[The application of artificial neural network based on ambiguous function to heart sound classification]

Area of Science:

• Biomedical Engineering
• Artificial Intelligence in Medicine
• Signal Processing

Background:

• Heart sound analysis is crucial for diagnosing cardiovascular conditions.
• Traditional methods face challenges with signal ambiguity and computational complexity.
• Developing robust and efficient heart sound classification techniques is essential.

Purpose of the Study:

• To apply artificial neural networks (ANNs) for heart sound classification.
• To circumvent the crossterm problem associated with Wigner distribution by using ambiguous function inputs.
• To develop and optimize an ANN model for accurate heart sound analysis.

Main Methods:

• Utilized artificial neural networks for heart sound signal processing.
• Employed ambiguous functions as input data to avoid Wigner distribution issues.
• Iteratively trained and validated a suboptimal network model to determine key performance factors.

Main Results:

• The application of ANNs proved to be an effective technique for heart sound classification.
• The proposed method successfully avoided the crossterm problem inherent in Wigner distribution.
• An optimized network model was developed through repeated training and validation cycles.

Conclusions:

• Artificial neural networks offer a promising and effective approach to heart sound classification.
• The input data strategy effectively mitigates signal processing challenges.
• Further discussion on factors influencing network performance provides insights for future research.