A cascaded CNN-LSTM framework for quantifying respiratory motion from surface electromyographic signals.

Key Points

• This research aims to develop a framework for real-time monitoring of respiratory motion using deep learning techniques.
• Developed a cascaded convolutional neural network (CNN) and long short-term memory (LSTM) model.
• Implemented artifact suppression methods to enhance data quality.
• Tested the framework in clinical environments for respiratory monitoring.
• Achieved high-fidelity artifact suppression for cleaner signal processing.
• Enabled real-time monitoring of respiratory motion with improved accuracy in clinical settings.