EEG Signal Classification via SWFP Algorithm for RSVP Tasks

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current single-trial EEG classification methods face challenges in real-time applications due to high-dimensional data and small sample sizes, particularly in tasks like rapid serial visual presentation (RSVP), where noise and overlapping responses complicate the identification of task-related brain responses.

Innovation Solution

The Spatially Weighted Fisher Linear Discriminant-Principal Component Analysis (SWFP) algorithm, which involves a 2-step linear classification using Fisher Linear Discriminant (FLD) and PCA for dimensionality reduction, amplifies spatio-temporal representations and applies PCA projections to enhance classification accuracy in RSVP tasks.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If non-linear classifiers such as Neural Networks are used to capture complex discriminative functions, then classification accuracy can be improved, but computational complexity increases and real-time application becomes difficult

Engineering Contradiction:
Improveclassification accuracyVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent transforms the classification problem by changing the parameter space representation. It applies PCA to project high-dimensional EEG data onto a lower-dimensional space defined by principal components, then uses linear classifiers in this transformed space. This parameter transformation reduces computational complexity while maintaining classification accuracy, resolving the contradiction between accuracy and complexity.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent explicitly applies dimensionality reduction by projecting the original high-dimensional EEG data onto a lower-dimensional subspace spanned by principal components. This dimensional transformation simplifies the data structure and reduces computational requirements for classification while preserving the essential discriminative information, thus resolving the complexity-accuracy trade-off.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Device complexity

If feature extraction is performed to reduce dimensionality, then computational complexity is reduced, but information loss may occur

Engineering Contradiction:
Improvecomputational complexityVSAvoidinformation loss
Core Design Contradiction:
Device complexityVSLoss of information

Solution Approach 1:

The patent performs preliminary dimensionality reduction using PCA before classification. By pre-projecting the data onto principal components that capture the most variance in the data, it prepares a compressed representation that maintains essential information. This preliminary action reduces subsequent computational complexity while minimizing information loss through the preservation of dominant variance patterns.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent changes the representation parameters by transforming data from the original high-dimensional space to a lower-dimensional space of principal components. This parameter transformation is designed to preserve the most informative variance patterns, thereby reducing dimensionality without significant information loss, thus resolving the contradiction between complexity reduction and information preservation.

Inventive Principle:
Principle #35Parameter changes

3Reliability

If traditional EEG averaging over trials is used, then noise is reduced, but task-related brain responses cannot be identified in single trials for real-time feedback

Engineering Contradiction:
Improvesignal reliabilityVSAvoidreal-time processing capability
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent segments the EEG analysis from traditional trial-averaging approaches. Instead of averaging across trials to reduce noise, it applies dimensionality reduction and classification to individual single-trial data segments. This segmentation enables real-time processing of individual trials while maintaining reliability through the use of robust linear classifiers and principal component projection, thus resolving the contradiction between reliability and real-time capability.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent replaces the mechanical averaging process with a computational projection and classification system. Instead of relying on temporal averaging to reduce noise, it uses PCA projection onto principal components and linear classification in the transformed space. This substitution enables single-trial analysis with maintained reliability, achieving real-time processing capability.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

4Speed

If high-dimensional EEG data is processed, then temporal resolution is maintained, but classification becomes difficult due to high dimensional space and small sample sizes

Engineering Contradiction:
Improvetemporal resolutionVSAvoidclassification difficulty
Core Design Contradiction:
SpeedVSDifficulty of detecting and measuring

Solution Approach 1:

The patent applies dimensionality reduction by projecting high-dimensional EEG data onto a lower-dimensional space of principal components. This dimensional transformation simplifies the classification problem by reducing the number of features while preserving the temporal structure and discriminative information. The principal components are ordered by variance, ensuring that the most informative dimensions are retained, thus resolving the difficulty of classification in high-dimensional space.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The patent changes the data representation parameters by transforming the original high-dimensional EEG data into a reduced-dimensional space spanned by principal components. This parameter transformation makes the data more manageable for classification algorithms, reducing the curse of dimensionality while maintaining temporal resolution. The transformation preserves the essential temporal dynamics needed for classification, resolving the contradiction between temporal resolution and classification difficulty.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP2986203B1Classifying EEG signals in response to visual stimulus
Publication Date: 2022.12.07 YISSUM RESEARCH DEVELOPMENT COMPANY OF THE HEBREW UNIVERSITY OF JERUSALEM LTD
  • EP2986203B1 patent drawingFigure 1~2
  • EP2986203B1 patent drawingFigure 3
  • EP2986203B1 patent drawingFigure 4

AI summary

Systems and method for conduction of single trial classification of EEG signals of a human subject generated responsive to a series of images containing target images and non-target images, the method comprising: obtaining said EEG signals in a spatio-temporal representation comprising time points and respective spatial distribution of said EEG signals; classifying said time points independently, using a linear discriminant classifier, to compute spatio-temporal discriminating weights; using said spatio-temporal discriminating weights to amplify said spatio-temporal representation by said spatio-temporal discriminating weights at tempo-spatial points respectively, to create a spatially-weighted representation; using Principal Component Analysis (PCA) on a temporal domain for dimensionality reduction, separately for each spatial channel of said EEG signals, to create a PCA projection; applying said PCA projection to said spatially-weighted representation onto a first plurality of principal components, to create a temporally approximated spatially weighted representation containing for each spatial channel, PCA coefficients for said plurality of principal temporal projections; and classifying said temporally approximated spatially weighted representation, over said number of channels, using said linear discriminant classifier, to yield a binary decisions series indicative of each image of the images series as either belonging to said target image or to said non-target image.