Multidomain Convolution Neural Network Models for Improved Event-Related Potential Classification

Overview

Journal Sensors (Basel)

Publisher MDPI

Specialty Biotechnology

Date 2023 Jul 11

PMID 37430568

Authors

Xiaoqian Chen

Resh S Gupta

Lalit Gupta

Affiliations

Soon will be listed here.

Abstract

Two convolution neural network (CNN) models are introduced to accurately classify event-related potentials (ERPs) by fusing frequency, time, and spatial domain information acquired from the continuous wavelet transform (CWT) of the ERPs recorded from multiple spatially distributed channels. The multidomain models fuse the multichannel Z-scalograms and the V-scalograms, which are generated from the standard CWT scalogram by zeroing-out and by discarding the inaccurate artifact coefficients that are outside the cone of influence (COI), respectively. In the first multidomain model, the input to the CNN is generated by fusing the Z-scalograms of the multichannel ERPs into a frequency-time-spatial cuboid. The input to the CNN in the second multidomain model is formed by fusing the frequency-time vectors of the V-scalograms of the multichannel ERPs into a frequency-time-spatial matrix. Experiments are designed to demonstrate (a) customized classification of ERPs, where the multidomain models are trained and tested with the ERPs of individual subjects for brain-computer interface (BCI)-type applications, and (b) group-based ERP classification, where the models are trained on the ERPs from a group of subjects and tested on single subjects not included in the training set for applications such as brain disorder classification. Results show that both multidomain models yield high classification accuracies for single trials and small-average ERPs with a small subset of top-ranked channels, and the multidomain fusion models consistently outperform the best unichannel classifiers.

References

Choi H, Park J, Yang Y . A Novel Quick-Response Eigenface Analysis Scheme for Brain-Computer Interfaces. Sensors (Basel). 2022; 22(15). PMC: 9370919. DOI: 10.3390/s22155860. View

Cohen M . A better way to define and describe Morlet wavelets for time-frequency analysis. Neuroimage. 2019; 199:81-86. DOI: 10.1016/j.neuroimage.2019.05.048. View

Amerineni R, Gupta L, Steadman N, Annauth K, Burr C, Wilson S . Fusion Models for Generalized Classification of Multi-Axial Human Movement: Validation in Sport Performance. Sensors (Basel). 2021; 21(24). PMC: 8706912. DOI: 10.3390/s21248409. View

Wei X, Zhou L, Chen Z, Zhang L, Zhou Y . Automatic seizure detection using three-dimensional CNN based on multi-channel EEG. BMC Med Inform Decis Mak. 2018; 18(Suppl 5):111. PMC: 6284363. DOI: 10.1186/s12911-018-0693-8. View

Chen X, Gupta R, Gupta L . Exploiting the Cone of Influence for Improving the Performance of Wavelet Transform-Based Models for ERP/EEG Classification. Brain Sci. 2023; 13(1). PMC: 9856575. DOI: 10.3390/brainsci13010021. View

Mammone N, Ieracitano C, Morabito F . A deep CNN approach to decode motor preparation of upper limbs from time-frequency maps of EEG signals at source level. Neural Netw. 2020; 124:357-372. DOI: 10.1016/j.neunet.2020.01.027. View

Gupta L, Chung B, Srinath M, Molfese D, Kook H . Multichannel fusion models for the parametric classification of differential brain activity. IEEE Trans Biomed Eng. 2005; 52(11):1869-81. DOI: 10.1109/TBME.2005.856272. View

Dien J, Spencer K, Donchin E . Parsing the late positive complex: mental chronometry and the ERP components that inhabit the neighborhood of the P300. Psychophysiology. 2004; 41(5):665-78. DOI: 10.1111/j.1469-8986.2004.00193.x. View

Amerineni R, Gupta R, Gupta L . Multimodal Object Classification Models Inspired by Multisensory Integration in the Brain. Brain Sci. 2019; 9(1). PMC: 6356735. DOI: 10.3390/brainsci9010003. View

10.

Turk O, Ozerdem M . Epilepsy Detection by Using Scalogram Based Convolutional Neural Network from EEG Signals. Brain Sci. 2019; 9(5). PMC: 6562774. DOI: 10.3390/brainsci9050115. View

11.

Guttmann-Flury E, Sheng X, Zhu X . Channel selection from source localization: A review of four EEG-based brain-computer interfaces paradigms. Behav Res Methods. 2022; 55(4):1980-2003. DOI: 10.3758/s13428-022-01897-2. View

12.

Jochumsen M, Knoche H, Kjaer T, Dinesen B, Kidmose P . EEG Headset Evaluation for Detection of Single-Trial Movement Intention for Brain-Computer Interfaces. Sensors (Basel). 2020; 20(10). PMC: 7287803. DOI: 10.3390/s20102804. View

13.

Pham Xuan R, Andreessen L, Zander T . Investigating the Single Trial Detectability of Cognitive Face Processing by a Passive Brain-Computer Interface. Front Neuroergon. 2024; 2:754472. PMC: 10790859. DOI: 10.3389/fnrgo.2021.754472. View

14.

Gupta R, Kujawa A, Vago D . A Preliminary Investigation of ERP Components of Attentional Bias in Anxious Adults using Temporospatial Principal Component Analysis. J Psychophysiol. 2021; 35(4):223-236. PMC: 8559726. DOI: 10.1027/0269-8803/a000275. View

15.

Myrden A, Chau T . A Passive EEG-BCI for Single-Trial Detection of Changes in Mental State. IEEE Trans Neural Syst Rehabil Eng. 2017; 25(4):345-356. DOI: 10.1109/TNSRE.2016.2641956. View

16.

Lilly J . Element analysis: a wavelet-based method for analysing time-localized events in noisy time series. Proc Math Phys Eng Sci. 2017; 473(2200):20160776. PMC: 5415685. DOI: 10.1098/rspa.2016.0776. View

17.

Aunon J, McGillem C, Childers D . Signal processing in evoked potential research: averaging and modeling. Crit Rev Bioeng. 1981; 5(4):323-67. View

18.

Aslan Z, Akin M . A deep learning approach in automated detection of schizophrenia using scalogram images of EEG signals. Phys Eng Sci Med. 2021; 45(1):83-96. DOI: 10.1007/s13246-021-01083-2. View

19.

Gupta L, Molfese D, Tammana R, Simos P . Nonlinear alignment and averaging for estimating the evoked potential. IEEE Trans Biomed Eng. 1996; 43(4):348-56. DOI: 10.1109/10.486255. View