Variation of temporal characteristics in human cerebral hemodynamic responses to electric median nerve stimulation: a near-infrared spectroscopic study

Neuroscience Letters 316 (2001) 75–78 www.elsevier.com/locate/neulet

Variation of temporal characteristics in human cerebral hemodynamic responses to electric median nerve stimulation: a near-infrared spectroscopic study Masato Tanosaki a,*, Yoko Hoshi a, Yoshinobu Iguchi a, Yukio Oikawa b, Ichiro Oda b, Motoki Oda c a

Department of Integrated Neuroscience, Tokyo Institute of Psychiatry, 2-1-8 Kamikitazawa, Setagaya-Ku, Tokyo 156-8585, Japan b Technology Research Laboratory, Shimadzu, 380-1 Horiyamashita, Hatano-City, Kanagawa 259-1304, Japan c Central Research Laboratory, Hamamatsu Photonics KK., 5000 Hirakuchi, Hamakita-City, Shizuoka 434-8601, Japan Received 11 September 2001; received in revised form 8 October 2001; accepted 9 October 2001

Abstract Using near-infrared spectroscopy, we studied cerebral hemodynamic responses to electric median nerve stimulation in ten subjects. The recordings were conducted by optical fibers placed over the left scalp. Electric stimuli were delivered to contra- and ipsilateral median nerves, respectively. Hemodynamic responses in the secondary somatosensory cortex were observed following each median nerve stimulation, except for three drowsy subjects. The contralateral stimulation tended to induce a larger response. The degree of change in oxygenated hemoglobin was hardly related to stimulus intensities, and was augmented by attention. Four subjects showed long-lasting responses throughout the stimulus periods, while three other subjects revealed transient responses. Thus, taking account of the temporal activation patterns is necessary for proper interpretation of the hemodynamic response following electric nerve stimulation. q 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Near-infra-red spectroscopy; Hemodynamic response; Secondary somatosensory cortex; Electric median nerve stimulation; Selective attention; Temporal characteristics; Interindividual variability

Somatosensory evoked potentials (SEPs) and magnetic fields (SEFs) elicited by electric stimulation of the peripheral nerve are widely prevailing techniques to explore the human somatosensory system [1,10,13]. Recent neuroimaging techniques, such as positron emission tomography (PET) and functional magnetic resonance imaging (fMRI), have also been employed to assess the somatosensory system. However, the reported neuroimaging data are not necessarily consistent: an increase of fMRI signals was observed in the secondary somatosensory cortex (SII) following electric median nerve stimulation with a frequency of 3 Hz [3], while no SII activation was found in a PET study with the same stimulation protocol [7]. In addition, activation in the primary somatosensory cortex (SI) has not always been detected in neuroimaging studies * Corresponding author. Department of Psychophysiology, Tokyo Institute of Psychiatry, 2-1-8 Kamikitazawa, Setagaya-Ku, Tokyo 156-8585, Japan. Tel.: 181-3-3304-5701; fax: 181-3-33298035. E-mail address: [email protected] (M. Tanosaki).

with electric nerve stimulation [2,3,7,9,11,12]. What causes such discrepancies has not yet been fully understood. PET and fMRI measure cerebral hemodynamic changes accompanying those in neural activities. It is conceivable that changes in hemodynamics following electric nerve stimulation are dynamic rather than static. This implies that the results obtained in PET and fMRI studies can vary with measurement periods. However, it is difficult to confirm this with these techniques. Unlike PET and fMRI, nearinfrared spectroscopy (NIRS) enables us to assess hemodynamic changes continuously [5,8]. In this study, therefore, we employed a multi-channel NIRS system to examine spatial and temporal characteristics of cerebral hemodynamic responses to electric median nerve stimulation. The subjects were ten healthy volunteers (eight male, aged 21–37 years). Written informed consent was obtained before examinations. Electric stimuli with 0.2 ms duration were delivered to each median nerve at the wrist with a repetition rate of 2 Hz. During recordings, subjects were sitting in a dark room and instructed to be awake. Stimulus

