonstrated 100% sensitivity and specificity for positive cases, and 99.34% recall for negative cases.
• Attained an F1 score of 96.24% and an AUC score of 94.51%, indicating robust diagnostic capability.

Conclusions:

• The developed LSTM model shows significant potential for enhancing early diabetes detection in Oman.
• Implementation of this AI-driven approach can improve public health outcomes and diabetes prevention.
• This study represents a significant advancement in applying AI to address critical healthcare challenges in Oman and globally.