Harnessing the synergy of statistics and deep learning for BCI competition 4 dataset 4: a novel approach.

Gauttam Jangir, Nisheeth Joshi, Gaurav Purohit
Author Information
  1. Gauttam Jangir: Banasthali Vidyapith, Tonk, India. gauttamj49@gmail.com.
  2. Nisheeth Joshi: Banasthali Vidyapith, Tonk, India.
  3. Gaurav Purohit: Central Electronics Engineering Research Institute, Pilāni, India.

Abstract

Human brain signal processing and finger's movement coordination is a complex mechanism. In this mechanism finger's movement is mostly performed for every day's task. It is well known that to capture such movement EEG or ECoG signals are used. In this order to find the patterns from these signals is important. The BCI competition 4 dataset 4 is one such standard dataset of ECoG signals for individual finger movement provided by University of Washington, USA. In this work, this dataset is, statistically analyzed to understand the nature of data and outliers in it. Effectiveness of pre-processing algorithm is then visualized. The cleaned dataset has dual polarity and gaussian distribution nature which makes Tanh activation function suitable for the neural network BC4D4 model. BC4D4 uses Convolutional neural network for feature extraction, dense neural network for pattern identification and incorporating dropout & regularization making the proposed model more resilient. Our model outperforms the state of the art work on the dataset 4 achieving 0.85 correlation value that is 1.85X (Winner of BCI competition 4, 2012) & 1.25X (Finger Flex model, 2022).

Keywords

References

  1. Neuroimage. 2021 Nov 15;242:118463 [PMID: 34384910]
  2. Biosensors (Basel). 2022 Jun 02;12(6): [PMID: 35735532]
  3. J Neural Eng. 2022 Jan 24;19(1): [PMID: 35008079]
  4. Neuropsychologia. 2020 Sep;146:107506 [PMID: 32497532]
  5. IEEE Trans Biomed Eng. 2004 Jun;51(6):1034-43 [PMID: 15188875]
  6. Brain Topogr. 2015 Jan;28(1):47-61 [PMID: 25228153]
  7. J Neural Eng. 2007 Sep;4(3):264-75 [PMID: 17873429]
  8. J Mot Behav. 2022;54(2):135-145 [PMID: 34180775]
  9. Neurosci Lett. 2013 Jun 28;547:6-9 [PMID: 23665528]
  10. Nat Neurosci. 2002 Nov;5 Suppl:1085-8 [PMID: 12403992]
  11. J Neurosci Methods. 2010 Sep 30;192(1):152-62 [PMID: 20654646]
  12. Nat Rev Neurol. 2016 Sep;12(9):513-25 [PMID: 27539560]
  13. J Neural Eng. 2016 Jun;13(3):036014 [PMID: 27137818]
  14. Sci Rep. 2019 Apr 16;9(1):6115 [PMID: 30992474]
  15. J Neural Eng. 2020 Feb 25;17(1):016073 [PMID: 31972552]
  16. Front Neurosci. 2022 Oct 19;16:1009878 [PMID: 36340769]
  17. Nat Med. 2018 Nov;24(11):1669-1676 [PMID: 30250141]
  18. J Neurophysiol. 2022 Feb 1;127(2):559-570 [PMID: 35044809]
  19. Clin Neurophysiol. 2009 May;120(5):868-77 [PMID: 19345611]
  20. Sci Rep. 2020 Feb 7;10(1):2087 [PMID: 32034277]
  21. Front Neurosci. 2012 Jun 28;6:91 [PMID: 22754496]
  22. IEEE Trans Neural Syst Rehabil Eng. 2006 Jun;14(2):147-52 [PMID: 16792281]
  23. Int J Neural Syst. 2021 Sep;31(9):2150036 [PMID: 34247553]
  24. Nature. 2006 Jul 13;442(7099):195-8 [PMID: 16838020]
  25. Neuroimage. 2006 Feb 1;29(3):712-20 [PMID: 16188458]
  26. Brain Topogr. 2020 Jan;33(1):37-47 [PMID: 31879854]
  27. Neurol Int. 2022 Sep 19;14(3):738-747 [PMID: 36135997]
  28. J Neural Eng. 2009 Dec;6(6):066001 [PMID: 19794237]
  29. Annu Rev Biophys Bioeng. 1973;2:157-80 [PMID: 4583653]
  30. Sensors (Basel). 2021 Jan 12;21(2): [PMID: 33445601]
  31. J Neural Eng. 2022 Feb 25;19(1): [PMID: 35078156]

Word Cloud

Created with Highcharts 10.0.0dataset4movementBCImodelECoGsignalscompetitionneuralnetworkfinger'smechanismEEGworknatureBC4D4&1PotentialHumanbrainsignalprocessingcoordinationcomplexmostlyperformedeveryday'staskwellknowncaptureusedorderfindpatternsimportantonestandardindividualfingerprovidedUniversityWashingtonUSAstatisticallyanalyzedunderstanddataoutliersEffectivenesspre-processingalgorithmvisualizedcleaneddualpolaritygaussiandistributionmakesTanhactivationfunctionsuitableusesConvolutionalfeatureextractiondensepatternidentificationincorporatingdropoutregularizationmakingproposedresilientoutperformsstateartachieving085correlationvalue85XWinner201225XFingerFlex2022Harnessingsynergystatisticsdeeplearning4:novelapproachBrainComputerInterfaceelectroencephalogramElectrocorticographyEventRelatedERPMotorImageryMIVisualEvokedVEP

Similar Articles

Cited By

No available data.