- Email: [email protected]
The effects of stimulus rates on high frequency oscillations of median nerve somatosensory-evoked potentials – direct recording study from the human cerebral cortex
Clinical Neurophysiology 113 (2002) 1794–1797 www.elsevier.com/locate/clinph
The effects of stimulus rates on high frequency oscillations of median nerve somatosensory-evoked potentials – direct recording study from the human cerebral cortex Eiichirou Urasaki a,*, Tetsuya Genmoto a, Naoki Akamatsu b, Shin-ichi Wada a, Akira Yokota a a
Department of Neurosurgery, School of Medicine, University of Occupational and Environmental Health, Iseigaoka 1-1, Yahata Nishi-ku, Kitakyushu City 807-8555, Japan b Department of Neurology, School of Medicine, University of Occupational and Environmental Health, Iseigaoka 1-1, Yahata Nishi-ku, Kitakyushu City 807-8555, Japan Accepted 21 August 2002
Abstract Objectives: To study the effects of different stimulus rates on high-frequency oscillations (HFOs) of somatosensory-evoked potentials (SEPs), we recorded median nerve SEPs directly from the human cerebral cortex. Methods: SEPs were recorded from subdural electrodes in 5 patients with intractable epilepsy, under the conditions of low (3.3 Hz) and high (12.3 Hz) stimulus rates. Results: Increased stimulus rates to the median nerve from 3.3 to 12.3 Hz showed a pronounced amplitude reduction of HFOs when compared with the primary N20–P20, area 3b, and P25, area 1, responses. Conclusions: HFOs were more sensitive to a high stimulus rate than the primary cortical responses, suggesting that the post-synaptic intracortical activities may greatly contribute to the HFO generation. q 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Somatosensory-evoked potentials; High-frequency oscillations; Direct recording; Stimulus rates
1. Introduction Primary cortical components of somatosensory-evoked potentials (SEPs) are known to have high-frequency oscillations (HFOs) that are clearly exposed by setting low bandpass filters greater than 300–500 Hz (Gobbele´ et al., 1998; Gobbele´ et al., 1999; Hashimoto et al., 1999; Klostermann et al., 1999; Nakano and Hashimoto, 1999; Gobbele´ et al., 2000; Halboni et al., 2000; Mackert et al., 2000; Maegaki et al., 2000; Kojima et al., 2001). The characteristics and significance of the HFOs have been investigated but there are some uncertainties and controversies. The stability of HFOs on cortical N20, for example, under various stimulus rates of the median nerve was reported, suggesting a presynaptic origin such as thalamo-cortical projection fibers (Gobbele´ et al., 1998; Gobbele´ et al., 1999). The intracortical activities were, on the other hand, proposed as generators based on the finding that the amplitude decrease was emphasized more in the HFOs than the corresponding cortical N20 when the stimulus rate was increased (Klostermann * Corresponding author. Tel.: 181-93-603-1611; fax: 181-93-691-8755. E-mail address: [email protected] (E. Urasaki).
et al., 1999). The present study provides direct recording data from the human cerebral cortex demonstrating a pronounced amplitude decrease in HFOs when compared with N20 by increasing the stimulus rates, and adds new findings that suggest that the same is true for cortical P20 and P25, an area 1 component. 2. Patients and methods Five male patients with medically intractable epilepsy aged between 16 and 42 years (30.6 ^ 10.3; mean ^ SD) were studied. Subdural electrodes were implanted intracranially to evaluate seizure localization and functional brain mapping prior to surgical treatment. The necessary number of intracranial electrodes was placed near the expected areas of epileptogenic, irritative, or symptomatogenic zones (Hahn and Lu¨ders, 1987; Lu¨ders and Awad, 1991). Intracranial electrode placement on areas superior to the Sylvian fissure including the central sulcus was required in all patients. Informed consent was obtained from all patients. This study was approved by the institutional Review Board of the University of Occupational and Environmental
1388-2457/02/$ - see front matter q 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S13 88- 2457(02)0029 1-2
CLINPH 2002077
E. Urasaki et al. / Clinical Neurophysiology 113 (2002) 1794–1797
1795
Table 1 Comparison of HFOs and the primary cortical response (N20, P20, and P25) between the conditions of 3.3 and 12.3 Hz stimulations a 3.3 Hz (mean ^ SD)
12.3 Hz (mean ^ SD)
N20 amplitude (mV) (n ¼ 5) HFO number of negative peaks HFO amplitudes (mV) HFOs/N20 amplitude ratio (%)
14.6 ^ 9.13 4.0 ^ 2.2 2.80 ^ 1.70 26.3 ^ 21.1
16.03 ^ 12.07 2.6 ^ 2.1 1.37 ^ 1.77 11.2 ^ 13.2*
P20 amplitude (mV) (n ¼ 5) HFO number of positive peaks HFO amplitudes (mV) HFOs/P20 amplitude ratio (%)
15.15 ^ 10.61 4.8 ^ 2.3 5.9 ^ 6.52 30.8 ^ 16.4
12.27 ^ 7.10 3.6 ^ 2.1 2.64 ^ 3.06* 17.5 ^ 12.3
P25 amplitude (mV) (n ¼ 2) HFO number of positive peaks HFO amplitudes (mV) HFOs/P25 amplitude ratio (%)
22.58 ^ 2.47 6.5 ^ 2.1 6.13 ^ 2.30 26.7 ^ 7.3
24.32 ^ 8.33 4.0 ^ 2.8 1.90 ^ 0.17 8.2 ^ 2.1
a
SD ¼ standard deviation; *P , 0:05.
