- Email: [email protected]
Magnetically recorded oscillatory responses to luminance stimulation in man1
Electroencephalography and clinical Neurophysiology 104 (1997) 91–95
Magnetically recorded oscillatory responses to luminance stimulation in man1 Luisa Lopez a, Walter G. Sannita b , c , d ,* a
Institute for Advanced Biomedical Technologies, University of Chieti, Chieti, Italy Center for Neuroactive Drugs, Department of Motor Science and Rehabilitation - Neurophysiopathology, University of Genova, Genova, Italy c Center for Cerebral Neurophysiology, National Council of Research, Genova, Italy d Department of Psychiatry and Behavioral Sciences, State University of New York, Stony Brook, NY, USA
b
Accepted for publication: 4 November 1996
Abstract Fast-frequency (ca. 100–110 Hz) oscillatory potentials superimposed on waves N2 and P2 of conventional broad-band VEP were magnetically recorded in man from occipital locations in response to monocular transient flash stimulation with full-field flashes (3.5 cd⋅s⋅m − 2 intensity) and in spots (1, 1.5, or 2.0 cm in diameter). These oscillations proved replicable between- and within-subject and were phase-locked to retinal oscillatory potentials, with maximum correlation at approximately 35 ms and mean delay (as measured between the first measurable peaks) of 27.4 ± 1.6 ms. When stimuli were in spots at increasing eccentricity (5, 15, or 25°) from foveal fixation, broad-band VEP were recorded regardless of diameter and eccentricity of spot, whereas oscillatory responses were not detectable at eccentricity of, or greater than, 15°. This observation suggests that broad-band VEP and the oscillatory response are generated by (partly) distinct neuronal populations and/or functional arrangements and that there is some functional connection between cortical oscillatory responses and stimulus-related events triggered in central retina. 1997 Elsevier Science Ireland Ltd. All rights reserved Keywords: Visual evoked potentials; Oscillatory responses; Luminance stimulation; Retinal eccentricity; Neuromagnetic fields; MEG
1. Introduction Under proper stimulation and recording conditions, short bursts of fast frequency (ca. 100–140 Hz) rhythmic oscillations have been recorded in response to flash stimulation at retinal level in animals and man (Wachtmeister, 1987), from lateral geniculate nucleus (LGN) of monkeys (Doty and Kimura, 1963; Schroeder et al., 1992), and from visual cortex of cat, monkey and man (Hughes and Mazurowski, 1962; Doty and Kimura, 1963; Steriade, 1968; Ducati et al., 1988). Oscillatory responses in the same
* Corresponding author. Center for Neuropsychoactive Drugs, Department of Motor Sciences, University of Genova, I-16132 Genova, Italy. Tel.: +39 10 3537464; fax: +39 10 3537699; e-mail: [email protected] 1 Presented, in part, at the Symposium ‘Oscillatory Event-related Brain Dynamics’, Mu¨nster, Germany, September 2–6, 1993 and the Ninth Annual Meeting of the American Academy of Clinical Neurophysiology, Boston, MA, USA, July 14–16, 1994.
0168-5597/97/$17.00 1997 Elsevier Science Ireland Ltd. All rights reserved PII S0921-884X(96)9652 9-7
frequency range are superimposed on the early waves of human flash-evoked potentials (VEP) recorded from scalp (Cracco and Cracco, 1978; Whittaker and Siegfried, 1983). These oscillatory events are phase-locked to stimulus and become apparent (or are enhanced) when the signal is high-pass filtered with cut-off at approximately 70–90 Hz, but do not depend on narrow-band filter distortion of the electroretinogram (ERG) or VEP (Tsuchida et al., 1973). The existing differences between retinal and scalp-recorded oscillations stand against volume-conducted contamination from retinal sources (Whittaker and Siegfried, 1983) and significant contributions from LGN to cortical responses are unlikely (Schroeder et al., 1992). The sources of retinal oscillatory potentials appear separated from those of ERG and conceivably involve activation of neuronal loops in inner retina (Karwoski and Kawasaki, 1991). Accordingly, scalp-recorded oscillatory responses are suggested to reflect cortical phenomena that are parallel to, and may have (partially) different origin than, broad-band VEP (Whittaker and Siegfried,
EEP 96529
92
L. Lopez, W.G. Sannita / Electroencephalography and clinical Neurophysiology 104 (1997) 91–95
1983). Neuromagnetic, reference-independent recording techniques have been used to support this hypothesis. 