Skip to main content
NCBI home page
The Journal of Physiology logoLink to The Journal of Physiology
. 2005 Aug 25;568(Pt 2):699–711. doi: 10.1113/jphysiol.2005.089722

Movement-related frequency modulation of beta oscillatory activity in the human subthalamic nucleus

G Foffani 1,2, AM Bianchi 3, G Baselli 3, A Priori 1
PMCID: PMC1474736  PMID: 16123109

Abstract

Event-related changes of brain electrical rhythms are typically analysed as amplitude modulations of local field potential (LFP) oscillations, like radio amplitude modulation broadcasting. In telecommunications, frequency modulation (FM) is less susceptible to interference than amplitude modulation (AM) and is therefore preferred for high-fidelity transmissions. Here we hypothesized that LFP rhythms detected from deep brain stimulation (DBS) electrodes implanted in the subthalamic nucleus (STN) in patients with Parkinson's disease could represent movement-related activity not only in AM but also in FM. By combining adaptive autoregressive identification with spectral power decomposition, we were able to show that FM of low-beta (13–20 Hz) and high-beta (20–35 Hz) rhythms significantly contributes to the involvement of the human STN in movement preparation, execution and recovery, and that the FM patterns are regulated by the dopamine levels in the system. Movement-related FM of beta oscillatory activity in the human subthalamic nucleus therefore provides a novel informational domain for rhythm-based pathophysiological models of cortico-basal ganglia processing.

How the human basal ganglia process information is still largely unknown (Brown & Marsden, 1998; Wichmann & DeLong, 1999; Bar-Gad & Bergman, 2001; Bevan et al. 2002; Wichmann & DeLong, 2003; Bar-gad et al. 2003; Brown, 2003; Hutchison et al. 2004). The study of local electroencephalographic activity – i.e. the local field potentials (LFPs) – recorded through electrodes stereotactically implanted in the basal ganglia to treat movement disorders with deep brain stimulation (DBS) offers a fascinating research tool in this field (Brown, 2003; Foffani et al. 2003; Priori et al. 2004). LFP studies revealed a pattern of oscillatory signals in the human basal ganglia far more complex than available box-and-arrow, rate-based models (Brown, 2003; Priori et al. 2004). Consistent with scalp EEG literature (Pfurtscheller & Lopes da Silva, 1999), basal ganglia rhythms correlate with movement by changing the amplitude (or power) of their oscillations (Foffani et al. 2002, 2003; Levy et al. 2002a; Priori et al. 2002; Cassidy et al. 2002; Kühn et al. 2004; Williams et al. 2005; Doyle et al. 2005), which corresponds to the classical telecommunication concept of amplitude modulation (AM) (Haykin, 1994; Clochon et al. 1996). However, recent EEG studies showed that event-related brain dynamics include phase-resetting phenomena of ongoing rhythms (Makeig et al. 2002, 2004; Jansen et al. 2003; David et al. 2005; Gruber et al. 2005), which cannot be explained in terms of AM (Penny et al. 2002). In communication theory, phase-resetting can be viewed as a particular form of phase modulation or, using a more popular equivalent, of frequency modulation (FM) (Haykin, 1994; Penny et al. 2002). FM is more robust than AM against interference and against signal amplitude fading phenomena, and is therefore used for high-fidelity radio transmission (Haykin, 1994). From a neurophysiological point of view, the electrochemical properties within neurones and the connectivity between neurones directly control the firing patterns of neuronal oscillatory activity, making the frequency of neuronal rhythms an attractive variable for information processing and communication in the brain (Varela et al. 2001; Salinas & Sejnowski, 2001; Schnitzler & Gross, 2005). We therefore hypothesized that FM could be an alternative informational domain for understanding event-related oscillatory electrical activity in the human basal ganglia. To test the FM hypothesis, we applied a recently validated approach (Foffani et al. 2004a) to study both movement-related AMs and FMs in the LFPs recorded from the subthalamic nucleus (STN) in patients with Parkinson's disease. We focused the attention on subthalamic rhythms in the low-beta (13–20 Hz) and high-beta (20–35 Hz) frequency ranges, because they received the highest degree of attention in the recent literature (Brown et al. 2001; Levy et al. 2002a; Priori et al. 2002, 2004; Cassidy et al. 2002; Williams et al. 2002, 2003, 2005; Silberstein et al. 2003; Kühn et al. 2004, 2005; Doyle et al. 2005; Foffani et al. 2005a) and are considered critical for the pathophysiological involvement of the human basal ganglia in motor control (Brown, 2003; Hutchison et al. 2004). To fully assess the neurophysiological relevance of subthalamic FMs compared to more standard AMs, we also investigated their dependence on dopaminergic medication.

Methods

Patients

Eleven patients (6 females, 5 males) with idiopathic Parkinson's disease were studied after their informed consent and local ethical committee approval. The average age was 56 years (range 38–69), years of disease history 13 (7–20), levodopa equivalent therapy pre-surgery 1438 mg day−1 (750–2800), 15–20 days post-surgery 311 mg day−1 (75–800), Unified Parkinson's Disease Rating Scale (UPDRS) III (motor part) pre-surgery off therapy 43 (27–72.5), pre-surgery on therapy 5 (1–23.5), 1 year after surgery off therapy on stimulation 9 (1.5–20, n = 9), UPDRS IV (complications, A + B) pre-surgery 10 (5–14), 1-year after surgery 2 (0–6, n = 9). Patients were treated with DBS only fulfilling specific inclusion criteria (L.I.M.P.E. 2003). All patients were clinically selected, surgically implanted, neurophysiologically assessed and clinically followed by the Milan DBS Group Policlinico, San Paolo, Politecnico (Italy). Part of the data has been used in previous works. Namely, we described the rest oscillations below 50 Hz of 13 nuclei from nine of these patients (Priori et al. 2004) and the 300 Hz oscillations of 11 nuclei from seven of these patients (Foffani et al. 2003, 2005b), recorded both before and after levodopa.

Surgical procedures

The methods for the localization of DBS electrodes within the STN are extensively reported elsewhere (Foffani et al. 2003). Briefly, the procedures included (a) the preoperative direct visualization of the nucleus through CT-MRI targeting (Egidi et al. 2002; Rampini et al. 2003), (b) the intraoperative neurophysiology with microrecordings (Priori et al. 2003), intraoperative stimulation (i.e. through the explorative microelectrode) and macrostimulation (i.e. through the implanted macroelectrode), (c) the postoperative CT-MRI verification of the final electrode position (Egidi et al. 2002; Rampini et al. 2003), and (d) the postoperative assessment of the optimal DBS contact. The model 3389 electrode (Medtronic, Minneapolis, USA) was implanted, which had four cylindrical contacts (diameter 1.27 mm), each 1.5 mm long and spaced 0.5 mm away from the others (i.e. 2 mm centre to centre). Contacts were denominated 0, 1, 2, and 3 beginning from the more caudal placement. The objective was to place contact 1 into the target position. Double blind assessment revealed marked clinical effects during intraoperative monopolar macrostimulation through contact 1. In all the patients included in this study for STN recordings contact 1 (and consistently less contact 2) induced a remarkably greater clinical effect (scored postoperatively by two independent blind observers) than contacts 0 and 3. All of these procedures taken together were consistent with the localization of contact 1 within the STN (Foffani et al. 2003; Priori et al. 2004). All 11 patients received bilateral STN implants.