Health. Neurological examinations including the somatosensory system showed normal findings in all patients except for their seizures. MRI revealed no abnormality in the postcentral gyrus or thalamus in any patient. The grid or strip electrodes in the perirolandic region were used for SEP recordings. The locations of the implanted subdural electrodes in the vicinity of the sensorimotor area were confirmed by a cortical stimulation study. The diameters of the recording electrodes in the subdural grid were 4 mm, and the interelectrode distance from center to center was 10 mm. To construct a functional brain map, biphasic 0.3 ms square wave electrical stimuli, 50/s frequency and 3–5 s duration were applied to the brain through the subdural electrodes. Each electrode was stimulated monopolarly by setting a common reference electrode chosen from the other intracranial electrodes that gave no response to electrical stimuli. The strength of the electrical current was 1–15 mA, or up to the appearance of the after discharge. SEPs were recorded after the functional mapping study. The reference electrode was placed on the ear contralateral to the side of stimulation. SEP was obtained using 0.2 ms square wave electrical pulses delivered transcutaneously to the median nerve at the wrist contralateral to the side of intracranial recording. Saddle-typed bipolar electrodes, a cathode 3 cm proximal to the anode, were used. The period of analysis was 50–100 ms. Stimulus intensity was adjusted to produce a thumb twitch. The filter setting was 10– 1000 Hz with an analysis time of 50–100 ms. At least two averages of 500–1000 responses were obtained to confirm reproducibility, and two sets of responses were averaged using a computer for later analysis. To maintain a stable vigilance level during the test, the patients were instructed to stay awake, and were interviewed about their vigilance level before and after every recording session. HFOs were extracted from underlying cortical responses by the digital filtering of the wide-band (10–1000) recorded
responses through a band-pass of 500–1000 Hz, using a Viking IV computer system (Nicolet). The amplitude of the primary cortical response was measured as the vertical distance between the onset and the peak of the primary cortical response. HFOs identified between the onset and endpoint of primary cortical responses (N20, P20, and P25) were analyzed. The number of negative or positive peaks of HFOs was counted between the onset and endpoint of respective N20, P20 and P25 responses. The amplitudes of the HFOs were averaged with the values measured as the vertical distances between two successive peaks. It was reported that the peak of N20 or P20 tended to coincide with the peak of maximal negative or positive HFOs (Ozaki et al., 1998; Maegaki et al., 2000). It may, therefore, be reasonable to measure the negative peaks or positive peaks to analyze the respective HFOs of N20 or P20 and P25. There is, however, no practical difference between the analysis of negative and positive peaks because the amplitude was measured from peak to peak. All negative HFO components were measured from the preceding positive peak to the following negative peak and then averaged for the amplitude calculation of HFOs superimposed on N20. Similarly, all positive components were measured and averaged for HFOs on P20 and P25. HFOs/primary cortical response amplitude ratio was calculated and revealed as a percentage. Two-way analysis of variance (ANOVA), multiple comparison test using rank values of variables, and nonparametric Wilcoxon sum rank test was used for statistical analysis, and a P value less than 0.05 was considered statistically significant.
3. Results Table 1 shows the comparison of the primary cortical response and HFO parameters between 3.3 and 12.3 Hz
1796
E. Urasaki et al. / Clinical Neurophysiology 113 (2002) 1794–1797
Fig. 1. Somatosensory evoked potentials (SEPs) after left median nerve stimulation, recorded from the subdural electrodes implanted in a 41year-old male patient with intractable epilepsy. Cortical SEPs such as P25, area 1 component, N20/P20, area 3b components, and their corresponding HFOs are shown. Note the marked decrease in HFOs despite the preserved primary cortical components at the higher stimulus rate. Electrical stimulation of electrodes 1, 2, and 3 elicited thumb motor, thumb sensory, and finger motor responses, respectively.
