Hybrid optimized temporal convolutional networks with long short-term memory for heart disease prediction with deep...

Area of Science:

• Cardiology
• Artificial Intelligence
• Medical Informatics

Background:

• Heart disease is a leading global cause of mortality.
• Early detection of heart conditions remains a significant challenge in healthcare.
• Accurate prediction models are crucial for timely intervention and improved patient survival rates.

Purpose of the Study:

• To develop and implement a novel heart disease prediction model using a hybrid deep learning strategy.
• To enhance the accuracy and efficiency of early heart disease detection.
• To provide a robust tool for medical professionals to aid in patient diagnosis.

Main Methods:

• A hybrid deep learning framework combining One-Dimensional Convolutional Neural Network (1DCNN) for feature extraction.
• Integration of Temporal Convolutional Networks (TCN) with Long Short-Term Memory (LSTM) for classification.
• Optimization of model parameters using the Enhanced Forensic-Based Investigation (EFBI) meta-optimization algorithm.

Main Results:

• The proposed model achieved a high accuracy rate of 98.67%.
• The precision rate of the developed heart disease prediction system was 99.48%.
• The system demonstrated superior performance compared to existing methods in evaluation metrics.

Conclusions:

• The novel hybrid deep learning approach offers a promising solution for accurate and early heart disease prediction.
• The developed model can significantly aid in the early diagnosis and management of heart conditions.
• This research contributes to advancing AI applications in cardiovascular disease detection.