2. Materials and methods Eight healthy volunteers (25–35 years old) were recruited after excluding any history or clinical evidence of ocular, neurological or systemic pathology, long-term treatment with, or abuse of, neuroactive drugs. In all cases visual acuity was better than 18/20, with negligible ametropia. Routine EEG, ERG and VEP were within normal limits. Flash stimuli were generated by a Grass PS22 stimulator and delivered via an optical system into the shielded room. Stimulation was monocular, at 0.6 Hz; stimuli were full-field (at 3.5 cd⋅s⋅m − 2 intensity) or in circular spots (with diameter of 1, 1.5, or 2.0 cm, respectively) located at increasing eccentricity in the left or right upper quadrant of visual field (5, 15, 25° from foveal fixation). A 70 dB masking white noise was applied to minimize the effect of noise artifact from the strobe. ERG recordings were performed via Ag/AgCl dermal electrodes positioned on the lower orbital margin and referred to linked mastoids; the ground electrode was at Cz. Neuromagnetic recordings (MEG) were performed with a 28-channel system (16 gradiometers, 9 magnetometers, 3 noise cancellation coils; tail diameter 16 cm) in a magnetically shielded room (Vacuumschmeltze GmbH). The head position with respect to the magnetic probe was kept constant in all stimulus conditions by referring to 3 magnetic coils placed over anatomical landmarks and checked prior to recordings. Two probe positions centered over the parietal-occipital region (P3/O1 and P4/O2 of the 10–20 International System) contralateral to the stimulated eye and the inion (Oz) were used for recordings. The EEG was monitored during and between recordings and subjects were not allowed to drowse. Continuous electric and magnetic signal acquisition was performed in a 0.16–250 Hz bandwidth (high-pass filter: 6 dB/octave; low-pass, Bessel-type filter: 48 dB/octave) at 1 kHz sampling rate. Oscillatory potentials were high-pass filtered at 80 Hz. Averages were on consecutive 60–70 epochs free of artifacts. The ethical principles of the Declaration of Helsinki (1964) by the World Medical Association concerning human experimentation were followed. The study was performed at the Institute of Electronics of the Solid State (the National Council of Research, Rome), with approval, and all volunteers signed an informed consent.
3. Results In all recordings after full-field flash stimulation, oscillatory responses appeared superimposed on the ERG as well as on waves N2 and P2 of the VEP and could be dissociated from the underlying broad-band response by high-pass filtering, without apparent filter distortion. The oscillatory responses proved replicable within-subject and consistent across subjects (Fig. 1). The average duration and maximum amplitude of the oscillations recorded at parietal-occipital locations were 40–50 ms and 60–150 fT, respectively, while the duration of retinal oscillatory potentials was 30–35 ms. The oscillatory fields recorded magnetically at parietal-occipital locations were correlated with the retinal oscillatory responses, with maximum correlation at a delay of approximately 35 ms (Fig. 1D); the mean delay between the onset of retinal and occipital responses (as measured between the first measurable peaks) was 27.4 ± 1.6 ms. ERG or retinal oscillatory potentials were not recorded after stimulation in spots, consistent with the stimulated area/response function of retinal responses to luminance stimuli. Broad-band VEP were conversely recorded under these stimulus conditions regardless of spot diameter and at any eccentricity in the 5–25° range. The VEP amplitude decreased with decreasing diameter of spot stimuli in the 1–2 cm range, but latencies were substantially unaffected. Oscillatory responses to spot stimulation were recorded from occipital locations at 5° eccentricity of stimulus (apparently with a limited effect of spot diameter), but were not observed at eccentricities of, or greater than, 15° (Fig. 2). 4. Discussion As opposed to the short delay between retinal and LGN oscillatory potentials in monkeys (Schroeder et al., 1992), the latency delay in man between retinal and parietal-occipital oscillations suggests complex operations and some higher degree of cellular interaction. In this regard, substantial cortical contributions in the generation of nearfield oscillatory responses recorded at parietal-occipital level are plausible. This hypothesis is based on theoretical and experimental evidence that only near-field currents related to intracellular activity are recordable by reference-free magnetic methods (Okada et al., 1987), and is supported by the observation of parietal-occipital oscilla-
Fig. 1. (A) Comparison of the oscillatory responses to full-field luminance stimulation as recorded electrically at retinal level (top; ‘retina’) and magnetically at occipital locations (bottom; ‘MEG’) in 4 healthy subjects. Superimposed recordings from 5 magnetic channels at a location corresponding to the occipital area contralateral to the stimulated eye are shown for each subject. Calibration as indicated. (B) Effect of high-pass filtering at increasing cut-offs on the response to full-field stimulation magnetically recorded at occipital location from a fifth subject. Note that oscillatory responses appear superimposed on conventional, broad-band (e.g. 1–200 Hz) VEP and are enhanced by narrow-band filtering, without apparent filter distortion. (C) Comparison on an expanded time scale between retinal and magnetically-recorded oscillatory responses. (D) Correlogram (computed over time and fitted to the shortest period within the selected 80–150 Hz frequency band) between the retinal oscillatory potential shown in (C) and the responses from 5 adjacent scalp locations of one subject. Note the peak correlation at approximately 35 ms.
