Analysis of Therapeutic Effect of Elderly Patients with Severe Heart Failure Based on LSTM Neural Model

Area of Science:

• Cardiology
• Artificial Intelligence
• Biomedical Engineering

Background:

• Cardiovascular diseases are a leading cause of mortality globally.
• Early detection of heart failure and arrhythmias is crucial for improving patient outcomes and quality of life.
• Existing diagnostic methods may require timely enhancements for greater accuracy and accessibility.

Purpose of the Study:

• To develop and evaluate a novel deep learning network for the automatic detection of heart failure.
• To assess the efficacy of a combined long- and short-term memory network (LSTM) and deep residual neural network (ResNet) architecture.
• To investigate the model's performance on continuous heart rate data for clinical application.

Main Methods:

• A hybrid deep learning model integrating LSTM and ResNet was designed for heart failure detection.
• The model was trained and validated using continuous heart rate data from the PhysioNet database.
• Performance was evaluated using unbiased testing methods across different data lengths (500, 1000, 2000).

Main Results:

• The proposed deep learning model demonstrated high accuracy in detecting heart failure.
• Accuracy rates of 99.67% (500 data length), 98.84% (1000 data length), and 96.63% (2000 data length) were achieved.
• The model effectively extracts high-dimensional features from continuous heart rate data, enhancing classification accuracy.

Conclusions:

• The developed LSTM-ResNet model shows significant potential for accurate and automated heart failure detection.
• This approach can provide valuable heart health information for portable electrocardiogram (ECG) monitoring systems.
• The findings highlight the clinical value and social significance of advanced AI in cardiovascular disease management.