stimulation in 5 subjects (Table 1). The cortical N20–P20 area 3b components were recorded from all 5 patients, and the P25 area 1 component from two patients. In all the 5 patients, HFOs on N20 and P20 were obtained at two different stimulus rates, and the high stimulus rate either induced loss or decreased the HFO amplitudes in 4 of the 5 patients (Fig. 1). One patient showed no decrease in HFOs on the corresponding N20. The peripheral N9 showed a slight decrease in amplitude by the high stimulus rate without statistical significance (5.53 ^ 2.27 mV at 3.3 Hz and 4.49 ^ 0.82 mV at 12.3 Hz). Two-way ANOVA and multi-
Fig. 2. SEPs elicited by left median nerve stimulation recorded from a 27year-old male patient with intractable epilepsy. Primary cortical N20 response and corresponding HFOs for the 3.3 and 12.3 Hz stimulus rates are superimposed. The early part of the HFOs before the N20 peak is relatively preserved, but the later part of the HFOs after the N20 peak is strongly decreased. Electrical stimulation of this recording site elicited a hand sensory response.
ple comparison test using rank values of variables showed significant changes in the HFO amplitudes (P ¼ 0:044) and the HFOs/primary cortical response ratios (P ¼ 0:035) between the 3.3 and 12.3 Hz stimulus conditions, suggesting a safeguard against errors due to multiple comparisons. There were no significant changes in the N20, P20 and P25 amplitudes and the HFO number between these two different stimuli. Overall, HFOs on N20–P20 and P25 were more sensitive to change in the stimulus rate than the primary cortical responses. Mean HFO amplitudes were decreased while all primary responses showed a relatively slight increase in amplitude when the stimulus rate became high (Fig. 1 and Table 1). The decrease in the P20 HFO amplitude under a high stimulus rate was significant. The number of HFO peaks also tended to decrease under the high stimulus conditions. The HFOs/N20 amplitude ratio was significantly smaller for 12.3 Hz stimulation than that after 3.3 Hz stimulation. A similar tendency was shown for HFOs/P20 or P25 amplitude ratios, but they did not reach the significance level (P ¼ 0:0796 and 0.1797 for HFOs/P20 and P25, respectively). A selective amplitude decrease in the later part of the HFOs after the N20 peak was found in one patient (Fig. 2). Three patients showed an amplitude decrease in both the early and later parts of N20 HFOs by the increase in the stimulus rate as shown in Fig. 1. One residual patient showed no decrease in the early and later N20 HFO parts.
4. Discussion The present results contrast with those of a previous study that showed stable HFOs among the 1.5, 3, 6, and 9 Hz stimulation rates, while the corresponding N20 amplitude was significantly decreased at higher stimulus rates (Gobbele´ et al., 1999). The difference between previous and the present findings may be due to the different vigilance levels of the subjects during the examination (Hashimoto et al., 1999; Gobbele´ et al., 2000; Halboni et al., 2000). The present study required a short time to obtain the SEPs of two different stimulus rates, and it was easy to maintain the subject at the same vigilance level. It appears to be difficult, however, to avoid changing the vigilance level when a large number of averages were required to obtain the SEP data. When the subject is sleepy during the lower stimulus rates, the HFO amplitude would be decreased (Hashimoto et al., 1999; Gobbele´ et al., 2000; Halboni et al., 2000), but when the subject is alert due to the unpleasant sensation induced by the higher stimulus rate, the HFO amplitude would be increased despite the essential decrease in the HFO by the increased stimulus rate (Hashimoto et al., 1999; Klostermann et al., 1999; Gobbele´ et al., 2000; Halboni et al., 2000). As a result, HFOs between different stimulus rates may show no apparent change. The present study demonstrated that the N20 and other cortical bursts are lost or diminished at a stimulus rate of
E. Urasaki et al. / Clinical Neurophysiology 113 (2002) 1794–1797
12.3 Hz, that could be related to a characteristic of the long recovery time constant of a slow calcium influx produced by the hyperpolarized thalamocortical relay cells (Klostermann et al., 1999). The deeply situated thalamocortical radiation fiber, or thalamocortical relay cells in the thalamus, however, may not be a main contributing factor of the cortically recorded HFOs in this study. HFOs were not widely distributed or recorded from the cortex, but they were recorded at a restricted brain area (Maegaki et al., 2000; Kojima et al., 2001), against the characteristics of deeply localized generator sources. In addition, the phase reversal of HFOs between the corresponding N20 and P20 strongly supports the cortical origin of HFOs close to area 3b (Ozaki et al., 1998; Kojima et al., 2001). Superficial preterminal thalamocortical radiation could be localized near or in the cortex, but such a preterminal activity appears to be more resistant to the higher stimulus rate than the post-synaptic primary cortical responses. g-Aminobutyric acid (GABA)ergic inhibitory interneurons, or pyramidal chattering cells if they exist in the sensory cortex, or other intracortical synaptic activities may be the other possible generators of the HFOs (Hashimoto et al., 1999; Klostermann et al., 1999; Mackert et al., 2000). Although it is uncertain whether the study by Gobbele et al. (1999) focused on the early HFO peak, two burst generators for the cortical HFOs were initially proposed by Klostermann et al. (1999)); an early burst with a resistant property, and a late burst with a pronounced amplitude decrease, to increasing the stimulus rates. Careful observation of individual findings or measurements disclosed one such case in the present study (Fig. 2). The amplitude of early HFOs and N20 before its peak did not change, but the later HFOs were decreased by the high stimulus rate. It could be interpreted that the decreased HFO amplitude after the N20 peak was associated with an increased negativity area of the N20 component in this case (Fig. 2). This finding could be a kind of reciprocal relation between N20 and HFOs as reported by Hashimoto et al., 1999). Another interpretation of this SEP change could be an amplitude reduction of the following positivity after N20 (Fig. 2). A selective decrease of the later part of HFOs after the N20 peak may be associated with the decrease in cortical positivity, that appeared to be a contrast finding of enhanced HFOs associated with increased cortical positivity in aged subjects and patients with Parkinson’s disease or myoclonic epilepsy (Mochizuki et al., 1999; Nakano and Hashimoto, 1999). The present study not only confirmed the characteristic that the decrease in amplitude with increasing stimulus rates is stronger for the HFOs than for N20, as reported by Klostermann et al., 1999, but also adds a new finding that similar
1797
characteristics exist in the HFOs on the P20 as well as P25, area 1, cortical responses. A specific common mechanism appears to be present in area 3b and area 1 for generation of HFOs, but further studies are required to clarify whether the area 1 HFOs are closely related to area 3b HFOs (Maegaki et al., 2000) or whether they are independent of each other (Klostermann et al., 1999; Mochizuki et al., 1999; Nakano and Hashimoto, 1999).
References Gobbele´ R, Buchner H, Curio G. High-frequency (600 Hz) SEP activities originating in the subcortical and cortical human somatosensory system. Electroenceph clin Neurophysiol 1998;108:182–189. Gobbele´ R, Buchner H, Scherg M, Curio G. Stability of high-frequency (600 Hz) components in human somatosensory evoked potentials under variation of stimulus rate – evidence for a thalamic origin. Clin Neurophysiol 1999;110:1659–1663. Gobbele´ R, Waberski TD, Kuelkens S, Sturm W, Curcio G, Buchner H. Thalamic and cortical high-frequency (600 Hz) somatosensory-evoked potential (SEP) components are modulated by slight arousal changes in awake subjects. Exp Brain Res 2000;133:506–513. Hahn JF, Lu¨ ders H. Placement of subdural grid electrodes at the Cleveland clinic. In: Engel JJ, editor. Surgical treatment of epilepsies, New York, NY: Raven Press, 1987. pp. 621–627. Halboni P, Kaminski R, Gobbele´ R, Zu¨ chner S, Waberski TD, Hermann CS, To¨ pper R, Buchner H. Sleep stage dependent changes of the highfrequency part of the somatosensory evoked potentials at the thalamus and cortex. Clin Neurophysiol 2000;111:2277–2284. Hashimoto I, Kimura T, Fukushima T, Iguchi Y, Sauto Y, Terasaki O, Sakuma K. Reciprocal modulation of somatosensory evoked N20m primary response and high-frequency oscillations by interference stimulation. Clin Neurophysiol 1999;110:1445–1451. Klostermann F, Nolte G, Curio G. Multiple generators of 600 Hz wavelets in human SEP unmasked by varying stimulus rates. NeuroReport 1999;10:1625–1629. Kojima Y, Uozumi T, Akamatsu N, Matsunaga K, Urasaki E, Tsuji S. Somatosensory evoked high frequency oscillations recorded from subdural electrodes. Clin Neurophysiol 2001;112:2261–2264. Lu¨ ders HO, Awad I. Conceptual considerations. In: Lu¨ ders H, editor. Epilepsy surgery, New York, NY: Raven Press, 1991. pp. 83–86. Mackert BM, Weisenbach S, Nolte G, Curio G. Rapid recovery (20 ms) of human 600 Hz electroencephalographic wavelets after double stimulation of sensory nerves. Neurosci Lett 2000;286:83–86. Maegaki Y, Najm I, Terada K, Morris HH, Bingaman WE, Kohaya N, Takenobu A, Kadonaga Y, Lu¨ ders H. Somatosensory evoked highfrequency oscillations recorded directly from the human cerebral cortex. Clin Neurophysiol 2000;111:1916–1926. Mochizuki H, Ugawa Y, Machii K, Terao Y, Hanajima R, Furubayashi T, Uesugi H, Kanazawa I. Somatosensory evoked high-frequency oscillation in Parkinson’s disease and myoclonus epilepsy. Clin Neurophysiol 1999;110:185–191. Nakano S, Hashimoto I. The later part of high-frequency oscillations in human somatosensory evoked potentials is enhanced in aged subjects. Neurosci Lett 1999;276:83–86. Ozaki I, Suzuki C, Yaegashi Y, Baba M, Matsunaga M, Hashimoto I. Highfrequency oscillations in early cortical somatosensory evoked potentials. Electroenceph clin Neurophysiol 1998;108:536–542.