0304-3940/01/$ - see front matter q 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S03 04 - 394 0( 0 1) 02 37 2- 2

76

M. Tanosaki et al. / Neuroscience Letters 316 (2001) 75–78

intensities with 150 and 100% of the motor threshold (MT) and 200, 110 and 100% of the sensory threshold (ST) were delivered to the right side, and with 200 and 100% of the ST to the left side. In addition, an attention paradigm was done, in which weak stimulation (with intensity of ST) was transiently interrupted at random intervals and the subject was instructed to count the number of interruptions. Each stimulus condition lasted for 1.5 min, and stimulation was started after 15 s or more of stable baseline hemodynamics were obtained. The recordings were conducted using a NIRS instrument containing six sets of incident and detecting optical fibers (OMM 2000, Shimadzu). Twelve optical fibers were mounted on a 10 £ 10 cm square pad placed on the left scalp. Since the left primary sensorimotor hand area neighbors the C3 of the 10–20 International System, the center of the pad was placed 3 cm lateral to the C3 to cover both the SI hand area and the SII (Fig. 1a). This arrangement of the fibers enabled us to measure relative concentration changes in oxygenated ([oxy-Hb]), deoxygenated ([deoxy-Hb]) and total hemoglobin ([t-Hb]) in a total of 16 channels (Ch1– 16). NIRS signals were obtained with a sampling rate of 10 Hz, and 59 serial data were averaged automatically to improve the signal-to-noise ratio. In each channel, we also determined the mean optical path length using a single channel time-resolved spectroscopic instrument (TRS-10, Hamamatus Photonics KK.) to obtain absolute values of concentration changes. Since [oxy-Hb] is the most sensitive parameter of activity-dependent hemodynamic changes in optical measurement [6], we mainly analyzed changes of [oxy-Hb]. In each subject, we first computed topographical images of [oxy-Hb] within the square area, and then selected the channel in which the hemodynamic change was the most prominent. Changes of [oxy-Hb] in this channel were analyzed from the stimulus onset to a certain point within the stimulus period. The period used for analysis, which was different between the subjects but identical in all stimulus conditions for each subject, was determined so that increases of [oxyHb] were observed in most conditions as compared with the pre-stimulus baseline. Thus, in cases where the response lasted beyond the stimulation period, the analysis period was the same as the stimulation period (Fig. 1b); in cases where the response was transient, the analysis period was shorter than the stimulation period (Fig. 2). The mean value of the concentration change was calculated by dividing the total change in the analysis period by the number of sampling points. Differences of these values for various intensity conditions were tested by analysis of variance (ANOVA), using non-parametric methods (the Kruskal– Wallis rank test). The tests were performed for the right and left side stimulations individually. P , 0:05 was defined as a significant level. The mean value for MT was 3.9 mA; those for ST in the right and left sides were 1.7 and 2.2 mA, respectively. Three subjects did not show significant responses and were