L. Lopez, W.G. Sannita / Electroencephalography and clinical Neurophysiology 104 (1997) 91–95
93
94
L. Lopez, W.G. Sannita / Electroencephalography and clinical Neurophysiology 104 (1997) 91–95
Fig. 2. Effects of diameter and eccentricity of spot stimuli on conventional broad-band VEP and narrow-band oscillatory response. Magnetic recording; 4 superimposed adjacent channels. The stimulus is indicated by vertical bars. Note that VEP are recordable at spot diameter as small as 1 cm and eccentricities of stimulus up to 25°, whereas the oscillatory response is not evident at eccentricities of 15° or more and approaches the noise level with spot diameter of 1 cm. No retinal oscillatory events were recorded after spot stimulation.
tory responses to focal (spot) stimuli eliciting no retinal response as well as by the described grading off a few millimeters outside the LGN of the oscillations originating from lamina 6 of LGN, without significant contributions to the scalp-recorded VEP (Schroeder et al., 1992). Further to this, the oscillatory components of cortical primary response of cat are suggested to be independent of responses in LGN or optic tract because of their enhancement in conditions of increased cortical excitability (Steriade, 1968). However, visual evoked potentials with short latency and suggested subcortical origin but still unidentified sources have been described in man (Harding and Rubinstein, 1980) and pre-cortical components of the oscillatory response cannot be excluded. Cortical circuitry may tentatively be held responsible for the delay between retinal and ‘cortical’ oscillatory macropotentials. Oscillatory events occurring at singleand multi-unit level over a relatively wide range of frequencies (ca. 20–100 Hz) are suggested to mediate in the synchronization of relatively large populations of neurons (Bressler, 1990; Singer, 1993; Jefferys et al., 1996), which is a prerequisite for any neural activity to be driven into patterns of oscillation recordable via macroelectrodes. Membrane oscillations recorded intracellularly in the visual cortex of cat depend on the physical properties of
stimulus (Gray et al., 1989; Sillito et al., 1994) and are believed to reflect cortical mechanisms operating in conjunction with intrinsic neuronal membrane properties and thalamocortical feedback loops (Gray et al., 1989; Llinas, 1992; Singer, 1993; Sillito et al., 1994; Jefferys et al., 1996). These mechanisms operate in sensory information processing and conceivably play a role in the generation of both oscillatory activities at cellular level and local-field potentials. The recording of broad-band VEP in the absence of oscillatory responses after stimulation with spots at eccentricities equal to, or greater than, 15° suggests that the generation of broad-band VEP and oscillatory responses depends on the activation of (partly) distinct neuronal populations and/or functional arrangements eventually overlapping in MEG space, and further indicates visual information processing through parallel subsystems. It should be noted in this regard that oscillatory responses to non-visual stimuli have also been described as being distinct from the underlying broad-band evoked potentials (Singer, 1993; Pantev et al., 1994for references). Unlike retinal oscillatory potentials and conventional broad-band VEP, the near-field oscillatory responses at parietal-occipital locations appear to depend on events triggered through central rather than reflect diffuse retinal stimula-
L. Lopez, W.G. Sannita / Electroencephalography and clinical Neurophysiology 104 (1997) 91–95