LFP recordings

The postoperative recording sessions took place 2–3 days after patients had received the electrode implants. The patients were comfortably seated on an armchair. LFPs were recorded during the execution of voluntary movements, 8–12 h after overnight withdrawal of dopaminergic medication, before and after administration of 100–200 mg oral fast-acting l-DOPA (Madopar Dispersibile, Roche, Italy). Levodopa dosages were chosen in order to induce a clinical response as similar as possible to the full on-state used for the pre- and postoperative UPDRS evaluation. After-medication recordings started 30–40 min after drug administration, when the patient showed clinical improvement. The latter was evaluated by an experienced neurologist and by self-assessment of the patient. Most patients had some dyskinesias after levodopa but none before. Patients were instructed to extend the second finger from an initial relaxed semiflexed position every 8–10 s as quickly as possible. Movements were self-paced. One extension corresponded to one trial; the mean number of artifact-free trials was 97 ± 40 before levodopa, 84 ± 43 after levodopa. The task execution was interrupted either by the patient will, if tired, or after a block of 50 trials. Multiple blocks of trials were usually recorded from each nucleus before and after levodopa separated by rest periods to avoid fatigue. Dyskinesias, when present, did not interfere with task execution. The EMG signal from the extensor indicis muscle was recorded through a pair of surface non-polarizable Ag–AgCl electrodes with a belly tendon montage. LFPs were captured from the contralateral 3389 electrode using a closely spaced pair of contacts (1–2). Signals were preamplified, differentially amplified and filtered (EMG: 20–1000 Hz, STN: 2–1000 Hz) through a Cambridge 1902 (Cambridge Electronic Design, Cambridge, UK), A/D converted (sampling rate 2500 Hz) through a Cambridge 1401 (Cambridge Electronic Design), on-line analysed on a personal computer and stored by Signal software (version 1.80, Cambridge Electronic Design). All further analysis was conducted off-line with the Matlab software (version 6.5, The Mathworks, Natick, MA, USA) with custom written programs described below unless otherwise specified. Due to time constraints during the experiments, three patients were studied bilaterally both before (6 nuclei) and after (6 nuclei) levodopa, three patients were studied unilaterally both before (3 nuclei) and after (3 nuclei) levodopa, one patient was studied bilaterally before (2 nuclei) and unilaterally after (1 nucleus) levodopa, one patient was studied unilaterally before (1 nucleus) and bilaterally after (2 nuclei) levodopa, two patients were studied unilaterally only before (2 nuclei) levodopa and one patient was studied bilaterally only after (2 nuclei) levodopa. In total, we studied 17 nuclei, 14 before levodopa and 14 after levodopa, from 11 patients.

Data analysis I: preprocessing

As a preprocessing step, LFPs were normalized by subtracting the mean and dividing by the standard deviation of the 600–1000 Hz band-pass filtered signals, in order to minimize variability within and between nuclei (Foffani et al. 2003; Priori et al. 2004). Signals were then low-pass filtered below 45 Hz and down-sampled at 125 Hz. Movement onsets, which triggered the analyses, were evaluated on the EMG signal using the following cumulative sum approach (Foffani et al. 2003): (1) the EMG signal was squared to obtain power; (2) the mean of the resulting signal was subtracted point-by-point; (3) the cumulative sum of the resulting signal was calculated and then smoothed using a moving average forward–backward (zero phase) filter; (4) the movement onsets were defined as the local minima (the signal at the previous and at the next instants is greater than the signal at the current instant) and the offsets as the local maxima (the signal at the previous and at the next instants is lower than the signal at the current instant) in the filtered cumulative sum; and (5) signals were visually inspected, and only epochs free from artifacts were used for the analysis (Fig. 1A).

Figure 1. Adaptive autoregressive (AAR) identification with spectral power decomposition for studying subthalamic movement-related activity.

Figure 1

Open in a new tab

A, the top trace (y-axis: amplitude in a.u.) shows a representative example (60 s) of low-pass filtered (below 45 Hz) local field potential (LFP) recording from the subthalamic nucleus (STN) in a patient with Parkinson's disease while executing voluntary second finger extensions. The central trace represents the squared EMG from the extensor indicis muscle (y-axis: squared amplitude in a.u.). The bottom trace (y-axis: cumsum amplitude in a.u.) shows the smoothed cumulative sum of the squared EMG signal (see central trace), after removing the mean. The vertical bars in correspondence to the local minima of the cumulative sum are the estimates of the movement onsets, which triggered the subsequent analyses. B, schematic representation of the AAR method with spectral power decomposition (Foffani et al. 2004a). Each signal y(t) is submitted to AAR identification. The resulting time-varying parameter vector Θ(t) is synchronously averaged across all available trials (i.e. movement executions), using movement onset as a trigger. The resulting average AAR model 〈Θ(t)〉 undergoes spectral power decomposition through the residual method in order to obtain the time-varying ‘poles’. Each real pole i or complex pole pair i is characterized by a power value Pi(t) and a frequency value Fi(t), both dependent on time, which separately describe movement-related power and frequency modulations.

Data analysis II: adaptive autoregressive identification with spectral power decomposition

Time–frequency analysis was applied to the movement-related LFPs using a method based on adaptive autoregressive (AAR) identification with spectral power decomposition (Fig. 1B). The method was described in detail and validated elsewhere (Foffani et al. 2004a). In brief, LFPs were modelled using a generic AAR model of the form:

graphic file with name tjp0568-0699-m1.jpg

where Θ(t) =[a1(t) a2(t) … an(t)] is the time-varying autoregressive parameter vector, ϕ(t) =[y(t− 1) y(t− 2) … y(t−n)]T is the observation vector, d(t) is a white noise (mean = 0, variance =σ2) and n is the model order. The parameter vector Θ(t) was identified from the data using an adaptive algorithm with ‘whale’ forgetting factor (Bianchi et al. 1997), given by the following recursive equations:

graphic file with name tjp0568-0699-m2.jpg (1)