The effects of stimulus rates on high frequency oscillations of median nerve somatosensory-evoked potentials – direct recording study from the human cerebral cortex Eiichirou Urasaki a,*, Tetsuya Genmoto a, Naoki Akamatsu b, Shin-ichi Wada a, Akira Yokota a a
Department of Neurosurgery, School of Medicine, University of Occupational and Environmental Health, Iseigaoka 1-1, Yahata Nishi-ku, Kitakyushu City 807-8555, Japan b Department of Neurology, School of Medicine, University of Occupational and Environmental Health, Iseigaoka 1-1, Yahata Nishi-ku, Kitakyushu City 807-8555, Japan Accepted 21 August 2002
Abstract Objectives: To study the effects of different stimulus rates on high-frequency oscillations (HFOs) of somatosensory-evoked potentials (SEPs), we recorded median nerve SEPs directly from the human cerebral cortex. Methods: SEPs were recorded from subdural electrodes in 5 patients with intractable epilepsy, under the conditions of low (3.3 Hz) and high (12.3 Hz) stimulus rates. Results: Increased stimulus rates to the median nerve from 3.3 to 12.3 Hz showed a pronounced amplitude reduction of HFOs when compared with the primary N20–P20, area 3b, and P25, area 1, responses. Conclusions: HFOs were more sensitive to a high stimulus rate than the primary cortical responses, suggesting that the post-synaptic intracortical activities may greatly contribute to the HFO generation. q 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Somatosensory-evoked potentials; High-frequency oscillations; Direct recording; Stimulus rates
1. Introduction Primary cortical components of somatosensory-evoked potentials (SEPs) are known to have high-frequency oscillations (HFOs) that are clearly exposed by setting low bandpass filters greater than 300–500 Hz (Gobbele´ et al., 1998; Gobbele´ et al., 1999; Hashimoto et al., 1999; Klostermann et al., 1999; Nakano and Hashimoto, 1999; Gobbele´ et al., 2000; Halboni et al., 2000; Mackert et al., 2000; Maegaki et al., 2000; Kojima et al., 2001). The characteristics and significance of the HFOs have been investigated but there are some uncertainties and controversies. The stability of HFOs on cortical N20, for example, under various stimulus rates of the median nerve was reported, suggesting a presynaptic origin such as thalamo-cortical projection fibers (Gobbele´ et al., 1998; Gobbele´ et al., 1999). The intracortical activities were, on the other hand, proposed as generators based on the finding that the amplitude decrease was emphasized more in the HFOs than the corresponding cortical N20 when the stimulus rate was increased (Klostermann * Corresponding author. Tel.: 181-93-603-1611; fax: 181-93-691-8755. E-mail address: [email protected] (E. Urasaki).
et al., 1999). The present study provides direct recording data from the human cerebral cortex demonstrating a pronounced amplitude decrease in HFOs when compared with N20 by increasing the stimulus rates, and adds new findings that suggest that the same is true for cortical P20 and P25, an area 1 component. 2. Patients and methods Five male patients with medically intractable epilepsy aged between 16 and 42 years (30.6 ^ 10.3; mean ^ SD) were studied. Subdural electrodes were implanted intracranially to evaluate seizure localization and functional brain mapping prior to surgical treatment. The necessary number of intracranial electrodes was placed near the expected areas of epileptogenic, irritative, or symptomatogenic zones (Hahn and Lu¨ders, 1987; Lu¨ders and Awad, 1991). Intracranial electrode placement on areas superior to the Sylvian fissure including the central sulcus was required in all patients. Informed consent was obtained from all patients. This study was approved by the institutional Review Board of the University of Occupational and Environmental
1388-2457/02/$ - see front matter q 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S13 88- 2457(02)0029 1-2
CLINPH 2002077
E. Urasaki et al. / Clinical Neurophysiology 113 (2002) 1794–1797
1795
Table 1 Comparison of HFOs and the primary cortical response (N20, P20, and P25) between the conditions of 3.3 and 12.3 Hz stimulations a 3.3 Hz (mean ^ SD)
12.3 Hz (mean ^ SD)
N20 amplitude (mV) (n ¼ 5) HFO number of negative peaks HFO amplitudes (mV) HFOs/N20 amplitude ratio (%)
14.6 ^ 9.13 4.0 ^ 2.2 2.80 ^ 1.70 26.3 ^ 21.1
16.03 ^ 12.07 2.6 ^ 2.1 1.37 ^ 1.77 11.2 ^ 13.2*
P20 amplitude (mV) (n ¼ 5) HFO number of positive peaks HFO amplitudes (mV) HFOs/P20 amplitude ratio (%)
15.15 ^ 10.61 4.8 ^ 2.3 5.9 ^ 6.52 30.8 ^ 16.4
12.27 ^ 7.10 3.6 ^ 2.1 2.64 ^ 3.06* 17.5 ^ 12.3
P25 amplitude (mV) (n ¼ 2) HFO number of positive peaks HFO amplitudes (mV) HFOs/P25 amplitude ratio (%)
22.58 ^ 2.47 6.5 ^ 2.1 6.13 ^ 2.30 26.7 ^ 7.3
24.32 ^ 8.33 4.0 ^ 2.8 1.90 ^ 0.17 8.2 ^ 2.1
a
SD ¼ standard deviation; *P , 0:05.