tion. Substantial involvement of geniculocalcarine pathway is plausible. The oscillatory responses to luminance and those recorded after pattern-reversal stimulation of 9° central retina, with ‘tuning’ at approximately 5 cycles/ degree (Sannita et al., 1995) may therefore reflect mechanisms of information processing and transfer within the visual system that are in part common to the two paradigms of visual stimulation, in spite of the different spectral contents. The integration of harmonics of carrying frequencies (at e.g. 35–40 Hz) (Pantev et al., 1994; Jefferys et al., 1996 for references) is also possible and the issue should be investigated further within a conceptual framework attributing a general functional role to oscillatory potentials. Further studies are in progress to investigate under these experimental conditions the dynamics of broad-band VEP and oscillatory responses and the contributions of their generators to the recorded magnetic field. Acknowledgements The study was performed at the Institute of Electronics of the Solid State, the National Council of Research, Rome, with substantial technical contributions from Drs. A. Pasquarelli, M. Peresson, V. Pizzella and G.L. Romani, to whom the authors feel indebted. Thanks are also due to Dr. R. Llinas for critical comments. References Bressler, S.L. The gamma wave: a cortical information carrier. Trends Neurosci., 1990, 13: 161–162. Cracco, R.Q. and Cracco, J.B. Visual evoked potentials in man: early oscillatory potentials. Electroenceph. clin. Neurophysiol., 1978, 45: 731–739. Doty, R.W. and Kimura, D.S. Oscillatory potentials in the visual system of cats and monkeys. J. Physiol. (London), 1963, 168: 205–218. Ducati, A., Fava, E. and Motti, E.D.F. Neural generators of the visual evoked potentials: intracerebral recording in awake humans. Electroenceph. clin. Neurophysiol., 1988, 71: 89–99. Gray, C.M., Ko¨nig, P., Engel, A.K. and Singer, W. Oscillatory responses
95
in cat visual cortex exhibit inter-columnar synchronization which reflects global stimulus properties. Nature, 1989, 338: 334–336. Harding, G.F.A. and Rubinstein, M.P. The scalp topography of the human visually evoked subcortical potential. Invest. Ophthalmol. Vis. Sci., 1980, 19: 318–321. Hughes, J.R. and Mazurowski, J.A. Rhythmical oscillatory activity in visual responses from the unanesthetized monkey. Physiologist, 1962, 5: 158–166. Jefferys, J.G.R., Traub, R.D. and Whittington, M.A. Neuronal networks for induced ‘40 Hz’ rhythms. Trends Neurosci., 1996, 16: 202–208. Karwoski, C. and Kawasaki, K. Oscillatory potentials. In: J.R. Heckenlively and G.B. Arden (Eds.), Principles and Practice of Clinical Electrophysiology of Vision. Mosby-Year Book, St. Louis, 1991, pp. 125– 128. Llinas, R.R. Oscillations in CNS neurons: a possible role for cortical interneurons in the generation of 40-Hz. In: E. Bas¸ar and T.H. Bullock (Eds.), Induced Rhythms in the Brain. Birkha¨user, Boston, 1992, pp. 269–283. Okada, Y., Lauritzen, M. and Nicholson, C. MEG source models and physiology. Phys. Med. Biol., 1987, 32: 43–51. Pantev, C., Elbert, T. and Lu¨tkenho¨ner, B. (Eds.) Oscillatory Eventrelated Brain Dynamics, NATO ASI Series A: Life Sciences, Vol. 271. Plenum Press, New York, 1994. Sannita, W.G., Lopez, L., Piras, C. and Di Bon, G. Scalp-recorded oscillatory potentials evoked by transient pattern-reversal stimulation in man. Electroenceph. clin. Neurophysiol., 1995, 96: 206–218. Schroeder, C.E., Tenke, C.E. and Givre, S.J. Subcortical contributions to the surface recorded flash-VEP in the awake macaque, Electroenceph. clin. Neurophysiol., 1992, 84: 219–231. Sillito, A.M., Jones, H.E., Gerstein, G.L. and West, D.C. Feature-linked synchronization of thalamic relay cell firing induced by feedback from the visual cortex. Nature, 1994, 369: 479–482. Singer, W. Synchronization of cortical activity and its putative role in information processing and learning. Annu. Rev. Physiol., 1993, 55: 349–374. Steriade, M. The flash-evoked afterdischarge, Brain Res., 1968, 9: 169– 212. Tsuchida, Y., Kawasaki, K., Fujimura, K. and Jacobson, J.H. Isolation of faster components in the electroretinogram and visually evoked response in man. Am. J. Ophthalmol., 1973, 75: 846–852. Wachtmeister, L. Basic research and clinical aspects of the oscillatory potentials of the electroretinogram. Doc. Ophthalmol., 1987, 66: 187– 194. Whittaker, S.G. and Siegfried, J.B. Origin of wavelets in the visual evoked potential. Electroenceph. clin. Neurophysiol., 1983, 55: 91– 101.