where α1 and α2 are such that the two roots p1 and p2 of z2−α1z−α2= 0 lie in the interval [0,1], Q is a n-dimensional vector (initialized to zero) and S is an n-by-n square matrix (initialized as a unitary matrix with gain 0.0001). The whale forgetting factor (α1≠ 0, α2≠ 0) was preferred to the more standard exponential forgetting factor (α1≠ 0, α2= 0) due to its better performance in terms of disturbance rejection when α1 and α2 are chosen such that the two roots p1 and p2 of z2−α1z−α2= 0 are a couple of coincident roots pp(0 < pp < 1). As a compromise between tracking capability and disturbance rejection, the whale forgetting factor pp was fixed at 0.93, which corresponds to an approximate memory length of about n = 55 samples according to the formula N≈−4/ln(pp). The values of α1 and α2 were then calculated as α1= 2pp and α2=−pp2. Because any movement-related change increases the prediction error, the input error variance σ2 was fixed to a normalized constant unitary value in order not to overestimate power increases or underestimate power decreases. This normalization also minimizes the signal-to-noise ratio variability across recordings, facilitating statistical comparisons. Power values were then expressed in arbitrary units (a.u.). The model order was set to n = 11 for all recordings. This model order is consistent with current neurophysiological knowledge, because the resulting poles match the main STN rhythms previously reported using non-parametric methods (Priori et al. 2004). Namely, one real pole follows the activity at very low frequencies (< 7 Hz), one pair of complex poles follows the alpha activity (7–13 Hz), one pair follows the low-beta (13–20 Hz), one pair follows the high-beta (20–35 Hz), one pair follows the gamma (35–45 Hz) and one pair accounts for residual high-frequencies (see also Foffani et al. 2004a). Trial averaging was performed in the parameter domain: the parameter vector was synchronously averaged across all the available trials triggered on movement onsets, obtaining an average movement-related AAR model given by following parameter vector:

graphic file with name tjp0568-0699-m3.jpg (2)

where Θτ(t) is the movement-related AAR parameter vector of trial τ and T is the total number of trials used in the average. The above trial averaging (2) assumes that each single trial is a different independent realization of the same AAR process. The average movement-related AAR parameter vector 〈Θ(t)〉 then contains all the time–frequency spectral information of the signal. The time-varying power spectrum of the average movement-related AAR model 〈Θ(t)〉 was divided into components relevant to real poles or to pairs of complex conjugate poles in the z-domain using the residual method (Baselli et al. 1997; see Foffani et al. 2004a for details). This spectral power decomposition was performed at every time instant t of the average movement-related AAR model 〈Θ(t)〉, so that each real pole i or pair of complex conjugate poles i was characterized by a time-dependent power value Pi(t) and a time-dependent frequency value Fi(t), allowing us to separately describe movement-related amplitude modulations (AMs, measured in log power) and frequency modulations (FMs, measured in Hz). The attention was concentrated on low-beta (13–20 Hz) and high-beta (20–35 Hz) poles (Priori et al. 2002; Foffani et al. 2004a) (i.e. poles whose frequency values Fi(t) varied within the above frequency ranges).

Data analysis III: parameter extraction and statistical comparisons

Four main parameters were extracted to quantify the movement-related modulations of each pole: the baseline value ‘b’ (4 s to 2 s before movement onset), the premovement value ‘p’ (0.5 s to 0 s before movement onset), the movement value ‘m’ (0 s to 1 s after movement onset) and the recovery value ‘r’ (2 s to 3 s after movement onset). Two separate statistical analyses were performed: at the single-nucleus level and at the population level. At the single-nucleus level, the significance of AMs and FMs was assessed by challenging the null hypothesis that premovement ‘p’, movement ‘m’ and recovery ‘r’ had the same mean value as the baseline ‘b’, using unpaired t tests (Matlab function ‘ttest2’). Bonferroni correction was applied for multiple comparisons and a cut-off significance value was set at P < 0.00014881. At the population level, the results were analysed with a two-way analysis of variance (ANOVA) design. The first main factor of the ANOVA was the movement-related modulation, with four levels (repeated measures): baseline ‘b’, premovement ‘p’, movement ‘m’ and recovery ‘r’. The second main factor was the dopaminergic condition, with two levels (independent measures): before and after levodopa administration. Each nucleus was considered as an individual sample. Four ANOVAs were separately performed to analyse the movement-related AMs (measured in log power) and FMs (measured in Hz) of the low-beta and high-beta poles. Finally, the latencies of AM and FM onsets (relative to movement onset) were measured as the last zero-crossing before a significant modulation, after subtracting the baseline mean value. This latency estimation is conceptually identical to the approach applied to subthalamic movement-related potentials by Paradiso et al. (2003).

Results

Amplitude modulations

Statistical analysis at the single-nucleus level revealed that most nuclei exhibited significant AMs in the low-beta and high-beta poles both before and after levodopa administration. Based on the average baseline ‘b’, pre-movement ‘p’, movement ‘m’ and recovery ‘r’ values, we observed two main movement-related AM patterns (Table 1; cut-off significance P < 0.000148810). The most consistent AM pattern was a movement-related power decrease (P(m) < P(b), i.e. negative AM; Fig. 2A and C, which typically started before movement onset (P(m) < P(b) and P(p) < P(b)) and was followed by a positive post-movement rebound (P(m) < P(b) and P(r) > P(b)). The second AM pattern consisted of a movement-related power increase (P(m) > P(b), i.e. positive AM; Fig. 2B and D, which also typically started before movement onset (P(m) > P(b) and P(p) > P(b)). See Table 1 for details.

Table 1.

Patterns of movement-related modulations of beta subthalamic rhythms

Low beta High beta
AM
 Pattern 1: Negative AM P(m) < P(b) P(m) < P(b) & P(p) < P(b) P(m) < P(b) & P(r) > P(b) P(m) < P(b) P(m) < P(b) & P(p) < P(b) P(m) < P(b) & P(r) > P(b)
  Before levodopa 10 8 10 11 10 9
  After levodopa 12 9 10 13 12 11
 Pattern 2: Positive AM P(m) > P(b) P(m) > P(b) & P(p) > P(b) P(m) > P(b) & P(r) < P(b) P(m) > P(b) P(m) > P(b) & P(p) > P(b) P(m) > P(b) & P(r) < P(b)
  Before levodopa 3 3 0 1 1 1
  After levodopa 2 1 0 1 0 0
FM
 Pattern 1: Positive FM F(m) > F(b) F(m) > F(b) & F(p) > F(b) F(m) > F(b) & F(r) < F(b) F(m) > F(b) F(m) > F(b) & F(p) > F(b) F(m) > F(b) & F(r) < F(b)
  Before levodopa 9 5 7 10 8 9
  After levodopa 2 1 1 11 6 6
 Pattern 2: Negative FM F(m) < F(b) F(m) < F(b) & F(p) > F(b) F(m) < F(b) & F(r) > F(b) F(m) < F(b) F(m) < F(b) & F(p) < F(b) F(m) < F(b) & F(r) > F(b)
  Before levodopa 4 2 0 4 2 0
  After levodopa 11 5 5 2 2 1
Open in a new tab

The table collects the results of the statistical analysis within individual nuclei, reporting the number of nuclei exhibiting significant amplitude modulations (AM) or frequency modulations (FM) in the low-beta pole or in the high-beta pole, before (total n = 14) and after (total n = 14) levodopa administration, based on the power ‘P’ and the frequency ‘F’ evaluated in the time windows corresponding to baseline ‘b’, pre-movement ‘p’, movement ‘m’ and recovery ‘r’. The average movement-related AM/FM curves corresponding to the patterns in this table are shown in Fig. 2.

