Brain Topography

, Volume 4, Issue 4, pp 291–307 | Cite as

Unrestricted principal components analysis of brain electrical activity: Issues of data dimensionality, artifact, and utility

  • Frank H. Duffy
  • Kenneth Jones
  • Peter Bartels
  • Gloria McAnulty
  • Marilyn Albert


Principal components analysis (PCA) was performed on the 1536 spectral and 2944 evoked potential (EP) variables generated by neurophysiologic paradigms including flash VER, click AER, and eyes open and closed spectral EEG from 202 healthy subjects aged 30 to 80. In each case data dimensionality of 1500 to 3000 was substantially reduced using PCA by magnitudes of 20 to over 200. Just 20 PCA factors accounted for 70% to 85% of the variance. Visual inspection of the topographic distribution of factor loading scores revealed complex loadings across multiple data dimensions (time-space and frequency-space). Forty-two non-artifactual factors were successful in classifying age, gender, and a separate group of 60 demented patients by linear discriminant analysis. Discrimination of age and gender primarily involved EP derived factors, whereas dementia primarily involved EEG derived factors. Thirty-eight artifactual factors were identified which, alone, could not discriminate age but were relatively successful in discriminating gender and dementia. The need to parsimoniously develop real neurophysiologic measures and to objectively exclude artifact are discussed. Unrestricted PCA is suggested as a step in this direction.

