CC BY
12Reflection of processes of long-term memory actualization
in characteristics of the event-related brain potentials on
production of time intervals
Khodanovich M.Y. (Mhodanovich@mail.ru) (2), Bushov J.V. (1), Ivanov A.S. (2), Rjabova G.A. (1), Vjachistaja J.V. (2)
(1) Tomsk State University, (2) Research Institute of Biology and Biophysics
It is common opinion that subjective unit of time measurement, or "time standard", which is used by subject for measurement of duration of events, is formed on the basis of subject's individual experience. This time standard is stored in long-term memory and has relative stability (OoHeoBa H.A., fflecroBa H.Á., 1988). But it is incomprehensible what this standard is, where is it stored in human brain and what is mechanisms of its extraction from long-term memory. These questions remain without answers today. Most of explorations of timing mechanisms do not examine processes of actualization of long-term memory (Hazeltine E., Helmuth L.L., Ivry R., 1997; Ivry R., Mangles J., 1997).
One of the most promising approaches to study these processes is to analyze electrophysiological correlates of time perception, in particular, event-related potentials (ERP). This method is one of basis in experimental psychophysiology, because this method allows observe main stages of information processing during estimation of time parameters of stimulus. The exploration of brain sources of ERP allows determine which brain structures take part in this process.
At present time, there is numerous experimental data that is being evidence that processes of long-term memory actualization are reflected in characteristics of ERP. The most of researchers relate the process of recollection to appearance of late positive component in piece of 400-800 ms from stimulus onset. (Paller K. A., 2001; Schweinberger R. S., Pickering C., Mike Burton A., Kaufmann M. J., 2002 u gp.). But there are investigations, in which this process relates to appearance of negative component of the same latency (Roesler F., Heil M., Hennighausen E., 1995).
The aim of our exploration is to search of EEG-correlates of reference to long-term memory during producing of time intervals. We suppose that production of time intervals, when the intervals are defined by digits on monitor screen, is accompanied by addressing to the time standard stored in long-term memory, and this process is reflected in characteristics of ERP.
Method
22 practically healthy subjects (9 male, 13 female) aged 18-24 years take part in experiments. Participants were volunteers and were paid for performing physiological tests.
Subject sat in a comfortable chair, in sound-attenuated, weakly illuminated, electrically isolated room. At the beginning of experimental session the subjects were told don't use mental arithmetic, knocking or other similar methods for time interval determination.
Subject was presented digits in range of 0.1-5.5 on monitor screen. Subject was instructed to produce time interval in seconds by double pressing a "space" key in case when presenting number was more or equal to 1 (stimulus type A) or to do quick double pressing a "space" key in case of representing number was less than 1 (stimulus type B). The sequence of stimuli formed in following manner: 1) 150 numbers in range of 0.1-0.9 (0.1, 0.2...0.9 - 10 types, 15 stimuli of each type) and 50 numbers in range of 1.0-5.5 (1.0, 1.5.5.5 - 10 types, 5 stimuli of each type); 2) stimuli presented in random sequence, but there were at least 3 stimuli of type A between stimuli of type B. These demands to sequence of stimuli were used in order to ensure, that subject will address to long-term memory, but not to image of previous stimulus in volatile storage. Duration of stimulus presentation was 400 ms, interval between the second pressing a key and appearance of the next stimulus was 600 ms. Average latent period (LP) of motor response and interval between first and second pressing a key were defined for both types of stimulus from test results.
Simultaneously with presentation of visual stimuli EEG, EOG and SGR were recorded. For EEG recording adhesive electrodes ("Nicolet") were used. EEG recorded by using 16-channel encephalograph ("Medicor") with band pass of 0.23-200 Hz monopolarly from 15 sites (F3, Fz, F4, C3, Cz, C4, T3, T4, T5, T6, P3, Pz, P4, 01, 02) in international 10-20% system. The common referent electrode was placed on the left and right lobes of the ears; the ground electrode was placed on the chin. Subject's eyes were fixed on the monitor screen.
EOG and SGR were recorded for rejection of artifacts related to eye movements and electrodermal activity. Weekly polarized ceramic electrodes were placed above eyebrow and on the infraorbital region of left eye. Ag-AgCl electrodes for SGR recordings were placed on the interior and external sides of wrist of left hand. All recordings were sampled at 1036 Hz. Native EEG and ERP were filtered with band pass 0-30 Hz (Hamming window).