Figure 2. Patterns of movement-related beta activity from local field potentials (LFPs) recorded in the human subthalamic nucleus (STN).

Figure 2

Open in a new tab

A–D, amplitude modulations (AMs) in the low-beta pole (A and B) and the high-beta pole (C and D), separating the negative AM patterns (A and C) from the positive AM patterns (B and D). E–H, frequency modulations (FMs) in the low-beta pole (E and F) and the high-beta pole (G and H), separating the positive FM patterns (E and G) from the negative FM patterns (F and H). Pole values were calculated for each nucleus from the adaptive autoregressive parameters averaged across trials (see Fig. 1; Foffani et al. 2004a). Only nuclei exhibiting the same significant modulation patterns (see Table 1) were averaged together. Gray lines show nuclei recorded before levodopa administration, black lines show nuclei recorded after levodopa administration. The number of nuclei averaged in each condition is also indicated in the corresponding colour. x-axes: movement-related time in seconds (s); vertical grey dashed lines (0 s) correspond to movement onset. y-axes: AMs in Log power (a.u.) (A–D) or FMs in frequency (Hz) (E–H). ‘b’= baseline, ‘p’= pre-movement, ‘m’= movement, ‘r’= recovery.

Statistical analysis of the data at the population level (Table 2) confirmed the presence of significant movement-related AMs (considering ‘b’, ‘p’, ‘m’ and ‘r’) in both the low-beta pole (ANOVA 1st main factor: P < 0.000001, Fig. 3A) and the high-beta pole (P < 0.000001, Fig. 3B). In the low-beta pole, levodopa administration significantly decreased the baseline amplitude (2nd main factor: P= 0.019, Fig. 3A), but did not affect the movement-related AM (interaction factor: P= 0.63, Fig. 3A). In the high-beta pole, levodopa had no significant effects on the baseline amplitude or its modulation (P= 0.55 and P= 0.93, respectively, Fig. 3B). Latency analysis on all nuclei exhibiting significant movement-related AMs (either P(m) < P(b) or P(m) > P(b)) confirmed that the modulations started well before movement onset (low-beta: 1.26 ± 0.97 s before levodopa, n = 13, and 0.93 ± 1.09 s after levodopa, n = 14; high-beta: 0.97 ± 0.76 s before levodopa, n = 12 and 1.01 ± 0.85 s after levodopa, n = 14).

Table 2.

Movement-related modulations of beta subthalamic rhythms

Low beta High beta
b p m r b P m r
AM (a.u.)
  Before levodopa Mean 0.74 0.74 0.71 0.84 0.37 0.29 0.21 0.44
s.d. 0.34 0.38 0.49 0.34 0.30 0.28 0.36 0.32
  After levodopa Mean 0.42 0.38 0.33 0.51 0.31 0.24 0.13 0.36
s.d. 0.37 0.34 0.37 0.39 0.35 0.28 0.27 0.37
FM (Hz)
  Before levodopa Mean 17.11 17.18 17.45 16.89 26.43 26.49 26.59 25.95
s.d. 1.46 1.47 1.41 1.29 1.12 1.33 1.64 1.00
  After levodopa Mean 17.07 17.1 16.68 17.06 26.55 26.68 26.81 26.48
s.d. 1.58 1.47 1.25 1.73 1.13 1.19 1.24 1.18
Open in a new tab

The table reports means and standard deviations (s.d.) of the amplitude modulations (AM, in log power arbitrary units, a.u.) and the frequency modulations (FM, in Hz) for the low-beta and the high-beta poles, of all nuclei recorded before (n = 14) and after (n = 14 nuclei) levodopa administration. Entries for each nucleus were obtained by averaging the pole values of individual nuclei in four movement-related time windows: b= baseline (4 s to 2 s before movement onset), p= pre-movement (0.5 s to 0 s before movement onset), m= movement (0 s to 1 s after movement onset), r= recovery (2 s to 3 s after movement onset). The full movement-related AM/FM curves corresponding to this table are shown in Fig. 3.

Figure 3. Grand-average of the movement-related beta activity from local field potentials (LFPs) recorded in the human subthalamic nucleus (STN).

Figure 3

Open in a new tab

A and B, amplitude modulations (AMs) in the low-beta pole (A) and the high-beta pole (B). C and D, frequency modulations (FMs) in the low-beta pole (C) and the high-beta pole (D). Pole values were calculated for each nucleus from the adaptive autoregressive parameters averaged across trials (see Fig. 1; Foffani et al. 2004a) and then were averaged across all nuclei recorded before (n = 14, grey lines) and after (n = 14, black lines) levodopa administration. x-axes: movement-related time in seconds (s); vertical grey dashed lines (0 s) correspond to movement onset. y-axes: AMs in Log power (a.u) (A and B) or FMs in frequency (Hz) (C and D). ‘b’= baseline, ‘p’= premovement, ‘m’= movement, ‘r’= recovery. Means and standard deviations in these time windows are reported in Table 2.

Frequency modulations

Besides AMs, FMs were also a consistent feature in our dataset. Statistical analysis at the single-nucleus level revealed that most nuclei exhibited significant FMs in the low-beta pole and in the high-beta pole both before and after levodopa administration. Two main movement-related FM patterns were observed (Table 1; cut-off significance P < 0.000148810). The most consistent FM pattern was a movement-related frequency increase (F(m) > P(b), i.e. positive FM; Fig. 2E and G, which typically started before movement onset (F(m) > P(b) & F(p) > P(b)) and was followed by a negative post-movement rebound (F(m) > P(b) and P(r) < P(b)). The second FM pattern consisted of movement-related frequency decrease (F(m) < F(b), i.e. negative FM; Figs 2F and H), which typically started before movement onset (F(m) < F(b) and F(p) < F(b)) with positive post-movement rebounds often observed after levodopa (F(m) < F(b) and F(r) > F(b)). Differently from the AM patterns, levodopa induced consistent effects on the FM patterns, particularly in the low-beta pole, where we observed an inversion of polarity from positive FM (Fig. 2E) to negative FM (Fig. 2F). This result was corroborated by observing that of 11 nuclei recorded in both pharmacological conditions (i.e. both before and after levodopa administration), 7 of the 9 nuclei exhibiting positive FM before levodopa reversed their FM polarity after levodopa, the remaining two nuclei exhibiting positive FM before levodopa decreased the FM after levodopa, and the two nuclei remaining exhibited negative FM both before and after levodopa.