Magnetically recorded oscillatory responses to luminance stimulation in man1 Luisa Lopez a, Walter G. Sannita b , c , d ,* a
Institute for Advanced Biomedical Technologies, University of Chieti, Chieti, Italy Center for Neuroactive Drugs, Department of Motor Science and Rehabilitation - Neurophysiopathology, University of Genova, Genova, Italy c Center for Cerebral Neurophysiology, National Council of Research, Genova, Italy d Department of Psychiatry and Behavioral Sciences, State University of New York, Stony Brook, NY, USA
b
Accepted for publication: 4 November 1996
Abstract Fast-frequency (ca. 100–110 Hz) oscillatory potentials superimposed on waves N2 and P2 of conventional broad-band VEP were magnetically recorded in man from occipital locations in response to monocular transient flash stimulation with full-field flashes (3.5 cd⋅s⋅m − 2 intensity) and in spots (1, 1.5, or 2.0 cm in diameter). These oscillations proved replicable between- and within-subject and were phase-locked to retinal oscillatory potentials, with maximum correlation at approximately 35 ms and mean delay (as measured between the first measurable peaks) of 27.4 ± 1.6 ms. When stimuli were in spots at increasing eccentricity (5, 15, or 25°) from foveal fixation, broad-band VEP were recorded regardless of diameter and eccentricity of spot, whereas oscillatory responses were not detectable at eccentricity of, or greater than, 15°. This observation suggests that broad-band VEP and the oscillatory response are generated by (partly) distinct neuronal populations and/or functional arrangements and that there is some functional connection between cortical oscillatory responses and stimulus-related events triggered in central retina. 1997 Elsevier Science Ireland Ltd. All rights reserved Keywords: Visual evoked potentials; Oscillatory responses; Luminance stimulation; Retinal eccentricity; Neuromagnetic fields; MEG
1. Introduction Under proper stimulation and recording conditions, short bursts of fast frequency (ca. 100–140 Hz) rhythmic oscillations have been recorded in response to flash stimulation at retinal level in animals and man (Wachtmeister, 1987), from lateral geniculate nucleus (LGN) of monkeys (Doty and Kimura, 1963; Schroeder et al., 1992), and from visual cortex of cat, monkey and man (Hughes and Mazurowski, 1962; Doty and Kimura, 1963; Steriade, 1968; Ducati et al., 1988). Oscillatory responses in the same
* Corresponding author. Center for Neuropsychoactive Drugs, Department of Motor Sciences, University of Genova, I-16132 Genova, Italy. Tel.: +39 10 3537464; fax: +39 10 3537699; e-mail: [email protected] 1 Presented, in part, at the Symposium ‘Oscillatory Event-related Brain Dynamics’, Mu¨nster, Germany, September 2–6, 1993 and the Ninth Annual Meeting of the American Academy of Clinical Neurophysiology, Boston, MA, USA, July 14–16, 1994.
0168-5597/97/$17.00 1997 Elsevier Science Ireland Ltd. All rights reserved PII S0921-884X(96)9652 9-7
frequency range are superimposed on the early waves of human flash-evoked potentials (VEP) recorded from scalp (Cracco and Cracco, 1978; Whittaker and Siegfried, 1983). These oscillatory events are phase-locked to stimulus and become apparent (or are enhanced) when the signal is high-pass filtered with cut-off at approximately 70–90 Hz, but do not depend on narrow-band filter distortion of the electroretinogram (ERG) or VEP (Tsuchida et al., 1973). The existing differences between retinal and scalp-recorded oscillations stand against volume-conducted contamination from retinal sources (Whittaker and Siegfried, 1983) and significant contributions from LGN to cortical responses are unlikely (Schroeder et al., 1992). The sources of retinal oscillatory potentials appear separated from those of ERG and conceivably involve activation of neuronal loops in inner retina (Karwoski and Kawasaki, 1991). Accordingly, scalp-recorded oscillatory responses are suggested to reflect cortical phenomena that are parallel to, and may have (partially) different origin than, broad-band VEP (Whittaker and Siegfried,
EEP 96529
92
L. Lopez, W.G. Sannita / Electroencephalography and clinical Neurophysiology 104 (1997) 91–95
1983). Neuromagnetic, reference-independent recording techniques have been used to support this hypothesis. 2. Materials and methods Eight healthy volunteers (25–35 years old) were recruited after excluding any history or clinical evidence of ocular, neurological or systemic pathology, long-term treatment with, or abuse of, neuroactive drugs. In all cases visual acuity was better than 18/20, with negligible ametropia. Routine EEG, ERG and VEP were within normal limits. Flash stimuli were generated by a Grass PS22 stimulator and delivered via an optical system into the shielded room. Stimulation was monocular, at 0.6 Hz; stimuli were full-field (at 3.5 cd⋅s⋅m − 2 intensity) or in circular spots (with diameter of 1, 1.5, or 2.0 cm, respectively) located at increasing eccentricity in the left or right upper quadrant of visual field (5, 15, 25° from foveal fixation). A 70 dB