Interference cancellation in respiratory sounds via a multiresolution joint time-delay and signal-estimation scheme

Area of Science:

• Biomedical Engineering
• Signal Processing

Background:

• Respiratory sound analysis is crucial for diagnosing lung conditions.
• Heart sounds often contaminate respiratory signals, hindering accurate interpretation.
• Existing methods for heart sound cancellation may lack precision or robustness.

Purpose of the Study:

• To develop and evaluate a new joint time-delay and signal-estimation (JTDSE) procedure for heart sound cancellation in respiratory sounds.
• To improve the accuracy and robustness of respiratory sound signal separation.

Main Methods:

• A multiresolution discrete wavelet transform (DWT) was used to decompose signals into subbands.
• Iterative time-delay estimation (TDE) was performed in each subband using Levenberg-Marquardt (LM) and block fast transversal filter (BFTF) algorithms.
• The JTDSE procedure incorporated complementary information across subbands and minimized errors between estimated signals.

Main Results:

• The proposed JTDSE methodology demonstrated robustness in noisy conditions and accuracy in time-delay estimation.
• Simulated and actual respiratory sound cases showed the effectiveness of the JTDSE scheme.
• Results indicated superior performance compared to standard adaptive filtering techniques.

Conclusions:

• The JTDSE procedure offers a promising approach for accurate time-delay estimation and effective heart sound cancellation.
• This method enhances the quality of respiratory sound recordings for improved clinical diagnostics.
• The technique's ability to integrate multi-subband information contributes to its robustness and accuracy.