Statistical analysis at the population level (Table 2) confirmed the FM patterns and the effects of levodopa observed at the single-nucleus level. Namely, in the low-beta pole, levodopa reversed the movement-related modulation from a positive FM to a negative FM (ANOVA interaction factor: P= 0.0003) (Fig. 3C). In the high-beta pole, the population exhibited a positive FM both before and after levodopa administration (1st main factor: P= 0.0001) (Fig. 3D). On average, levodopa had no significant effects on baseline frequencies. Latency analysis on all nuclei exhibiting significant movement-related FMs (either F(m) > F(b) or F(m) < F(b)) confirmed that the modulations started well before movement onset (low-beta: 0.63 ± 1.18 s before levodopa, n = 13, and 0.52 ± 0.74 s after levodopa, n = 13; high-beta: 1.05 ± 1.17 s before levodopa, n = 14 and 1.15 ± 1.04 s after levodopa, n = 13).

Discussion

The main finding of this study is that low-beta (13–20 Hz) and high-beta (20–35 Hz) rhythms detected from DBS electrodes implanted in the STN of patients with Parkinson's disease can represent movement-related activity not only in amplitude modulation, like a radio AM, but also in frequency modulation, like a radio FM. At the population level, the dopaminergic system regulates beta baseline amplitudes – specifically in the low-beta (13–20 Hz) rhythm – but have no consistent effect on their movement-related modulation. In contrast, the dopaminergic system does not affect beta baseline frequencies but regulates their movement-related modulation. Taken together, these results suggest that FM could represent a new domain for information processing and communication in the human STN, with specific reactivity to pharmacological and behavioural cues.

Methodological considerations

The concept of FM as used in this paper refers to the modulation of the frequency of neural oscillatory rhythms distributed in large populations of neurones and should not be confused with the modulation of single-neurone firing rates (Riehle et al. 1997). This difference should be taken into account when comparing our results at the LFP level with studies at the single-neurone level (Amirnovin et al. 2004). LFPs represent the synchronous presynaptic and postsynaptic activity of large populations of neurones and can detect focal network oscillatory behaviours that are not necessarily observable in single neurones or in neurone pairs (Creutzfeldt et al. 1966; Frost, 1968; Murthy & Fetz, 1992, 1996a, b; Baker et al. 1997; Donoghue et al. 1998; Magill et al. 2004; Goldberg et al. 2004). In the human STN, the local action of LFP beta oscillations is strongly supported by polarity reversals and amplitude gradients between adjacent electrode contacts (Brown et al. 2001; Levy et al. 2002a; Kühn et al. 2004; Doyle et al. 2005), and by the synchronization between LFP beta oscillations and single-neurone activity, reflecting focal topography (Levy et al. 2002a; Kühn et al. 2005). The possibility that these oscillations represent volume conduction from cortical or nearby brainstem regions is therefore unlikely. In fact, when simultaneous (differential) recordings are performed from several of these regions in humans (e.g. STN, globus pallidus internus (GPi) and cortex), the phase differences of beta oscillations between regions are typically consistent with synaptic transmission, being too long for volume conduction (Brown et al. 2001; Cassidy et al. 2002; Williams et al. 2002). Possible sources of variability involved with LFP recordings from DBS electrodes in human patients were discussed in our previous works and include the time elapsed from electrode implant, the clinical features of the patients and possible biases in the localization of the electrodes within the STN (Foffani et al. 2003, 2005b; Priori et al. 2004). Of note, the possible subtle variability in the localization of the electrodes could have contributed to the variability in the patterns of movement-related modulation described in the present study, which is consistent with the increase in complexity of STN processing that is emerging from LFP studies. As we previously remarked, ‘this variability is an important result by itself, as it suggests that the human STN is not a single homogeneous nuclear structure but it contains multiple, functionally segregated subsystems’ (Priori et al. 2004). Two additional aspects should be specifically considered in this study. Firstly, from the experimental point of view, we asked our patients to perform fine self-paced movements (second finger extensions), which require much smaller and more focal muscular activation than the hand-held joystick movements employed in previous studies (Cassidy et al. 2002; Doyle et al. 2005). Differences in muscular activation patterns are likely to produce differences in the STN LFP modulation patterns. Secondly, from the data analysis point of view, we employed a novel approach (adaptive autoregressive identification with spectral power decomposition, Foffani et al. 2004a) that describes the movement-related behaviour of LFP rhythms in terms of ‘poles’, which can vary both in amplitude and frequency, instead of the classical frequency bands, which can vary only in amplitude. This allowed us to study not only the movement-related AM but also the movement-related FM of subthalamic beta LFP activity. Nonetheless, it should be remarked that all studies based on LFP recordings from DBS electrodes, including the present work, are performed in patients with movement disorders and the extension to physiological conditions should be cautious.

Movement-related amplitude modulations (AMs)

In scalp EEG studies, the idea of event-related AM is almost as old as EEG itself. Since the pioneering works by Berger (1929), EEG rhythms have been described to be modulated in amplitude by a great variety of events, including attention, cognitive tasks and movement executions (Pfurtscheller & Lopes da Silva, 1999). Movement-related AM can be easily observed over the sensorimotor cortex, most consistently in the alpha, beta and gamma frequency ranges (Pfurtscheller & Aranibar, 1979; Defebvre et al. 1994; Toro et al. 1994; Defebvre et al. 1998; Crone et al. 1998a, b; Magnani et al. 1998; Wang et al. 1999; Pfurtscheller & Lopes da Silva, 1999; Ohara et al. 2000; Alegre et al. 2003; Devos et al. 2003; Pfurtscheller et al. 2003; Szurhaj et al. 2003; Alegre et al. 2004; Foffani et al. 2004a, b). Thanks to the advent of DBS for treating movement disorders, these concepts have been deepened into the human basal ganglia – in particular STN and GPi – through LFP recordings from electrodes implanted for DBS (Brown, 2003). Subthalamo-pallidal rhythms encompass a wide range of frequencies, from as low as a few Hz (Silberstein et al. 2003; Priori et al. 2004; Foffani et al. 2005a) to as high as 300 Hz (Foffani et al. 2003, 2005b). However, the highest degree of attention has been devoted to the beta frequency range (13–35 Hz). LFP beta oscillations in the human STN are tightly related to local single-unit activity (Levy et al. 2002a; Kühn et al. 2005), display long-range synchronization across multiple structures in the cortico-basal ganglia loop (Brown et al. 2001; Cassidy et al. 2002; Williams et al. 2002; Brovelli et al. 2004; Schnitzler & Gross, 2005; Foffani et al. 2005a), and are modulated in amplitude by both dopaminergic medication (Brown et al. 2001; Levy et al. 2002a; Priori et al. 2004) and movement execution (Foffani et al. 2002, 2004a; Levy et al. 2002a; Priori et al. 2002; Cassidy et al. 2002; Kühn et al. 2004; Williams et al. 2005; Doyle et al. 2005). Here we corroborated that the low-beta (13–20 Hz) rhythm has a greater baseline sensitivity to levodopa than the high-beta (20–35 Hz) rhythm (Priori et al. 2004), showed that movement-related AMs in these two rhythms can display different patterns, and confirmed that movement-related beta AM are present not only at off in the full parkinsonian state (Foffani et al. 2002; Cassidy et al. 2002), but also at on after dopaminergic medication, consistent with our preliminary observations (Priori et al. 2002) and with recent findings (Doyle et al. 2005). These results support the idea that beta rhythms can play an important – not only pathological – role in basal ganglia motor control, which is consistent with the movement-related beta AM observed in the globus pallidus internus (GPe) of epileptic patients (Sochurkova & Rektor, 2003) and with the widespread modulation of beta activity observed in the striatum of normal primates (Courtemanche et al. 2003).

