EEG Artifact Detection Using Pre-calculated Templates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting and removing artifacts from electrophysiological signals, such as EEG, are inadequate due to reliance on reference channels, computational intensity, and inability to operate in real-time, leading to data loss and inaccurate results.
Innovation Solution
A system and method for real-time detection and removal of artifacts using a combination of sensitivity and specificity algorithms, which analyze EEG signals with at least two separate measures to provide probabilities of artifact presence and absence, allowing for accurate identification and removal without compromising the underlying signal, and is designed for use in various physiological monitoring applications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If batch processing techniques are used for artifact identification, then artifact detection accuracy may be improved, but real-time detection capability is lost
Solution Approach 1:
The patent pre-calculates and stores artifact templates during system initialization or calibration phases. These templates represent characteristic artifact patterns that can be quickly matched against incoming signals during real-time operation, eliminating the need for computationally intensive batch processing while maintaining detection accuracy
Solution Approach 2:
The system dynamically adjusts detection thresholds and parameters based on the current signal characteristics and operational context. This allows the artifact detection algorithm to adapt its sensitivity and processing depth in real-time, optimizing the balance between detection accuracy and processing speed for different recording conditions
2Measurement precision
If computationally intensive techniques are used for artifact removal, then artifact detection accuracy may be improved, but processing time increases
Solution Approach 1:
The patent divides the continuous signal into discrete segments or epochs, applying artifact detection and removal algorithms to each segment independently. This segmentation allows for more efficient processing by limiting the computational scope to manageable time windows while maintaining overall signal integrity through proper segment reconstruction
Solution Approach 2:
The system employs lightweight, optimized algorithms that can be executed rapidly on standard hardware, replacing computationally intensive methods. These simplified algorithms use approximate matching and heuristic approaches that achieve sufficient accuracy with minimal processing overhead, enabling real-time operation without specialized computing resources
3Loss of information
If automated artifact detection is implemented, then data loss from artifact rejection is reduced, but false positive identification increases
Solution Approach 1:
The patent implements a feedback mechanism where detection results are continuously evaluated and used to adjust detection parameters. The system monitors false positive rates and automatically recalibrates thresholds and sensitivity settings based on observed performance, learning from accumulated data to improve discrimination between true artifacts and genuine signal features
Solution Approach 2:
The system applies multiple detection criteria and requires consensus among several independent indicators before declaring an artifact presence. This multi-criteria approach may occasionally detect some additional potential artifacts but significantly reduces false positives by requiring multiple lines of evidence, achieving a practical balance between sensitivity and specificity
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention relates to a physiological monitor and system, particularly to an electroencephalogram (EEG) monitor and system, and a method of detecting the presence and absence of artifacts and possibly removing artifacts from an EEG, other physiological signal or sensor signal without corrupting or compromising the signal. The accurate, real-time detection of the presence or absence of artifacts and removal of artifacts in EEG or other signals allows for increased reliability in the efficacy of those signals. The strategy of rejecting artifact-corrupted EEG can result in unacceptable data loss, and asking subjects to minimize movements in order to minimize artifacts is not always feasible. The present invention allows for increased accuracy in detection and removal of artifacts from physiological signals, substantially in real time, and without loss or corruption of signal or data in order to increase the accuracy of such signals for diagnosis and treatment purposes.