Wavelet EOG Artifact Removal for Head Movement Noise
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems fail to effectively remove head movement artifacts from electrooculography (EOG) signals, which degrade signal quality and increase misclassification rates in eye movement detection, particularly in unconstrained environments without a chin rest.
Innovation Solution
A system utilizing a combination of filters and discrete wavelet transform to remove head movement artifacts from EOG signals, including a 4th order FIR bandpass filter, 1-dimensional median filter, polynomial fitting to remove DC drifts, and biorthogonal 'bior2.8' mother wavelet decomposition to filter noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional filtering techniques (band pass, median, moving average) are used to remove artifacts from EOG signals, then power line noise, facial EMG, and blink artifacts can be removed, but head movement artifacts cannot be removed because they are in the same frequency range and morphologically similar to EOG signals
Solution Approach 1:
The patent applies discrete wavelet transform to change the parameter domain from time-frequency to wavelet coefficient domain, enabling separation of head movement artifacts from EOG signals based on their different wavelet coefficient characteristics across multiple decomposition levels, rather than relying on frequency domain separation which fails for head movement artifacts
2Measurement precision
If constrained lab environments with chin rests are used to minimize head movement artifacts, then EOG signal quality improves, but the system becomes less adaptable to real-world unconstrained applications
Solution Approach 1:
The patent extracts and removes head movement artifacts from EOG signals using wavelet-based methods, enabling the system to maintain high measurement precision in unconstrained environments without requiring physical constraints like chin rests, thus improving adaptability to real-world applications
3Measurement precision
If head movement artifacts are present in EOG signals, then signal quality degrades and misclassification rate increases, but the artifacts cannot be removed by conventional filtering methods
Solution Approach 1:
The patent introduces wavelet transform as an intermediary method that bridges the gap between raw EOG signals contaminated with head movement artifacts and clean processed signals, enabling effective artifact removal through multi-level decomposition and selective coefficient thresholding
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Improves the accuracy of eye movement classification by reducing head movement artifacts, achieving higher classification accuracies and reducing computational load and signal reconstruction time.
Implementation Method 1
applying a discrete wavelet transform on the second set of filtered electrooculography (EOG) signals to filter a plurality of head movement noise
Data Source
Figure 1
Figure 2
Figure 3(a)~3(b)
AI summary
This disclosure relates generally to head movement noise removal from electrooculography (EOG) signals, and more particularly to systems and methods for wavelet based head movement artifact removal from electrooculography (EOG) signals. Embodiments of the present disclosure provide for head movement noise removal from the EOG signals by acquiring EOG signals of a user, filtering the acquired EOG signals to obtain a first set of filtered EOG signals, smoothening the first set of filtered EOG signals to obtain smoothened EOG signals, removing one or more redundant patterns and one or more direct current (DC) drifts from the smoothened EOG signals to obtain a second set of filtered EOG signals, and applying, a discrete wavelet transform on the second set of filtered EOG signals to filter a plurality of head movement noise from the second set of filtered EOG signals of the user.