Movement-related frequency modulations (FMs)

Very little evidence about event-related FM can be found in EEG studies. This probably has a methodological rather than physiological explanation, as in classical Fourier time-frequency analysis (or similar approaches) frequency and time are the independent variables whereas the spectral amplitude (or power) is the dependent variable. New methodological approaches recently allowed EEG researchers to observe phase-resetting phenomena in alpha oscillations (∼10 Hz) when studying brain dynamics in terms of event-related alterations of ongoing-rhythms (Makeig et al. 2002, 2004; Gruber et al. 2005). From a communication theory point of view, FM is a more general principle than phase-resetting, in the sense that phase-resetting is a sufficient but not necessary condition for FM. FM phenomena have also been observed at very-low frequencies (∼1 Hz) in the Limax olfactory lobe (Kimura et al. 1998; Cooke & Gelperin, 2001), at theta frequencies (3–10 Hz) in guinea pig inferior olive neurones (Benardo & Foster, 1986), in the rat hippocampus (Kirk, 1998) and in the primate occipitotemporal pathway (Purpura et al. 2003), and at beta/gamma frequencies (20–80 Hz) in the cat auditory cortex (Karmos et al. 2002; Lakatos et al. 2004), in the rabbit sensory cortices (Freeman & Rogers, 2002) and, most notably, in the rat subthalamic nucleus (Sharott et al. 2005). This evidence – although sparse, non-systematic and mostly related to sensory processing – suggests that FM could serve as a general principle for brain information processing and communication. By combining adaptive autoregressive identification with spectral power decomposition (Foffani et al. 2004a) we found that FMs could contribute to the involvement of the human STN in movement preparation, execution and recovery (Foffani et al. 2002, 2003; Priori et al. 2002; Levy et al. 2002a; Cassidy et al. 2002; Paradiso et al. 2003; Kühn et al. 2004; Amirnovin et al. 2004; Williams et al. 2005; Doyle et al. 2005). Moreover, we showed that dopaminergic medication induces significant changes in the FM patterns, particularly in the low-beta (13–20 Hz) rhythm. Even though the FMs could be an epiphenomenon of movement-related network changes (i.e. of the AMs), the different effects induced by levodopa on beta frequencies (movement-related modulation) compared to beta amplitudes (baseline modulation) suggest that FM could represent a new domain for information processing in the human basal ganglia.

Neurophysiological significance

The classical basal ganglia model (Albin, 1995; Brooks, 1995; Chesselet & Delfs, 1996; Wichmann & Delong, 1996) is exclusively based on firing rates, which are modulated in each nucleus by excitatory or inhibitory inputs from other nuclei. The limits of the classical model are apparent, and it is becoming progressively clear that rhythms and oscillations play a primary role in the pathophysiology of the human basal ganglia (Brown & Marsden, 1998; Wichmann & Delong, 1999; Bevan et al. 2002; Brown, 2003; Dovstrovsky & Bergman, 2004; Hutchison et al. 2004). Alternative rhythm-based models of basal ganglia pathophysiology are emerging from LFP studies in DBS patients (Brown, 2003; Priori et al. 2004). Intriguingly, Hoppensteadt and Izhikevich mathematically demonstrated that FM is a general principle of interaction between brain structures when the following two main assumptions are verified: (1) each structure can be considered as an autonomous oscillator, and (2) oscillators are weakly connected (Hoppensteadt & Izhikevich 1997, 1998; Izhikevich, 1999). Their study was focused on thalamo-cortical interactions, but it can be easily generalized to cortico-basal ganglia processing. In this circuit, the assumption about autonomous oscillations is supported by modelling studies suggesting that ‘different parts of the STN-GPe system may oscillate separately from each other’ (Terman et al. 2002), by experimental work in vitro showing the autonomous oscillatory behaviour of single cells in the STN–GP circuit (Plenz & Kital, 1999; Beurrier et al. 1999; Stanford, 2003) and by the evidence of independent oscillators in the pathophysiology of tremor (Hurtado et al. 1999, 2000, 2005). The assumption of weak connections is experimentally verified in the human basal ganglia, as LFP beta oscillations are transmitted throughout the STN–GP–cortical circuit (Brown et al. 2001; Cassidy et al. 2002; Williams et al. 2002; Foffani et al. 2005a), but pairs of single neurones in STN, GPi and GPe are not necessarily correlated at beta frequencies (Levy et al. 2000, 2002a, b). According to Hoppensteadt and Izhikevich, the FM interaction between brain structures/oscillators occurs when there is a certain nearly resonant relation between their frequencies, which well agrees with FMs in the order of ∼1 Hz as observed in our study. Our finding that beta subthalamic rhythms can represent information not only in AM but also in FM therefore suggests that movement execution is associated not only with linear modulations of ongoing rhythms, but also with non-linear interactions between nearly resonant, independent and weakly coupled oscillators in the cortico-basal ganglia circuit. Hence, movement-related AMs and FMs in the human STN might reflect the presence of linear and non-linear communication channels, which could separately or complementarily contribute to subthalamic information processing. The actual way by which FMs might exert their putative mechanistic influence is suggested by Hoppensteadt & Izhikevich (1998): ‘An entire cortex can be dynamically partitioned into interwoven, but relatively independent, ensembles that process information without cross-interference. A cortical oscillator may participate in different ensembles by changing its frequency, without changing the strengths of synaptic connections…. The functional partition of the cortex may be dynamically reorganized by a rhythmic input from the thalamus. Indeed, the thalamus can link any “non-interacting” cortical columns if it has an appropriate frequency’. These considerations can be easily extended to the STN: if multiple beta rhythms operate in the human STN (Terman et al. 2002) and control different body parts in a somatotopically segregated fashion (Rodriguez-Oroz et al. 2001; Romanelli et al. 2005), then movement-related FMs could serve to regulate the degree of synchrony between rhythms and therefore between body parts for coordinated movement control. The excitatory nature of the human STN and its widespread efferent connectivity to most of the basal ganglia and other structures in the brainstem (Hamani et al. 2004) suggest that beta FMs could be relevant for the ‘normal and pathological oscillatory communication in the brain’ (Brown & Marsden, 1998; Brown, 2003; Schnitzler & Gross, 2005). In conclusion, the subthalamic FM mode provides a novel informational domain for rhythm-based pathophysiological models of cortico-basal ganglia processing.

