sEMG Signal Time-Localization via Dynamic Wavelet Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The Discrete Wavelet Transform (DWT) struggles with time-localization and characterization of spiking events in surface electromyographic (sEMG) signals due to periodicity and variability in burst return times, leading to misalignment of motor unit action potentials and noise interference.
Innovation Solution
A system that includes sensors and a processor to receive and process sEMG signals using a Daubechies 3 wavelet decomposition, applying time delays to achieve optimal waveform matching between raw and transformed signals through expert rules, minimizing errors and enhancing time-localization of specific bursts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If Discrete Wavelet Transform is used to analyze sEMG signals, then frequency decomposition is achieved, but time-localization of spiking events deteriorates due to periodicity and variability in burst return times
Solution Approach 1:
The patent applies dynamic time-warping algorithms that adaptively adjust the time-scaling of wavelet coefficients based on the actual burst return times observed in the signal. This dynamic adjustment allows the transform to follow the variable timing patterns of motor unit bursts, maintaining both frequency decomposition accuracy and time-localization precision despite the non-stationary nature of sEMG signals.
Solution Approach 2:
The invention changes the parameter of wavelet transform by using multiple scales and selectively emphasizing different time-frequency resolutions based on the local characteristics of the signal. By adapting the wavelet scale parameters to match the burst patterns, the system achieves accurate time-localization while preserving frequency information.
2Productivity
If standard DWT is applied to sEMG signals, then signal decomposition is achieved, but waveform matching between raw and transformed signals deteriorates
Solution Approach 1:
The patent implements a feedback mechanism where the transformed wavelet coefficients are compared with the original raw signal, and the time-delay parameter is iteratively adjusted to maximize the correlation between them. This feedback loop ensures optimal waveform matching by finding the time-shift that best aligns the decomposed components with the original signal structure.
Solution Approach 2:
The system performs preliminary time-delay estimation on the wavelet coefficients before final reconstruction, pre-aligning the transformed signal components with the original signal timing. This preliminary action prevents timing misalignment from propagating through the reconstruction process, ensuring accurate waveform matching.
3Difficulty of detecting and measuring
If multiple MUAPs are aligned with mother wavelet, then spiking events are detected, but misalignment occurs due to periodicity and burst variability
Solution Approach 1:
The patent segments the continuous sEMG signal into individual burst intervals based on detected envelope peaks, then applies separate time-localization analysis to each segment. This segmentation allows the system to handle the variability of each burst independently, improving alignment precision by adapting to local timing patterns rather than assuming uniform periodicity across the entire signal.
Data Source
AI summary
Methods, systems, and apparatus for detecting rhythmic synchronization of motor neurons. The system includes one or more sensors for receiving a first signal that measures an electrical signal of one or more neurons, the first signal having a first plurality of specific bursts. The system includes a processor connected to the one or more sensors. The processor is configured to receive the first signal. The processor is configured to generate a second signal based on the first signal using a discrete wavelet transform, the second signal having a second plurality of specific bursts. The processor is configured to determine a time delay between a specific burst within the first plurality of specific bursts and a specific burst within the second plurality of specific bursts using one or more expert rules. The processor is configured to apply the time delay to the first signal.