The octave approach to EEG analysis

Abstract:

A “tonal” approach to EEG spectral analysis is presented which is compatible with the concept of physical octaves, thus providing a constant resolution of partial tones over the full frequency range inherent to human brain waves, rather than for equidistant frequency steps in the spectral domain. The specific advantages of the tonal approach, however, mainly pay off in the field of EEG sleep analysis where the interesting information is predominantly located in the lower octaves. In such cases the proposed method reveals a fine structure which displays regular maxima possessing typical properties of “overtones” within the three octaves 1-2 Hz, 2-4 Hz and 4-8 Hz. Accordingly, spectral patterns derived from tonal spectral analyses are particularly suited to measure the fine gradations of mutual differences between individual EEG sleep patterns and will therefore allow a more efficient investigation of the genetically determined proportion of sleep EEGs. On the other hand, we also tested the efficiency of tonal spectral analyses on the basis of our 5-year follow-up data of 30 healthy volunteers. It turned out that 28 persons (93.3%) could be uniquely recognized after five years by means of their EEG spectral patterns. Hence, tonal spectral analysis proved to be a powerful tool also in cases where the main EEG information is typically located in the medium octave 8-16 Hz.

Key-words

EEG - Sleep - Spectral Analysis - Tonal Approach

¹ The optimum discrimination between individuals as well as an optimum within-subject similarity served together as a single validation criterion during the iteration.

• REFERENCES

• 1 Dumermuth G. Numerical spectral analysis of the electroencephalogram. In: Remond A. ed. Handbook of Electroencephalography and Clinical Neurophysiology. Amsterdam: Elsevier Science Publ; 1973: 33-60.
• 2 Priestley MB. Spectral Analysis and Time Series. Vol. 1, 2. London, New York: Academic Press; 1981: 816-90.
• 3 Stassen HH. Computerized recognition of persons by EEG spectral patterns. Electroenceph Clin Neurophysiol 1980; 49: 190-4.
• 4 Stassen HH. The similarity approach to EEG analysis. Meth Inform Med 1985; 24: 200-12.
• 5 Stassen HH, Lykken DT, Propping P, Bom-ben G. Genetic determination of the human EEG (survey of recent results from twins reared together and apart). Human Genetics 1988; 80: 76.
• 6 Bracewell RN. The Fourier Transform and its Applications. (2nd rev. ed.) New York: McGraw Hill; 1986
• 7 Stassen HH. Masking artifacts in EEG time scries. Mcth Inform Med 1987; 26: 201-4.
• 8 Stassen HH, Bomben G. Longterm stability of EEG: A 5-year follow-up study. (to be published).
• 9 Dumermuth G, Lange B, Lehmann D, Meier CA, Dinkelmann R, Molinari L. Spectral analysis of all-night sleep EEG in healthy adults. Eur Neurol 1983; 22: 322-39.
• 10 Borbely AA, Baumann F, Brandeis D, Strauch I, Lehmann D. Sleep Deprivation: effect on sleep stages and EEG power density in man. Electroenceph Clin Neurophysiol 1981; 51: 483-93.
• 11 Schlegel T, Kurella B, Meister K. Automatic sleep analysis. I. Scoring of parameters. EEG-EMG 1990; 21: 13-9.
• 12 Schlegel T, Kurella B, Heitmann A, Meister K, Bzrozowski M. Automatic sleep analysis. II. Determination of sleep stages. EEG-EMG 1990; 21: 103-10.
• 13 Obergottsberger S, Zeitlhofer J, Mayr N, Marschnigg E, Deeke L. Possibilities and limitations of the automatic analysis of sleep stages using the Oxford system. EEG-EMG 1990; 21: 29-34.
• 14 Dijk DJ, Brunner DP, Borbely AA. Time course of EEG power density during long sleep in humans. Am J Physiol 1990; 258: R650-61.
• 15 Linkowski P, Kerkhofs M, Hauspie R, Susanne C, Mendlewicz J. EEG sleep patterns in man: a twin study. Electroenceph Clin Neurophysiol 1989; 73: 279-84.