Only trials free of artifacts (as determined visually) were included in the each averaging. The analysis epoch including prestimulus 100 ms was 700 ms for visual stimuli and 600 ms for pressing a key "space". ERP were corrected by prestimulus baseline of 100 ms duration. ERP to visual stimulus which specified interval for production, ERP to visual stimulus which demand only double pressing, ERP to first and second pressing a key for each type of stimulus were averaged separately.
Data analyzed with help of application Statistica 6.0 and programs worked out in our lab on language Borland Delphi 5.0 and Visual Basic for Applications in Statistica 6.0.
Estimation of significance of distinctions is carried out pointwise using nonparametric Wilcoxon test.
Results
Accuracy of task performance and the latent period of motor response. The basic results are submitted in table 1. As follows from the table, LP of motor response to stimulus of type A significantly (p < 0.01) exceeds LP of reactions to stimulus which demanded simple double pressing a key "space". So, average LP of motor response was 379.50 ± 46.09 ms on producing intervals with timing, and 241.04 ± 37.58 ms - when subject just double pressed a key. Accuracy of timing of neighbor intervals (1.5 and 2 ms, 2 and 2,5 ms, 2.5 and 3 ms, 3 and 3.5 ms, 3.5 and 4 ms, 4 and 4.5 ms, 4.5 and 5 ms, 5 and 5.5 ms) did not differ statistically significantly, except for intervals 1 and 1.5 ms. Interval of 1 sec was produced most precisely.
Table 1. Accuracy of task performance and LP of motor response
Stimulus
'1.0"
'1.5"
"2.0"
"2.5"
"3.0"
"3.5"
"4.0"
"4.5"
"5.0"
"5.5"
"0.1" -"0.9"
Notice. M
Standard, ms
1000
1500
2000
2500
3000
3500
4000
4500
5000
5500
Double pressing
Produced interval, ms
893.13±102.92 **
1068.40±107.11 **
1291.18±175.71 **
1393.63±171.82 **
1687.10±257.16
1886.50±267.55 **
2192.02±350.01 **
2300.93±386.61 **
2613.47±408.84 **
2788.28±526.44 **
492.96± 51.30
Standard deviation
331.67
373.57
361.48
540.33
494.73
559.28
568.38
490.01
675.45
667.97
Standard error, %
-10.69
-28.77
-35.44
-44.26
-43.76
-46.10
-45.20
-48.87
-47.73
-49.30
Stimulus type LP of motor response, ms
A, "0.1" - "0.9" 241.04±37.58 **
B, "1.0" - "5.5" 379.50±46.09
of motor response and produced intervals for
arked significant differences between LP ranges 0.1-0.9 and 1.0-5.5 sec: * - p<0.05, ** - p<0.01 (Wilcoxon test). LP of motor response was measured from the ending of visual stimulus (digits) by 400 ms duration to first pressing a key "space".
**
ERP visual stimuli of different type (A or B). From 13 (5 male and 8 female) of 22 subjects were received ERP to relevant visual stimulus (digits), which in one case meant an interval of time for producing, in the other case just should be answered with simple motor reaction (double pressing a key). Significant distinctions between these potentials were found in all sites, except for T3 (figure 1). The earliest distinctions between the specified curves were observed in temporal and occipital sites: in piece of 40-75, 160-180 and 230-345 ms from the beginning of visual stimulus in
site T4, 180-200 ms in T5 site, 70-90 ms in T6 site and 90-95, 220-235 ms in 01 site. The ERP in these pieces were more negative for cases when timing of interval was required. Later distinctions in piece from 400-450 ms up to 600 ms from stimulus onset were observed in all sites, except for T3 and T4. ERP in this piece was more positive for cases when timing of interval was required. The distinctions found were, probably, related to presence of at least two ERP components: a negative component of small amplitude (1.2 mcV) with peak latency of 240 ms from stimulus onset and the most expressed in site T6, and a late positive component of amplitude 2.7 mcV with peak latency of 520-540 ms from stimulus onset most expressed in sites Pz and F3.
ERP to first and second pressing a key "space" on producing of intervals with and without timing. It is established that ERP on producing of intervals with and without timing in piece before the first pressing a key «space» practically did not differ. Significant distinctions between these potentials were found in an interval from 200 up to 500 ms after the first pressing a key in all parietal, occipital, frontal (F3, F4), central (C3, C4) and right temporal (T6) sites. ERP to producing intervals with timing were more negative, than the same ERP to producing intervals without timing. Apparently from figure 1, distinctions become significant after 200 ms after the first pressing a key (in site 02 - after 115 ms), reach maximum (4 mkB) after 500 ms in sites F4 and C3 and last until second pressing a key in both occipital and temporal (T3, T4, T6) sites. Also in piece after second pressing a key «space» ERP to producing intervals with timing were significantly more negative in occipital and left temporal (T3) sites and were more positive in right central site (C4) in comparison with ERP to simple double pressing a key.
