A Characteristic Filtering Method for Pulse Wave Signal Quality Assessment.

Area of Science:

• Biomedical Engineering
• Physiological Signal Processing
• Medical Data Analysis

Background:

• Pulse wave recordings are crucial physiological signals in clinical practice.
• Common filtering methods often fail to effectively remove interferences in pulse wave data.
• Noisy outliers can significantly compromise the accuracy of pulse wave analysis.

Purpose of the Study:

• To develop and validate a novel filtering method for removing noisy outliers from pulse wave recordings.
• To enhance the reliability of pulse wave data for clinical applications.
• To provide a robust preprocessing technique for physiological signals.

Main Methods:

• Selected five characteristic metrics: short-term energy (SE), ascending intensity difference (AID), descending intensity difference (DID), ascending time difference (ATD), and descending time difference (DTD).
• Calculated these metrics from cardiac pulse wave data.
• Utilized median filters to obtain median lines for each metric.
• Defined acceptable value ranges around median lines using histogram analysis.
• Identified and discarded cardiac cycles where characteristic values fell outside the acceptable ranges.

Main Results:

• The proposed method efficiently identified and removed noisy outlier segments from pulse wave recordings.
• The characteristic-based filtering approach proved effective in handling difficult-to-eliminate noise.
• The method demonstrated potential for improving the performance of pulse-wave-based clinical assessments.

Conclusions:

• The developed filtering method offers an effective solution for preprocessing noisy pulse wave data.
• This technique can enhance the accuracy and reliability of clinical applications relying on pulse wave analysis.
• The methodology shows promise for extension to other physiological signal preprocessing, such as ECG and blood pressure waves.