masking white noise was applied to minimize the effect of noise artifact from the strobe. ERG recordings were performed via Ag/AgCl dermal electrodes positioned on the lower orbital margin and referred to linked mastoids; the ground electrode was at Cz. Neuromagnetic recordings (MEG) were performed with a 28-channel system (16 gradiometers, 9 magnetometers, 3 noise cancellation coils; tail diameter 16 cm) in a magnetically shielded room (Vacuumschmeltze GmbH). The head position with respect to the magnetic probe was kept constant in all stimulus conditions by referring to 3 magnetic coils placed over anatomical landmarks and checked prior to recordings. Two probe positions centered over the parietal-occipital region (P3/O1 and P4/O2 of the 10–20 International System) contralateral to the stimulated eye and the inion (Oz) were used for recordings. The EEG was monitored during and between recordings and subjects were not allowed to drowse. Continuous electric and magnetic signal acquisition was performed in a 0.16–250 Hz bandwidth (high-pass filter: 6 dB/octave; low-pass, Bessel-type filter: 48 dB/octave) at 1 kHz sampling rate. Oscillatory potentials were high-pass filtered at 80 Hz. Averages were on consecutive 60–70 epochs free of artifacts. The ethical principles of the Declaration of Helsinki (1964) by the World Medical Association concerning human experimentation were followed. The study was performed at the Institute of Electronics of the Solid State (the National Council of Research, Rome), with approval, and all volunteers signed an informed consent.
3. Results In all recordings after full-field flash stimulation, oscillatory responses appeared superimposed on the ERG as well as on waves N2 and P2 of the VEP and could be dissociated from the underlying broad-band response by high-pass filtering, without apparent filter distortion. The oscillatory responses proved replicable within-subject and consistent across subjects (Fig. 1). The average duration and maximum amplitude of the oscillations recorded at parietal-occipital locations were 40–50 ms and 60–150 fT, respectively, while the duration of retinal oscillatory potentials was 30–35 ms. The oscillatory fields recorded magnetically at parietal-occipital locations were correlated with the retinal oscillatory responses, with maximum correlation at a delay of approximately 35 ms (Fig. 1D); the mean delay between the onset of retinal and occipital responses (as measured between the first measurable peaks) was 27.4 ± 1.6 ms. ERG or retinal oscillatory potentials were not recorded after stimulation in spots, consistent with the stimulated area/response function of retinal responses to luminance stimuli. Broad-band VEP were conversely recorded under these stimulus conditions regardless of spot diameter and at any eccentricity in the 5–25° range. The VEP amplitude decreased with decreasing diameter of spot stimuli in the 1–2 cm range, but latencies were substantially unaffected. Oscillatory responses to spot stimulation were recorded from occipital locations at 5° eccentricity of stimulus (apparently with a limited effect of spot diameter), but were not observed at eccentricities of, or greater than, 15° (Fig. 2). 4. Discussion As opposed to the short delay between retinal and LGN oscillatory potentials in monkeys (Schroeder et al., 1992), the latency delay in man between retinal and parietal-occipital oscillations suggests complex operations and some higher degree of cellular interaction. In this regard, substantial cortical contributions in the generation of nearfield oscillatory responses recorded at parietal-occipital level are plausible. This hypothesis is based on theoretical and experimental evidence that only near-field currents related to intracellular activity are recordable by reference-free magnetic methods (Okada et al., 1987), and is supported by the observation of parietal-occipital oscilla-
Fig. 1. (A) Comparison of the oscillatory responses to full-field luminance stimulation as recorded electrically at retinal level (top; ‘retina’) and magnetically at occipital locations (bottom; ‘MEG’) in 4 healthy subjects. Superimposed recordings from 5 magnetic channels at a location corresponding to the occipital area contralateral to the stimulated eye are shown for each subject. Calibration as indicated. (B) Effect of high-pass filtering at increasing cut-offs on the response to full-field stimulation magnetically recorded at occipital location from a fifth subject. Note that oscillatory responses appear superimposed on conventional, broad-band (e.g. 1–200 Hz) VEP and are enhanced by narrow-band filtering, without apparent filter distortion. (C) Comparison on an expanded time scale between retinal and magnetically-recorded oscillatory responses. (D) Correlogram (computed over time and fitted to the shortest period within the selected 80–150 Hz frequency band) between the retinal oscillatory potential shown in (C) and the responses from 5 adjacent scalp locations of one subject. Note the peak correlation at approximately 35 ms.
