EEG Artifact Correction Using Adaptive Template Subtraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for removing EEG artifacts during fMRI scanning are ineffective when the patient's head moves, as they rely on averaging techniques that assume static head positions, leading to deteriorated results.
Innovation Solution
A method and system that segment EEG data based on MR scan time periods, detect movement using an acceleration sensor, and adaptively update templates for artifact subtraction, allowing for reliable artifact removal even with head movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the average subtraction method is used to remove artifacts, then artifact removal is effective for static head positions, but the method deteriorates when head movement occurs
Solution Approach 1:
The patent applies dynamics by making the template adaptive rather than static. The template is continuously updated using a sliding window approach that incorporates recent data segments, allowing it to track changes in artifact characteristics caused by head movement. This dynamic adaptation resolves the contradiction by maintaining measurement precision across varying head positions.
Solution Approach 2:
The patent uses preliminary action by pre-processing EEG data to detect head movement before artifact removal. A head movement detection mechanism identifies when the subject moves, and only segments without movement are used to construct the template. This preliminary detection ensures that the template reflects stable conditions, improving artifact removal accuracy even when subsequent movement occurs.
2Quantity of substance
If EEG data segments with head movement are included in template construction, then more data is available for averaging, but the template accuracy deteriorates
Solution Approach 1:
The patent applies local quality by differentiating between good-quality segments (without head movement) and poor-quality segments (with head movement). Only segments meeting quality criteria are included in template construction, ensuring that each local contribution to the template is accurate. This selective inclusion maintains template accuracy while still utilizing sufficient data segments.
Solution Approach 2:
The patent implements feedback through a quality assessment mechanism that evaluates each EEG segment before inclusion in the template. Head movement detection provides feedback about segment quality, and this information is used to selectively include or exclude segments from template construction. This feedback loop ensures that only accurate segments contribute to the template, maintaining precision while accumulating sufficient data.
Data Source
AI summary
Artifacts are removed from EEG signals by segmenting EEG data that are continuously recorded during an MR scan with respect to a time period based on the MR scan, thereby obtaining n temporally consecutive segments of EEG data. For each segment j of the n segments, it is determined whether movement is detected when the segment j is recorded. In the case of no movement, the segment j is selected for a template k. In case movement is detected, EEG data of segments of the n temporally consecutive segments which have been selected for the template k are averaged, thereby obtaining the template k, the template k is subtracted from the EEG data of the segments, k is incremented, and the segment j is selected for the template k. In case no movement is detected and j=n, the segment j is selected for the template k and EEG data of segments of the n temporally consecutive segments which have been selected for the template k are averaged, thereby obtaining the template k, and the template k is subtracted from the EEG data of the segments.