Acknowledgments

The authors whish to thank Dr G. Ardolino, Dr B. Bossi, Dr S. Barbieri, Dr E. Caputo, Professor Ing. S. Cerutti, Dr M. Egidi, Dr M. Locatelli, Dr K. A. Moxon, Dr A. Pesenti, Dr P. Rampini, Dr F. Tamma for their helpful cooperation, and S. Garlaschi, G. Gherardi, A. Marsilio, M. Pastori, for their assistance. The study was supported by IRCCS Ospedale Maggiore di Milano, Centro Dino Ferrari for Neurodegenerative Disorders (Milan, Italy), Ministero della Sanità (Rome, Italy), Ministero dell'Università e della Ricerca Scientifica e Tecnologica (Rome, Italy), Fundación Hospital Nacional Parapléjicos para la Investigación y la Integración (Toledo, Spain), and National Parkinson Foundation (Miami, FL, USA).

References

  1. Albin R. The pathophysiology of chorea/ballism and Parkinsonism. Parkinonism Relat Disord. 1995;1:3–11. doi: 10.1016/1353-8020(95)00011-t. [DOI] [PubMed] [Google Scholar]
  2. Alegre M, de Gurtubay IG, Labarga A, Iriarte J, Malanda A, Artieda J. Alpha and beta oscillatory activity during a sequence of two movements. Clin Neurophysiol. 2004;115:124–130. doi: 10.1016/s1388-2457(03)00311-0. [DOI] [PubMed] [Google Scholar]
  3. Alegre M, Gurtubay IG, Labarga A, Iriarte J, Malanda A, Artieda J. Alpha and beta oscillatory changes during stimulus-induced movement paradigms: effect of stimulus predictability. Neuroreport. 2003;14:381–385. doi: 10.1097/00001756-200303030-00017. [DOI] [PubMed] [Google Scholar]
  4. Amirnovin R, Williams ZM, Cosgrove GR, Eskandar EN. Visually guided movements suppress subthalamic oscillations in Parkinson's disease patients. J Neurosci. 2004;24:11302–11306. doi: 10.1523/JNEUROSCI.3242-04.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Baker SN, Olivier E, Lemon RN. Coherent oscillations in monkey motor cortex and hand muscle EMG show task-dependent modulation. J Physiol. 1997;501:225–241. doi: 10.1111/j.1469-7793.1997.225bo.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bar-Gad I, Bergman H. Stepping out of the box: information processing in the neural networks of the basal ganglia. Curr Opin Neurobiol. 2001;11:689–695. doi: 10.1016/s0959-4388(01)00270-7. [DOI] [PubMed] [Google Scholar]
  7. Bar-Gad I, Morris G, Bergman H. Information processing, dimensionality reduction and reinforcement learning in the basal ganglia. Prog Neurobiol. 2003;71:439–473. doi: 10.1016/j.pneurobio.2003.12.001. [DOI] [PubMed] [Google Scholar]
  8. Baselli G, Porta A, Rimoldi O, Pagani M, Cerutti S. Spectral decomposition in multichannel recordings based on multivariate parametric identification. IEEE Trans Biomed Eng. 1997;44:1092–1101. doi: 10.1109/10.641336. [DOI] [PubMed] [Google Scholar]
  9. Berger H. Über das Elektroenkephalogram des Menschen. Arch Psychiatr Nervenkr. 1929;87:527–570. [Google Scholar]
  10. Benardo LS, Foster RE. Oscillatory behavior in inferior olive neurons: mechanism, modulation, cell aggregates. Brain Res Bull. 1986;17:773–784. doi: 10.1016/0361-9230(86)90089-4. [DOI] [PubMed] [Google Scholar]
  11. Beurrier C, Congar P, Bioulac B, Hammond C. Subthalamic nucleus neurons switch from single-spike activity to burst-firing mode. J Neurosci. 1999;19:599–609. doi: 10.1523/JNEUROSCI.19-02-00599.1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Bevan MD, Magill PJ, Terman D, Bolam JP, Wilson CJ. Move to the rhythm: oscillations in the subthalamic nucleus-external globus pallidus network. Trends Neurosci. 2002;25:525–531. doi: 10.1016/s0166-2236(02)02235-x. [DOI] [PubMed] [Google Scholar]
  13. Bianchi AM, Mainardi LT, Cerutti S. The whale forgetting factor in recursive AR spectral analysis of heart rate variability signals. Meth Inf Med. 1997;36:241–245. [PubMed] [Google Scholar]
  14. Brovelli A, Ding M, Ledberg A, Chen Y, Nakamura R, Bressler SL. Beta oscillations in a large-scale sensorimotor cortical network: directional influences revealed by Granger causality. Proc Natl Acad Sci U S A. 2004;101:9849–9854. doi: 10.1073/pnas.0308538101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Brooks DJ. The role of basal ganglia in motor control: contributions from PET. J Neurol Sci. 1995;128:1–13. doi: 10.1016/0022-510x(94)00206-4. [DOI] [PubMed] [Google Scholar]
  16. Brown P. Oscillatory nature of human basal ganglia activity: relationship to the pathophysiology of Parkinson's disease. Mov Disord. 2003;18:357–363. doi: 10.1002/mds.10358. [DOI] [PubMed] [Google Scholar]
  17. Brown P, Marsden CD. What do the basal ganglia do? Lancet. 1998;351:1801–1804. doi: 10.1016/s0140-6736(97)11225-9. [DOI] [PubMed] [Google Scholar]
  18. Brown P, Oliviero A, Mazzone P, Insola A, Tonali P, Di Lazzaro V. Dopamine dependency of oscillations between subthalamic nucleus and pallidum in Parkinson's disease. J Neurosci. 2001;21:1033–1038. doi: 10.1523/JNEUROSCI.21-03-01033.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Cassidy M, Mazzone P, Oliviero A, Insola A, Tonali P, Di Lazzaro V, Brown P. Movement-related changes in synchronization in the human basal ganglia. Brain. 2002;125:1235–1246. doi: 10.1093/brain/awf135. [DOI] [PubMed] [Google Scholar]
  20. Chesselet MF, Delfs JM. Basal ganglia and movement disorders: an update. Trends Neurosci. 1996;19:417–422. doi: 10.1016/0166-2236(96)10052-7. [DOI] [PubMed] [Google Scholar]