L. Lopez, W.G. Sannita / Electroencephalography and clinical Neurophysiology 104 (1997) 91–95
93
94
L. Lopez, W.G. Sannita / Electroencephalography and clinical Neurophysiology 104 (1997) 91–95
Fig. 2. Effects of diameter and eccentricity of spot stimuli on conventional broad-band VEP and narrow-band oscillatory response. Magnetic recording; 4 superimposed adjacent channels. The stimulus is indicated by vertical bars. Note that VEP are recordable at spot diameter as small as 1 cm and eccentricities of stimulus up to 25°, whereas the oscillatory response is not evident at eccentricities of 15° or more and approaches the noise level with spot diameter of 1 cm. No retinal oscillatory events were recorded after spot stimulation.
tory responses to focal (spot) stimuli eliciting no retinal response as well as by the described grading off a few millimeters outside the LGN of the oscillations originating from lamina 6 of LGN, without significant contributions to the scalp-recorded VEP (Schroeder et al., 1992). Further to this, the oscillatory components of cortical primary response of cat are suggested to be independent of responses in LGN or optic tract because of their enhancement in conditions of increased cortical excitability (Steriade, 1968). However, visual evoked potentials with short latency and suggested subcortical origin but still unidentified sources have been described in man (Harding and Rubinstein, 1980) and pre-cortical components of the oscillatory response cannot be excluded. Cortical circuitry may tentatively be held responsible for the delay between retinal and ‘cortical’ oscillatory macropotentials. Oscillatory events occurring at singleand multi-unit level over a relatively wide range of frequencies (ca. 20–100 Hz) are suggested to mediate in the synchronization of relatively large populations of neurons (Bressler, 1990; Singer, 1993; Jefferys et al., 1996), which is a prerequisite for any neural activity to be driven into patterns of oscillation recordable via macroelectrodes. Membrane oscillations recorded intracellularly in the visual cortex of cat depend on the physical properties of
stimulus (Gray et al., 1989; Sillito et al., 1994) and are believed to reflect cortical mechanisms operating in conjunction with intrinsic neuronal membrane properties and thalamocortical feedback loops (Gray et al., 1989; Llinas, 1992; Singer, 1993; Sillito et al., 1994; Jefferys et al., 1996). These mechanisms operate in sensory information processing and conceivably play a role in the generation of both oscillatory activities at cellular level and local-field potentials. The recording of broad-band VEP in the absence of oscillatory responses after stimulation with spots at eccentricities equal to, or greater than, 15° suggests that the generation of broad-band VEP and oscillatory responses depends on the activation of (partly) distinct neuronal populations and/or functional arrangements eventually overlapping in MEG space, and further indicates visual information processing through parallel subsystems. It should be noted in this regard that oscillatory responses to non-visual stimuli have also been described as being distinct from the underlying broad-band evoked potentials (Singer, 1993; Pantev et al., 1994for references). Unlike retinal oscillatory potentials and conventional broad-band VEP, the near-field oscillatory responses at parietal-occipital locations appear to depend on events triggered through central rather than reflect diffuse retinal stimula-
L. Lopez, W.G. Sannita / Electroencephalography and clinical Neurophysiology 104 (1997) 91–95