excluded from further analyses. They reported drowsiness in the dark recording room. In the remaining seven subjects, increases of [oxy-Hb] and [t-Hb] and a decrease of [deoxyHb] were detected following median nerve stimulation (Fig. 1b). Locations of the most prominently activated channel were distributed over the parietal operculum (Ch1 and Ch13 each in two subjects, and Ch2, Ch6 and Ch9 each in one subject). In contrast, there was little intraindividual variation of spatial activation patterns in different stimulus conditions (Fig. 2). The SI hand area was not activated in all the subjects. The temporal configuration of the response was reproducible in each subject between different conditions (Fig. 2). Four out of the seven subjects showed long-lasting increases of [oxy-Hb] (Fig. 1b), and in the other three subjects, transient responses were observed at the beginning of the stimulation (Fig. 2). The differences between changes in [oxy-Hb] under various intensity conditions were not significant in each side stimulation (ANOVA). The contralateral stimulation tended to induce larger changes than the ipsilateral stimulation. The attention task augmented the response (Fig. 3). In the present study, the combination of the multi-channel NIRS system and the single channel time-resolved apparatus enabled us to obtain the quantitative functional map, which was analyzed by statistical procedures. Reproducible hemodynamic responses were observed over the parietal operculum following electric median nerve stimulation of each side. The locations of the responses were distinct from the C3, which is neighboured by the SI hand area. The SII is located in the upper bank of the lateral sulcus within the parietal operculum, and has many neurons responding to bilateral stimuli [4]. Thus, the location of responses is considered to correspond to the SII. In addition, we performed MRI scans with spherical lipid markers in two of our participants and confirmed that their activation areas were over the parietal operculum. SII activation is further supported by the following characteristics of the response: the increase of [oxy-Hb] in this area was not detected in drowsy subjects, was hardly related to stimulus intensities, and was augmented by attention. Previous fMRI studies showed that activation in the SII was abolished by low isoflurane anesthesia [2] and irrelevant to the intensities [3]. In addition, attention-related augmentation of the SII activity was found in previous fMRI [3] and SEF studies [10]. Our results first showed the great interindividual variability of temporal characteristics in hemodynamic responses to electric nerve stimulation. The response did not last throughout the stimulation period in three subjects. This may be explained by long-term habituation of the SII responses [13]. This implies that we should be careful in the interpretation of the grand-averaged data. For example, a transient response will decline or disappear after all the data throughout the stimulation period are averaged. Besides, the great interindividual variation in temporal profiles suggests that the hemodynamic states of the post-

M. Tanosaki et al. / Neuroscience Letters 316 (2001) 75–78

77

Fig. 1. (a) The arrangement of optical fibers. Hemodynamic changes were measured at 16 regions between each pair of incident and detecting fibers. We termed these areas as channels, and tentatively decided their nomenclature as shown in the figure. (b) A representative hemodynamic response to right median nerve stimulation with an intensity of 150% of MT. Violet bars denote the stimulus period, which is identical to the analysis period in this subject. The inserted topographical image in (a) was reconstructed from the data of [oxy-Hb] at the time when the most prominent change was seen (dotted vertical lines in Ch9). Note that the response lasted throughout the stimulation periods.

stimulus resting interval may differ from subject to subject. In most studies, the cerebral blood flow is assumed to return to its baseline level at the end of the resting interval. However, this assumption is difficult to testify without real-time monitoring. We could not systematically assess

the long-lasting responses and could not always perform the recording until the hemodynamics returned to baseline levels, because the experiment time was restricted to avoid fatigue of the subjects. However, the stable resting state recordings in all of the 16 channels, which often required several minutes to be obtained, strongly suggested that there was no remaining effect of the proceeding stimuli at the stimulus onset. We consider that the interindividual variation of the temporal activation patterns and the post-stimulus resting hemodynamics provides one possible explanation for the inconsistency between previous PET

Fig. 2. Topographical images of [oxy-Hb] and hemodynamic changes in a representative channel (Ch13). Both high (100% of MT) and low (110% of ST) stimulus intensity conditions are shown. Violet bars and black bars denote the stimulation and analysis periods for the changes of [oxy-Hb] in this subject, respectively. Note that the increases of [oxy-Hb] were transient, but in close association with the stimulus onset, and that the areas over Ch13 were activated in both high and low intensity conditions.

78

M. Tanosaki et al. / Neuroscience Letters 316 (2001) 75–78

reported during electric nerve stimulation at low frequencies, such as 0.5–4 Hz [7,12], whereas there were other reports emphasizing that activation in the SI appeared only after high frequency stimulation of more than 15 Hz [9,11]. This controversy might also be explained by interindividual variation of the temporal patterns of activation. Thus, future work should be focused on the relationship between the hemodynamic changes and the stimulus frequency in both the SI and SII areas.