Discussion
The carried out observation had shown that among researched intervals of time one second long intervals are produced most precisely. It corresponds to the literary data and specifies absence of a continuity in function of dependence of accuracy of an estimation and reproduction of intervals of time from size of an interval (Madison G., 2001). Now there is no common opinion concerning a point of break of this function. It is underlined that this point is located in duration close to 1 sec (Freyd J.J., Johnson J., 1987), 1.4 c (Madison G., 2001), 1.8 sec (Getty D.J., 1975). There is data that different brain structures participate in perception of short (less than 1 sec) and long (more than 1 sec) intervals of time (Hazeltine E., Helmuth, L.L., Ivry R., 1997): in perception of short intervals the basic role is played by a cerebellum, while in recognition of long intervals basal ganglia and a frontal cortex are involved. However, as some researchers (Ivry R., Mangles J., 1997) consider, the contribution of a frontal cortex can be not specific to processing the time information and finds out itself every time when requirements to working memory and attention raise.
:les/2004/136e.pdf
-6 n 0
6 -I
mcV
600ms
Stimuli "0.1"-"0.9", demanding simple double pressing a key
Stimuli "1.0"-"5.5", demanding producing of time interval
Figure 1. Averaged (N=13) ERP to visual stimuli of type A and B, demanding to produce interval with and without timing. Marked pieces with significant distinctions between curves (Wilcoxon test, p<0.05 - p<0.001).
Revealed by us early (in piece up to 400 ms from stimulus onset) distinctions in ERP to visual stimulus of different type (A or B), which demand different reaction from subjects, most likely were not related to the reference to long-term memory. It is known that during 300 - 500 ms after presentation of visual stimulus the estimation by a brain of its physical parameters, novelty and the importance, synthesis of this information and identification of stimulus is carried out (MBaHH^HH A.M., 1999). The specified stages of information processing necessarily proceed at participation of mechanisms of memory, but it is common for both types of stimulus. Probably, found out distinctions of ERP were caused by different frequency of presentation of stimulus of different type (stimulus of type B were presented more often than others).
Apparently, two found components may apply for a role of correlates of address to the subjective time standard: positive component ERP in piece 400-600 ms from the beginning of visual stimulus, and negative component in piece with 200 ms from the beginning of interval producing (the first pressing a key) up to the end of an interval (the second pressing a key). Some literary data assert in favour of it. In particular, at performance of recollection tasks the positive component in piece 400-800 ms from stimulus onset was found out at presentation of different types of visual stimulus: new and old words (Paller K.A., Kutas M., Mayes A.R., 1987), familiar and unfamiliar faces (Barrett S.E., Rugg M.D., Perrett D.I., 1988; Smith M.E., Halgren E., 1987), simple objects (Friedman D., Sutton S., 1987), slides with the image of people, landscapes, pictures (Neville H., Snyder E., Woods D., Galambos R., 1982). In a number of works later negative shift of potential, which is related to reference to long-term memory, is revealed also, and it has specific topography for different physical characteristics of stimulus (Roesler F., Heil M., Hennighausen E., 1995).
Changes of ERP observed in piece from 400 up to 800 ms from stimulus onset, apparently, were related to a stage of preparation of the motor response of the subject, with a stage of formation of the appropriate motor program. In a case of producing of interval with timing this stage is significantly longer than that in case of simple double pressing a key. Probably, it is related to necessity of formation for the first case of more complex motor program, that in turn demands participation of the greater number of the brain structures responsible for programming of motor acts. From this point of view the found out distinctions of ERP can be caused by different complexity of the motor programs formed producing of intervals with and without timing. Whether these distinctions may reflect the reference to long-term memory, to the subjective standard of time stored in it? Obviously, it is possible only if the appropriate program includes this stage. In the literature there is no common opinion on this account. Some researchers consider, that the motor program completely should be generated even before reproduction of an interval (Pasler H., 2001). In this case the reference to long-term memory occurs prior to the beginning of interval producing,
and found out by us positive wave distinctions on 400-600 ms from the visual stimulus onset reflect this process. On the other hand, the appropriate program only can order the reference to the subjective standard at the certain stage of its performance. Therefore now it is not clear, in what measure the found out wave distinctions of ERP at a stage of preparation of the motor response reflect the reference to the standard stored in long-term memory. The most probable correlate of this process is found out wave distinction of ERP in an interval between the first and second pressing a key «space». There is opinion that on time interval producing the subject is compelled to compare duration of an adjustable pause to the standard stored in memory (Костандов Э.А., Захарова Н.Н., Важнова Т.Н. и др., 1988). By simple double pressing a key «space» necessity for such checking is absent.