tion. Substantial involvement of geniculocalcarine pathway is plausible. The oscillatory responses to luminance and those recorded after pattern-reversal stimulation of 9° central retina, with ‘tuning’ at approximately 5 cycles/ degree (Sannita et al., 1995) may therefore reflect mechanisms of information processing and transfer within the visual system that are in part common to the two paradigms of visual stimulation, in spite of the different spectral contents. The integration of harmonics of carrying frequencies (at e.g. 35–40 Hz) (Pantev et al., 1994; Jefferys et al., 1996 for references) is also possible and the issue should be investigated further within a conceptual framework attributing a general functional role to oscillatory potentials. Further studies are in progress to investigate under these experimental conditions the dynamics of broad-band VEP and oscillatory responses and the contributions of their generators to the recorded magnetic field. Acknowledgements The study was performed at the Institute of Electronics of the Solid State, the National Council of Research, Rome, with substantial technical contributions from Drs. A. Pasquarelli, M. Peresson, V. Pizzella and G.L. Romani, to whom the authors feel indebted. Thanks are also due to Dr. R. Llinas for critical comments. References Bressler, S.L. The gamma wave: a cortical information carrier. Trends Neurosci., 1990, 13: 161–162. Cracco, R.Q. and Cracco, J.B. Visual evoked potentials in man: early oscillatory potentials. Electroenceph. clin. Neurophysiol., 1978, 45: 731–739. Doty, R.W. and Kimura, D.S. Oscillatory potentials in the visual system of cats and monkeys. J. Physiol. (London), 1963, 168: 205–218. Ducati, A., Fava, E. and Motti, E.D.F. Neural generators of the visual evoked potentials: intracerebral recording in awake humans. Electroenceph. clin. Neurophysiol., 1988, 71: 89–99. Gray, C.M., Ko¨nig, P., Engel, A.K. and Singer, W. Oscillatory responses
95
in cat visual cortex exhibit inter-columnar synchronization which reflects global stimulus properties. Nature, 1989, 338: 334–336. Harding, G.F.A. and Rubinstein, M.P. The scalp topography of the human visually evoked subcortical potential. Invest. Ophthalmol. Vis. Sci., 1980, 19: 318–321. Hughes, J.R. and Mazurowski, J.A. Rhythmical oscillatory activity in visual responses from the unanesthetized monkey. Physiologist, 1962, 5: 158–166. Jefferys, J.G.R., Traub, R.D. and Whittington, M.A. Neuronal networks for induced ‘40 Hz’ rhythms. Trends Neurosci., 1996, 16: 202–208. Karwoski, C. and Kawasaki, K. Oscillatory potentials. In: J.R. Heckenlively and G.B. Arden (Eds.), Principles and Practice of Clinical Electrophysiology of Vision. Mosby-Year Book, St. Louis, 1991, pp. 125– 128. Llinas, R.R. Oscillations in CNS neurons: a possible role for cortical interneurons in the generation of 40-Hz. In: E. Bas¸ar and T.H. Bullock (Eds.), Induced Rhythms in the Brain. Birkha¨user, Boston, 1992, pp. 269–283. Okada, Y., Lauritzen, M. and Nicholson, C. MEG source models and physiology. Phys. Med. Biol., 1987, 32: 43–51. Pantev, C., Elbert, T. and Lu¨tkenho¨ner, B. (Eds.) Oscillatory Eventrelated Brain Dynamics, NATO ASI Series A: Life Sciences, Vol. 271. Plenum Press, New York, 1994. Sannita, W.G., Lopez, L., Piras, C. and Di Bon, G. Scalp-recorded oscillatory potentials evoked by transient pattern-reversal stimulation in man. Electroenceph. clin. Neurophysiol., 1995, 96: 206–218. Schroeder, C.E., Tenke, C.E. and Givre, S.J. Subcortical contributions to the surface recorded flash-VEP in the awake macaque, Electroenceph. clin. Neurophysiol., 1992, 84: 219–231. Sillito, A.M., Jones, H.E., Gerstein, G.L. and West, D.C. Feature-linked synchronization of thalamic relay cell firing induced by feedback from the visual cortex. Nature, 1994, 369: 479–482. Singer, W. Synchronization of cortical activity and its putative role in information processing and learning. Annu. Rev. Physiol., 1993, 55: 349–374. Steriade, M. The flash-evoked afterdischarge, Brain Res., 1968, 9: 169– 212. Tsuchida, Y., Kawasaki, K., Fujimura, K. and Jacobson, J.H. Isolation of faster components in the electroretinogram and visually evoked response in man. Am. J. Ophthalmol., 1973, 75: 846–852. Wachtmeister, L. Basic research and clinical aspects of the oscillatory potentials of the electroretinogram. Doc. Ophthalmol., 1987, 66: 187– 194. Whittaker, S.G. and Siegfried, J.B. Origin of wavelets in the visual evoked potential. Electroenceph. clin. Neurophysiol., 1983, 55: 91– 101.