Key words

Spectral analysis Evoked potentials Principal components analysis Singular value decomposition Discriminant function analysis Artifact Dimensionality Aging Dementia 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Abt, K. Descriptive data analyses: A concept between confirmatory and exploratory data analysis. Methods of Information in Medicine, 1987, 26: 77–88.Google Scholar
  2. Abt, K.. Descriptive data analysis (DDA) in Quantitative EEG studies. In D. Samson-Dollfus,J. D. Guieu,J. Gotman, and P. Etvènon (Eds.), Statistics and Topography in Quantitative EEG (pp. 150–160). Amsterdam: Elsevier, 1988.Google Scholar
  3. Achim, A., Richer, F., Alain, C., and Saint-Hilaire, J. M. A test of model adequacy applied to the dimensionality of multichannel average auditory evoked potentials. In D. Samson-Dollfus,J. D. Guieu,J. Gotman, and P. Etvenon (Eds.), Statistics and Topography in Quantitative EEG (pp. 161–171). Amsterdam: Elsevier, 1988.Google Scholar
  4. Bartels, P. H. Numerical evaluation of cytologic data IX. Search for data structure by principal components transformation. Anal. Quant. Cytol., 1981a, 3: 167–177.Google Scholar
  5. Bartels, P. H. Numerical evaluation of cytologic data VIII. Computation of the principal components. Anal. Quant. Cytol., 1981b, 3: 83–90.Google Scholar
  6. Bartlett, M. S. Tests of significance in factor analysis. Brit. J. of Stat. Psychol., 1950, 3: 77–85.Google Scholar
  7. Bickford, R. G., Cody, D. T., Jacobsen, J. L., and Lambert, E. H. Fast motor systems in man: Physiopathology of the sonomotor response. Trans. Am. Neurol. Assoc., 1964, 89: 56–58.Google Scholar
  8. Cooley, W. W., and Lohnes, P. R. Multivariate Data Analysis. New York: J. Wiley and Sons, 1971.Google Scholar
  9. Donchin, E. A multivariate approach to the analysis of average evoked potentials. IEEE Trans. Biomed. Eng., 1966, BME-13: 131–139.Google Scholar
  10. Duffy, F. H. Topographic display of evoked potentials: Clinical applications of brain electrical activity mapping (BEAM). Ann. N.Y. Acad. Sci., 1982, 388: 183–196.Google Scholar
  11. Duffy, F. H. Clinical decision making in quantified electroencephalographic analysis. In D. Samson-Dollfus,J. Gotman,J. D. Guieu, and P. Etevenon (Eds.), Statistics and Topography in Quantitative EEG (pp. 9–26). Paris: Elsevier, 1988.Google Scholar
  12. Duffy, F. H.. Topographic mapping of brain electrical activity: Clinical applications and issues. In K. Maurer (Eds.), Topographic Brain Mapping of EEG and Evoked Potentials (pp. 19–52). Heidelberg, Germany: Springer-Verlag, 1989.Google Scholar
  13. Duffy, F. H., Albert, M. S., and McAnulty, G. Brain electrical activity in patients with presenile and senile dementia of the Alzheimer's type. Ann. Neurol., 1984a, 16: 439–448.Google Scholar
  14. Duffy, F. H., Albert, M. S., McAnulty, G., and Garvey, A. J. Age- related differences in brain electrical activity mapping of healthy subjects. Ann. Neurol., 1984b, 16: 430–438.Google Scholar
  15. Duffy, F. H., Bartels, P. H., and Neff, R. N. A response to Oken and Chiappa. Ann. Neurol., 1986, 19: 494–495.Google Scholar
  16. Duffy, F. H., Burchfiel, J. L., and Lombroso, C. T. Brain electrical activity mapping (BEAM): A method for extending the clinical utility of EEG and evoked potential data. Ann. Neurol., 1979, 5: 309–321.Google Scholar
  17. Duffy, F. H., Jones, K., Bartels, P., Albert, M., McAnulty, G. B., and Als, H. Quantified neurophysiology with mapping: Statistical inference, exploratory and confirmatory data analysis. Brain Topog., 1990, 3: 3–12.Google Scholar
  18. Gasser, T., Bacher, P., and Mocks, J. Transformation towards the normal distribution of broad band spectral parameters of the EEG. Electroencephalogr. Clin. Neurophysiol., 1982,. 53: 119–124.Google Scholar
  19. Golub, G. H. Matrix Computations (2 ed.). Baltimore, MD: Johns Hopkins University Press, 1989.Google Scholar
  20. Hachinski, J. Cerebral blood differentiation of Alzheimer's disease from multi-infarct dementia. In R. Katzman,R. D. Terry, and K. T. Bick (Eds.), Alzheimer's Disease: Senile Dementia and Related Disorders (pp. 97–104). New York: Raven Press, 1978.Google Scholar
  21. Harner, R. N., and Riggio, S. Application of singular value decomposition to topographic analysis of flash-evoked potentials. Brain Topog., 1989, 2: 91–98.Google Scholar
  22. Harner, R. N., Warner, J. A., and Riggio, S. Localization of disordered brain function by singular value decomposition and spatial rotation of orthogonal features of the multichannel flash-evoked potential(in press). Brain Topog., 1991.Google Scholar
  23. Horst, P. Factor Analysis of Data Matrices. New York: Holt, Rinehart, Winston, 1965. Hotelling, H. Analysis of complex statistical variables into principal components. J. Ed. Psychol., 1933, 24: 417–441.Google Scholar
  24. John, E. R., Karmel, B. Z., Corning, W. C., Easton, P., Brown, D., Ahn, H., John, M., Harmony, T., Prichep, L., Toro, A., Gerson, I., Bartlett, F., Thatcher, R., Kaye, H., Valdes, P., and Schwartz, E. Neurometrics. Science, 1977, 196: 1393–1410.Google Scholar
  25. John, E. R., Prichep, L. S., Fridman, J., and Easton, P. Neurometrics: Computer-assisted differential diagnosis of brain dysfunctions. Science, 1988, 239: 162–169.Google Scholar
  26. John, E. R., Walker, P., Cawood, D., Rush, M., and Gehrmann, J. Factor analysis of evoked potentials. Electroencephalogr. Clin. Neurophysiol., 1973, 24: 33–43.Google Scholar