Found out wave distinctions ERP after the second pressing a key «space», probably, reflect the distinctions related to value judgment of results of activity on producing of interval with and without timing.
Conclusion
Thus, the analysis of the achieved results in comparison to the literary data testifies that process of extraction of subjective time standard from long-term memory can be related to positive component ERP in piece 400-600 ms from the onset of visual stimulus specifying an interval of time, but the most probable correlate of this process is the negative component of ERP which has been found out in piece between the first and second pressing a key «space» at producing of time intervals.
The maximum of a positive component is observed in an interval 520-540 ms from visual stimulus onset in parietal and frontal areas of a cortex. It also is well expressed in the central and occipital zones of a brain. Negative component of ERP to producing of time intervals reaches a maximum after 500 ms after the first pressing a key and is most expressed in right frontal and left central sites.
Asknowledgments: This work was supported by the grant of the Competitive Centre of Fundamental Natural Sciences PD 02-1.4 - 433.
References
1. Фонсова Н.А., Шестова И.А. Восприятие околосекундных интервалов времени// Биологические науки. - 1988. - №3. - С.59-72.
2. Hazeltine E., Helmuth L. L., Ivry R. Neural mechanisms of timing. Trends in Cognitive Sciences// 1997, №1, P. 163-169.
3. Ivry R., Mangles J. The many manifestations of a cerebellar timing mechanism. Presented at the Fourth Annual Meeting of the Cognitive Neuroscience Society, 1997, March 23.
4. Paller K.A. Neurocognitive foundation of human memory// In: The philosophy of learning and motivation. 2001, V.40, p. 121-145. D.L. Medin (Ed.), Academic Press, San Diego.
5. Schweinberger R. S., Pickering C., Mike Burton A., Kaufmann M. J. Human brain potential correlates of repetition priming in face and name recognition// Neuropsychologia. 2002, V.40, P.2057-2073.
6. Roesler, F., Heil, M. & Hennighausen, E. Distinct cortical activation patterns during long-term memory retrieval of verbal, spatial, and color information// Journal of Cognitive Neuroscience. 1995, N 7, P. 51-65.
7. Madison G. Functional modelling of human timing mechanism. Acta Universitatis Upsaliensis. Comprehansive summaries of Upsala Dissertations from the faculty of social sciences. 2001. 101. 77 p. Upsala. ISBN 91-554-5012-1.
8. Freyd J.J., Johnson J. Probing the time course of representational momentum// Jornal of Experimental Psychology: Learning, Memory and Cognition, 1987. 13. P. 259-268.
9. Getty D.J. Discriminations of short temporal intervals: A comparsion of two models// Perception and Psychophysics, 1975, 18, P. 1-8.
10. Иваницкий А. М. Главная загадка природы: как на основе работы мозга возникают субъективные переживания// Психол. журн. 1999, Т. 20, N3, С. 93.
11. Paller, K. A., Kutas, M., Mayes, A. R. Neural correlates of encoding in an incidental learning paradigm// Electroencephalography and Clinical Neurophysiology. 1987, V. 67., P. 360-371.
12. Barrett, S. E., Rugg, M. D., & Perrett, D. I. Event-related potentials and the matching of familiar and unfamiliar faces// Neuropsychologia. 1988. 26. 105-117.
13. Smith, M. E., Halgren, E. Event-related potentials elicited by familiar and unfamiliar faces// Electroencephalography and Clinical Neurophysiology. 1987.Supplement 40, 422-426.
14. Friedman, D., Sutton, S. Event-related potentials during continuous recognition memory// Electroencephalography and Clinical Neurophysiology. 1987. Supplement 40, 316-321.
15. Neville, H., Snyder, E., Woods, D., Galambos, R. Recognition and surprise alter the human visual evoked response. Proceedings of the National Academy of Sciences USA, 1982. 79, 2121-2123.
16. Pasler. H. Perception and production of brief durations: beet-based versus interval-based timing// Jornal of Experimental Psychology, 2001, V.27, N2, P.485-493.
17. Костандов Э.А., Захарова Н.Н., Важнова Т.Н. и др. Различение микроинтервалов времени эмоционально возбудимыми личностями// Ж. высш. нервн. деят-сти, 1988, Т. 38, вып. 4, С. 614.