Fig. 3. (a) Relationship between changes of [oxy-Hb] and stimulus intensities in right median nerve stimulation. The intensities were graded based on MTs and STs. The error bars represent standard deviations. (b) Comparison of changes of [oxy-Hb] during right versus left median nerve stimulation (RM versus LM), and with versus without selective attention to the stimuli with an intensity of 100% of ST (attention). Right median nerve stimulation tended to induce a larger change of [oxy-Hb]. The selective attention augmented the change of [oxy-Hb].

and fMRI studies on the SII activation following low frequency electric median nerve stimulation [3,7]. The protocol of electric stimulation employed here was identical with those routinely used for SEPs and SEFs. Nevertheless, activation of the SI, in which most of the electrophysiological responses were generated [1], was not detected. In the SI, the topographic differentiation is highly organized, whereas that in the SII is not so clear [4,7]. Although the overall surface area of the SI is larger than that of the SII [4], it is plausible that the activated SI area during electric median nerve stimulation is so confined that its hemodynamic change cannot be observed over the scalp. Alternatively, the stimulus frequency of 2 Hz employed here may explain the absence of activation in the SI. The relationship between activation in the SI and the stimulus frequency is controversial. Significant SI activation was

[1] Allison, T., McCarthy, G., Wood, C.C., Darcey, T.M., Spencer, D.D. and Williamson, P.D., Human cortical potentials evoked by stimulation of the median nerve. I. Cytoarchitectonic areas generation short-lasting activity, J. Neurophysiol., 62 (1989) 694–710. [2] Antognini, J.F., Buonocore, M.H., Disbrow, E.A. and Carstens, E., Isoflurane anesthesia blunts cerebral responses to noxious and innocuous stimuli: a fMRI study, Life Sci., 61 (1997) 349–354. [3] Backes, W.H., Mess, W.H., van Kranen-Mastenbroek, V. and Reulen, J.P.H., Somatosensory cortex responses to median nerve stimulation: fMRI effects of current amplitude and selective attention, Clin. Neurophysiol., 111 (2000) 1738– 1744. [4] Burton, H., Second somatosensory cortex and related areas, In E.G. Jones and A. Peter (Eds.), Cerebral Cortex, Plenum, New York, 1986, pp. 31–98. [5] Hoshi, Y., Oda, I., Wada, Y., Ito, Y., Yamashita, Y., Oda, M., Ohta, K., Yamada, Y. and Tamura, M., Visuospatial imagery is a fruitful strategy for the digit span backward task: a study with near-infrared optical tomography, Cogn. Brain Res., 9 (2000) 339–342. [6] Hoshi, Y., Kobayashi, N. and Tamura, M., Interpretation of near-infrared spectroscopy signal: a study with a newly developed perfused rat brain model, J. Appl. Physiol., 90 (2001) 1657–1662. [7] Iba´n˜ez, V., Deiber, M.P., Sadato, N., Toro, C., Grisson, J., Wood, R.P., Mazziotta, J.C. and Hallett, M., Effects of stimulus rate on regional cerebral blood flow after median nerve stimulation, Brain, 118 (1995) 1339–1351. [8] Jo¨bsis, F.F., Noninvasive infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters, Science, 198 (1977) 1264–1267. [9] Kampe, K.K.W., Jones, R.A. and Auer, D.P., Frequency dependence of the functional MRI response after electric median nerve stimulation, Hum. Brain Mapp., 9 (2000) 106–114. [10] Mima, T., Nagamine, T., Nakamura, K. and Shibasaki, H., Attention modulates both primary and second somatosensory cortical activities in humans: a neuromagnetic study, J. Neurophysiol., 80 (1998) 2215–2221. [11] Puce, A., Constable, T., Luby, M.L., McCarthy, G., Nobre, A.C., Spencer, D.D., Gore, J.C. and Allison, T., Functional magnetic resonance imaging of sensory and motor cortex: comparison with electrophysiological location, J. Neurosurg., 83 (1995) 262–270. [12] Spiegel, J., Tintera, J., Gawehn, J., Stoeter, P. and Treede, R.D., Functional MRI of human primary somatosensory and motor cortex during median nerve stimulation, Clin. Neurophysiol., 110 (1999) 47–52. [13] Wegner, K., Foss, N. and Salenius, S., Characteristics of the human contra- versus ipsilateral SII cortex, Clin. Neurophysiol., 111 (2000) 894–900.