  27. Jones, K. J. Relation of R-analysis factor scores and Q-analysis factor loadings. Psychol. Reports, 1967, 20: 247–249.Google Scholar
  28. Kaiser, H. F. The application of electronic computers to factor analysis. Ed. Psych. Measmt., 1960, 20: 141–151.Google Scholar
  29. Kaiser, H. J. A Varimax criterion for analytic rotation in factor Analysis. Psychometrika, 1958,. 23: 187–200.Google Scholar
  30. Kaiser, H. J. A computer program for Varimax rotation factor analysis. Ed. Psych. Measmt., 1959, 19: 413–420.Google Scholar
  31. Kavanagh, R. N., Darcey, T. M., and Fender, D. H. The dimensionality of the human visual evoked scalp potential. Electroencephalogr. Clin. Neurophysiol., 1976, 40: 633–644.Google Scholar
  32. Lopes da Silva, F. H.. Event-related potentials: Methodology and quantification. In E. Niedermeyer and F. Lopes da Silva (Eds.), Electroencephalography Basic Principles, Clinical Applications and Related Fields (pp. 763–772). Baltimore-Munich: Urban and Schwarzenberg, 1987.Google Scholar
  33. Maier, J., Dagnelie, G., Spekreijse, H., and van Dijk, B. W. Principal components analysis for source localization of VEPs in man. Vision Res., 1987, 27: 165–177.Google Scholar
  34. Molfese, D. L., Buhrke, R. A., and Wang, S. L. The right hemisphere and temporal processing of consonant transition durations: electrophysiological correlates. Brain Lang., 1985,. 26, 49–62.Google Scholar
  35. Oken, B. S., and Chiappa, K. H. Statistical issues concerning computerized analysis of brainwave topography. Ann. Neurol., 1986, 19: 493–494.Google Scholar
  36. Rawlings, R. R., Rohrbaugh, J. W., Begleiter, H., and Eckardt, M. J. Spectral methods for principal components analysis of event-related brain potentials. Comput. and Biomed. Res., 1968, 19: 497–507.Google Scholar
  37. Rogers, L. J., and Douglas, R. R. Comparative factor analysis models for an empirical study of EEG data, III: resolution of stationary generating signals with confirmation by variable deletion. Internat. J. Neurosci., 1984, 24: 167–183.Google Scholar
  38. Rösler, F., and Manzey, D. Principal components and varimax-rotated components in event-related potential research: some remarks on their interpretation. Biol. Psychiatry, 1981, 13: 3–26.Google Scholar
  39. Scherg, M., and Von Cramon, D. Two bilateral sources of the late AEP as identified by a spatio-temporal dipole model. Electroencephalogr. Clin. Neurophysiol., 1985, 62: 32–44.Google Scholar
  40. Seal, H. L.. Principal components. In H. L. Seal (Eds.), Multivariate Statistical Analysis for Biologists (pp. 101–122). New York: John Wiley and Sons, 1964.Google Scholar
  41. Semlitsch, H. V., Anderer, P., Schuster, P., and Presslich, O. A solution for reliable and valid reduction of ocular artifacts, applied to the P300 ERP. Pschophysiol., 1986, 23: 695–703.Google Scholar
  42. Silberstein, R. B., Cadusch, P. J., and Schier, M. A. Effects of conduction on spatial components analysis applied to brain electrical activity. Brain Topography, 1992, 4(4): 267–276.Google Scholar
  43. Skalka, H., and Holman, J. Effect of pupillary dilatation in flash VER testing. Doc. Ophthalmol., 1986, 63: 321–324.Google Scholar
  44. Skrandies, W., and Lehmann, D. Spatial principal components of multichannel maps evoked by lateral visual half-field stimuli. Electroencephalogr. Clin. Neurophysiol., 1982, 54: 662–667.Google Scholar
  45. Suter, C. M. Principal component analysis of average evoked potentials. Exp. Neurol., 1970, 29(2): 317–327.Google Scholar
  46. Valdes, P., Bosch, J., Grave, R., Hernandez, J., Riera, J., Pascual, R., and Boscay, R. Frequency domain models of the EEG. Brain Topography, 1992, 4(4): 309–319.Google Scholar
  47. Van Rotterdam, A. Limitations and difficulties in signal processing by means of the principal-components analysis. I.E.E.E. Trans. Biomed. Eng., 1970,. BME-17: 268–269.Google Scholar
  48. Wastell, D. G. On the independence of P300 and the CNV: A short critique of the principal components analysis of donchin et al. (1975). Biol. Psychiat., 1979, 9: 171–176.Google Scholar
  49. Weintraub, S., Baratz, R., and Mesulam, M. M. Assessment of daily living in Alzheimer's disease. In S. Corkin,K. L. Davis, and J. H. Growdon (Eds.), Alzheimer's Disease: A Review of Progress (pp. 189–192). New York: Raven Press, 1982.Google Scholar
  50. Wood, C. C., and McCarthy, G. Principal component analysis of event-related potentials: Simulation studies demonstrate misallocation of variance across components. Electroenceph. Clin. Neurophysiol., 1984, 59: 249–260.Google Scholar
  51. Zar, J. H. Biostatistical Analysis. Englewood Cliffs, N. J.: Prentice Hall, 1984.Google Scholar

Copyright information

© Human Sciences Press, Inc 1992

Authors and Affiliations

  • Frank H. Duffy
    • 1
  • Kenneth Jones
    • 2
  • Peter Bartels
    • 3
  • Gloria McAnulty
    • 1
  • Marilyn Albert
    • 4
  1. 1.Department of NeurologyChildrens Hospital and Harvard Medical SchoolBostonUSA
  2. 2.Florence Heller Graduate School for Advanced Studies in Social WelfareBrandeis UniversityWalthamUSA
  3. 3.Optical Sciences CenterUniversity of ArizonaTusconUSA
  4. 4.Departments of Neurology and PsychiatryMassachusetts General Hospital and Harvard Medical SchoolBostonUSA

Personalised recommendations