Health. Neurological examinations including the somatosensory system showed normal findings in all patients except for their seizures. MRI revealed no abnormality in the postcentral gyrus or thalamus in any patient. The grid or strip electrodes in the perirolandic region were used for SEP recordings. The locations of the implanted subdural electrodes in the vicinity of the sensorimotor area were confirmed by a cortical stimulation study. The diameters of the recording electrodes in the subdural grid were 4 mm, and the interelectrode distance from center to center was 10 mm. To construct a functional brain map, biphasic 0.3 ms square wave electrical stimuli, 50/s frequency and 3–5 s duration were applied to the brain through the subdural electrodes. Each electrode was stimulated monopolarly by setting a common reference electrode chosen from the other intracranial electrodes that gave no response to electrical stimuli. The strength of the electrical current was 1–15 mA, or up to the appearance of the after discharge. SEPs were recorded after the functional mapping study. The reference electrode was placed on the ear contralateral to the side of stimulation. SEP was obtained using 0.2 ms square wave electrical pulses delivered transcutaneously to the median nerve at the wrist contralateral to the side of intracranial recording. Saddle-typed bipolar electrodes, a cathode 3 cm proximal to the anode, were used. The period of analysis was 50–100 ms. Stimulus intensity was adjusted to produce a thumb twitch. The filter setting was 10– 1000 Hz with an analysis time of 50–100 ms. At least two averages of 500–1000 responses were obtained to confirm reproducibility, and two sets of responses were averaged using a computer for later analysis. To maintain a stable vigilance level during the test, the patients were instructed to stay awake, and were interviewed about their vigilance level before and after every recording session. HFOs were extracted from underlying cortical responses by the digital filtering of the wide-band (10–1000) recorded
responses through a band-pass of 500–1000 Hz, using a Viking IV computer system (Nicolet). The amplitude of the primary cortical response was measured as the vertical distance between the onset and the peak of the primary cortical response. HFOs identified between the onset and endpoint of primary cortical responses (N20, P20, and P25) were analyzed. The number of negative or positive peaks of HFOs was counted between the onset and endpoint of respective N20, P20 and P25 responses. The amplitudes of the HFOs were averaged with the values measured as the vertical distances between two successive peaks. It was reported that the peak of N20 or P20 tended to coincide with the peak of maximal negative or positive HFOs (Ozaki et al., 1998; Maegaki et al., 2000). It may, therefore, be reasonable to measure the negative peaks or positive peaks to analyze the respective HFOs of N20 or P20 and P25. There is, however, no practical difference between the analysis of negative and positive peaks because the amplitude was measured from peak to peak. All negative HFO components were measured from the preceding positive peak to the following negative peak and then averaged for the amplitude calculation of HFOs superimposed on N20. Similarly, all positive components were measured and averaged for HFOs on P20 and P25. HFOs/primary cortical response amplitude ratio was calculated and revealed as a percentage. Two-way analysis of variance (ANOVA), multiple comparison test using rank values of variables, and nonparametric Wilcoxon sum rank test was used for statistical analysis, and a P value less than 0.05 was considered statistically significant.
3. Results Table 1 shows the comparison of the primary cortical response and HFO parameters between 3.3 and 12.3 Hz
1796
E. Urasaki et al. / Clinical Neurophysiology 113 (2002) 1794–1797
Fig. 1. Somatosensory evoked potentials (SEPs) after left median nerve stimulation, recorded from the subdural electrodes implanted in a 41year-old male patient with intractable epilepsy. Cortical SEPs such as P25, area 1 component, N20/P20, area 3b components, and their corresponding HFOs are shown. Note the marked decrease in HFOs despite the preserved primary cortical components at the higher stimulus rate. Electrical stimulation of electrodes 1, 2, and 3 elicited thumb motor, thumb sensory, and finger motor responses, respectively.
