[The method of instantaneous pulse detection based on hybrid wavelet transform]

Area of Science:

• Signal Processing
• Biomedical Engineering
• Mathematical Analysis

Context:

• Traditional signal detection methods often lack adaptability to transient signals.
• The relationship between matched filtering and wavelet transforms is explored, highlighting WT's potential as a flexible matched filter.
• Interference pulse detection in electroencephalogram (EEG) signals presents a significant challenge in neuroscience and clinical applications.

Purpose:

• To propose a novel signal detection method, Hybrid Wavelet Transform (HWT), based on the conceptual equivalence of Wavelet Transform (WT) and matched filtering.
• To leverage the adaptive nature of WT by employing two different mother wavelets during decomposition and reconstruction phases.
• To enhance the detection and characterization of transient interference pulses within complex biological signals like EEG.

Summary:

• This study introduces Hybrid Wavelet Transform (HWT), a signal detection technique conceptualized as a matched filter with a changeable template.
• HWT utilizes distinct mother wavelets for the decomposition and reconstruction stages: one for pulse detection (template) and another for characteristic enhancement.
• The method was successfully applied to detect interference pulses in EEG signals, demonstrating its efficacy for instantaneous signal detection.

Impact:

• HWT offers an improved approach for detecting transient signals, particularly interference pulses in noisy biological data.
• The findings suggest that HWT can enhance the accuracy and reliability of signal analysis in applications like EEG monitoring.
• This research provides a new perspective on WT as a versatile tool for adaptive signal detection, bridging signal processing theory and practical application.