stimulation in 5 subjects (Table 1). The cortical N20–P20 area 3b components were recorded from all 5 patients, and the P25 area 1 component from two patients. In all the 5 patients, HFOs on N20 and P20 were obtained at two different stimulus rates, and the high stimulus rate either induced loss or decreased the HFO amplitudes in 4 of the 5 patients (Fig. 1). One patient showed no decrease in HFOs on the corresponding N20. The peripheral N9 showed a slight decrease in amplitude by the high stimulus rate without statistical significance (5.53 ^ 2.27 mV at 3.3 Hz and 4.49 ^ 0.82 mV at 12.3 Hz). Two-way ANOVA and multi-
Fig. 2. SEPs elicited by left median nerve stimulation recorded from a 27year-old male patient with intractable epilepsy. Primary cortical N20 response and corresponding HFOs for the 3.3 and 12.3 Hz stimulus rates are superimposed. The early part of the HFOs before the N20 peak is relatively preserved, but the later part of the HFOs after the N20 peak is strongly decreased. Electrical stimulation of this recording site elicited a hand sensory response.
ple comparison test using rank values of variables showed significant changes in the HFO amplitudes (P ¼ 0:044) and the HFOs/primary cortical response ratios (P ¼ 0:035) between the 3.3 and 12.3 Hz stimulus conditions, suggesting a safeguard against errors due to multiple comparisons. There were no significant changes in the N20, P20 and P25 amplitudes and the HFO number between these two different stimuli. Overall, HFOs on N20–P20 and P25 were more sensitive to change in the stimulus rate than the primary cortical responses. Mean HFO amplitudes were decreased while all primary responses showed a relatively slight increase in amplitude when the stimulus rate became high (Fig. 1 and Table 1). The decrease in the P20 HFO amplitude under a high stimulus rate was significant. The number of HFO peaks also tended to decrease under the high stimulus conditions. The HFOs/N20 amplitude ratio was significantly smaller for 12.3 Hz stimulation than that after 3.3 Hz stimulation. A similar tendency was shown for HFOs/P20 or P25 amplitude ratios, but they did not reach the significance level (P ¼ 0:0796 and 0.1797 for HFOs/P20 and P25, respectively). A selective amplitude decrease in the later part of the HFOs after the N20 peak was found in one patient (Fig. 2). Three patients showed an amplitude decrease in both the early and later parts of N20 HFOs by the increase in the stimulus rate as shown in Fig. 1. One residual patient showed no decrease in the early and later N20 HFO parts.
4. Discussion The present results contrast with those of a previous study that showed stable HFOs among the 1.5, 3, 6, and 9 Hz stimulation rates, while the corresponding N20 amplitude was significantly decreased at higher stimulus rates (Gobbele´ et al., 1999). The difference between previous and the present findings may be due to the different vigilance levels of the subjects during the examination (Hashimoto et al., 1999; Gobbele´ et al., 2000; Halboni et al., 2000). The present study required a short time to obtain the SEPs of two different stimulus rates, and it was easy to maintain the subject at the same vigilance level. It appears to be difficult, however, to avoid changing the vigilance level when a large number of averages were required to obtain the SEP data. When the subject is sleepy during the lower stimulus rates, the HFO amplitude would be decreased (Hashimoto et al., 1999; Gobbele´ et al., 2000; Halboni et al., 2000), but when the subject is alert due to the unpleasant sensation induced by the higher stimulus rate, the HFO amplitude would be increased despite the essential decrease in the HFO by the increased stimulus rate (Hashimoto et al., 1999; Klostermann et al., 1999; Gobbele´ et al., 2000; Halboni et al., 2000). As a result, HFOs between different stimulus rates may show no apparent change. The present study demonstrated that the N20 and other cortical bursts are lost or diminished at a stimulus rate of
E. Urasaki et al. / Clinical Neurophysiology 113 (2002) 1794–1797
12.3 Hz, that could be related to a characteristic of the long recovery time constant of a slow calcium influx produced by the hyperpolarized thalamocortical relay cells (Klostermann et al., 1999). The deeply situated thalamocortical radiation fiber, or thalamocortical relay cells in the thalamus, however, may not be a main contributing factor of the cortically recorded HFOs in this study. HFOs were not widely distributed or recorded from the cortex, but they were recorded at a restricted brain area (Maegaki et al., 2000; Kojima et al., 2001), against the characteristics of deeply localized generator sources. In addition, the phase reversal of HFOs between the corresponding N20 and P20 strongly supports the cortical origin of HFOs close to area 3b (Ozaki et al., 1998; Kojima et al., 2001). Superficial preterminal thalamocortical radiation could be localized near or in the cortex, but such a preterminal activity appears to be more resistant to the higher stimulus rate than the post-synaptic primary cortical responses. g-Aminobutyric acid (GABA)ergic inhibitory interneurons, or pyramidal chattering cells if they exist in the sensory cortex, or other intracortical synaptic activities may be the other possible generators of the HFOs (Hashimoto et al., 1999; Klostermann et al., 1999; Mackert et al., 2000). Although it is uncertain whether the study by Gobbele et al. (1999) focused on the early HFO peak, two burst generators for the cortical HFOs were initially proposed by Klostermann et al. (1999)); an early burst with a resistant property, and a late burst with a pronounced amplitude decrease, to increasing the stimulus rates. Careful observation of individual findings or measurements disclosed one such case in the present study (Fig. 2). The amplitude of early HFOs and N20 before its peak did not change, but the later HFOs were decreased by the high stimulus rate. It could be interpreted that the decreased HFO amplitude after the N20 peak was associated with an increased negativity area of the N20 component in this case (Fig. 2). This finding could be a kind of reciprocal relation between N20 and HFOs as reported by Hashimoto et al., 1999). Another interpretation of this SEP change could be an amplitude reduction of the following positivity after N20 (Fig. 2). A selective decrease of the later part of HFOs after the N20 peak may be associated with the decrease in cortical positivity, that appeared to be a contrast finding of enhanced HFOs associated with increased cortical positivity in aged subjects and patients with Parkinson’s disease or myoclonic epilepsy (Mochizuki et al., 1999; Nakano and Hashimoto, 1999). The present study not only confirmed the characteristic that the decrease in amplitude with increasing stimulus rates is stronger for the HFOs than for N20, as reported by Klostermann et al., 1999, but also adds a new finding that similar
1797
characteristics exist in the HFOs on the P20 as well as P25, area 1, cortical responses. A specific common mechanism appears to be present in area 3b and area 1 for generation of HFOs, but further studies are required to clarify whether the area 1 HFOs are closely related to area 3b HFOs (Maegaki et al., 2000) or whether they are independent of each other (Klostermann et al., 1999; Mochizuki et al., 1999; Nakano and Hashimoto, 1999).
References Gobbele´ R, Buchner H, Curio G. High-frequency (600 Hz) SEP activities originating in the subcortical and cortical human somatosensory system. Electroenceph clin Neurophysiol 1998;108:182–189. Gobbele´ R, Buchner H, Scherg M, Curio G. Stability of high-frequency (600 Hz) components in human somatosensory evoked potentials under variation of stimulus rate – evidence for a thalamic origin. Clin Neurophysiol 1999;110:1659–1663. Gobbele´ R, Waberski TD, Kuelkens S, Sturm W, Curcio G, Buchner H. Thalamic and cortical high-frequency (600 Hz) somatosensory-evoked potential (SEP) components are modulated by slight arousal changes in awake subjects. Exp Brain Res 2000;133:506–513. Hahn JF, Lu¨ ders H. Placement of subdural grid electrodes at the Cleveland clinic. In: Engel JJ, editor. Surgical treatment of epilepsies, New York, NY: Raven Press, 1987. pp. 621–627. Halboni P, Kaminski R, Gobbele´ R, Zu¨ chner S, Waberski TD, Hermann CS, To¨ pper R, Buchner H. Sleep stage dependent changes of the highfrequency part of the somatosensory evoked potentials at the thalamus and cortex. Clin Neurophysiol 2000;111:2277–2284. Hashimoto I, Kimura T, Fukushima T, Iguchi Y, Sauto Y, Terasaki O, Sakuma K. Reciprocal modulation of somatosensory evoked N20m primary response and high-frequency oscillations by interference stimulation. Clin Neurophysiol 1999;110:1445–1451. Klostermann F, Nolte G, Curio G. Multiple generators of 600 Hz wavelets in human SEP unmasked by varying stimulus rates. NeuroReport 1999;10:1625–1629. Kojima Y, Uozumi T, Akamatsu N, Matsunaga K, Urasaki E, Tsuji S. Somatosensory evoked high frequency oscillations recorded from subdural electrodes. Clin Neurophysiol 2001;112:2261–2264. Lu¨ ders HO, Awad I. Conceptual considerations. In: Lu¨ ders H, editor. Epilepsy surgery, New York, NY: Raven Press, 1991. pp. 83–86. Mackert BM, Weisenbach S, Nolte G, Curio G. Rapid recovery (20 ms) of human 600 Hz electroencephalographic wavelets after double stimulation of sensory nerves. Neurosci Lett 2000;286:83–86. Maegaki Y, Najm I, Terada K, Morris HH, Bingaman WE, Kohaya N, Takenobu A, Kadonaga Y, Lu¨ ders H. Somatosensory evoked highfrequency oscillations recorded directly from the human cerebral cortex. Clin Neurophysiol 2000;111:1916–1926. Mochizuki H, Ugawa Y, Machii K, Terao Y, Hanajima R, Furubayashi T, Uesugi H, Kanazawa I. Somatosensory evoked high-frequency oscillation in Parkinson’s disease and myoclonus epilepsy. Clin Neurophysiol 1999;110:185–191. Nakano S, Hashimoto I. The later part of high-frequency oscillations in human somatosensory evoked potentials is enhanced in aged subjects. Neurosci Lett 1999;276:83–86. Ozaki I, Suzuki C, Yaegashi Y, Baba M, Matsunaga M, Hashimoto I. Highfrequency oscillations in early cortical somatosensory evoked potentials. Electroenceph clin Neurophysiol 1998;108:536–542.