  21. Clochon P, Fontbonne J, Lebrun N, Etevenon P. A new method for quantifying EEG event-related desynchronization: amplitude envelope analysis. Electroencephalogr Clin Neurophysiol. 1996;98:126–129. doi: 10.1016/0013-4694(95)00192-1. [DOI] [PubMed] [Google Scholar]
  22. Cooke IR, Gelperin A. In vivo recordings of spontaneous and odor-modulated dynamics in the Limax olfactory lobe. J Neurobiol. 2001;46:126–141. [PubMed] [Google Scholar]
  23. Courtemanche R, Fujii N, Graybiel AM. Synchronous, focally modulated beta-band oscillations characterize local field potential activity in the striatum of awake behaving monkeys. J Neurosci. 2003;23:11741–11752. doi: 10.1523/JNEUROSCI.23-37-11741.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Creutzfeldt OD, Watanabe S, Lux HD. Relations between EEG phenomena and potentials of single cortical cells. I. Evoked responses after thalamic and erpicortical stimulation. Electroencephalogr Clin Neurophysiol. 1966;20:1–18. doi: 10.1016/0013-4694(66)90136-2. [DOI] [PubMed] [Google Scholar]
  25. Crone NE, Miglioretti DL, Gordon B, Lesser RP. Functional mapping of human sensorimotor cortex with electrocorticographic spectral analysis. II. Event-related synchronization in the gamma band. Brain. 1998b;121:2301–2315. doi: 10.1093/brain/121.12.2301. [DOI] [PubMed] [Google Scholar]
  26. Crone NE, Miglioretti DL, Gordon B, Sieracki JM, Wilson MT, Uematsu S, Lesser RP. Functional mapping of human sensorimotor cortex with electrocorticographic spectral analysis. I. Alpha and beta event-related desynchronization. Brain. 1998a;121:2271–2299. doi: 10.1093/brain/121.12.2271. [DOI] [PubMed] [Google Scholar]
  27. David O, Harrison L, Friston KJ. Modelling event-related responses in the brain. Neuroimage. 2005;25:756–770. doi: 10.1016/j.neuroimage.2004.12.030. [DOI] [PubMed] [Google Scholar]
  28. Defebvre L, Bourriez JL, Derambure P, Duhamel A, Guieu JD, Destee A. Influence of chronic administration of L-DOPA on event-related desynchronization of mu rhythm preceding voluntary movement in Parkinson's disease. Electroencephalogr Clin Neurophysiol. 1998;109:161–167. doi: 10.1016/s0924-980x(97)00085-4. [DOI] [PubMed] [Google Scholar]
  29. Defebvre L, Bourriez JL, Dujardin K, Derambure P, Destee A, Guieu JD. Spatiotemporal study of Bereitschaftspotential and event-related desynchronization during voluntary movement in Parkinson's disease. Brain Topogr. 1994;6:237–244. doi: 10.1007/BF01187715. [DOI] [PubMed] [Google Scholar]
  30. Devos D, Labyt E, Derambure P, Bourriez JL, Cassim F, Guieu JD, Destee A, Defebvre L. Effect of L-Dopa on the pattern of movement-related (de)synchronisation in advanced Parkinson's disease. Neurophysiol Clin. 2003;33:203–212. doi: 10.1016/j.neucli.2003.10.001. [DOI] [PubMed] [Google Scholar]
  31. Donoghue JP, Sanes JN, Hatsopoulos NG, Gaal G. Neural discharge and local field potential oscillations in primate motor cortex during voluntary movements. J Neurophysiol. 1998;79:159–173. doi: 10.1152/jn.1998.79.1.159. [DOI] [PubMed] [Google Scholar]
  32. Dostrovsky J, Bergman H. Oscillatory activity in the basal ganglia – relationship to normal physiology and pathophysiology. Brain. 2004;127:721–722. doi: 10.1093/brain/awh164. [DOI] [PubMed] [Google Scholar]
  33. Doyle LM, Kühn AA, Hariz M, Kupsch A, Schneider GH, Brown P. Levodopa-induced modulation of subthalamic beta oscillations during self-paced movements in patients with Parkinson's disease. Eur J Neurosci. 2005;21:1403–1412. doi: 10.1111/j.1460-9568.2005.03969.x. [DOI] [PubMed] [Google Scholar]
  34. Egidi M, Rampini P, Locatelli M, Farabola M, Priori A, Pesenti A, Tamma F, Caputo E, Chiesa V, Villani RM. Visualisation of the subthalamic nucleus: a multiple sequential image fusion (MuSIF) technique for direct stereotaxic localisation and postoperative control. Neurol Sci. 2002;23(Suppl. 2):S71–S72. doi: 10.1007/s100720200075. [DOI] [PubMed] [Google Scholar]
  35. Foffani G, Ardolino G, Meda B, Egidi M, Rampini P, Caputo E, Baselli G, Priori A. Altered subthalamo-pallidal synchronisation in parkinsonian dyskinesias. J Neurol Neurosurg Psychiatry. 2005a;76:426–428. doi: 10.1136/jnnp.2004.043547. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Foffani G, Ardolino G, Rampini P, Tamma F, Caputo E, Egidi M, Cerutti S, Barbieri S, Priori A. Physiological recordings from electrodes implanted in the basal ganglia for deep brain stimulation in Parkinson's disease. The of relevance of fast subthalamic rhythms. Acta Neurochir Suppl. 2005b;93:97–99. doi: 10.1007/3-211-27577-0_16. [DOI] [PubMed] [Google Scholar]
  37. Foffani G, Bianchi AM, Cincotti F, Babiloni C, Carducci F, Babiloni F, Rossini PM, Cerutti S. Independent component analysis compared to laplacian filtering as ‘Deblurring’ techniques for event related desynchronization/synchronization. Meth Inf Med. 2004b;43:74–78. [PubMed] [Google Scholar]
  38. Foffani G, Bianchi AM, Priori A, Baselli G. Adaptive autoregressive identification with spectral power decomposition for studying movement-related activity in scalp EEG signals and basal ganglia local field potentials. J Neural Eng. 2004a;1:165–173. doi: 10.1088/1741-2560/1/3/006. [DOI] [PubMed] [Google Scholar]
  39. Foffani G, Priori A, Egidi M, Rampini P, Tamma F, Caputo E, Moxon KA, Cerutti S, Barbieri S. 300-Hz subthalamic oscillations in Parkinson's disease. Brain. 2003;126:2153–2163. doi: 10.1093/brain/awg229. [DOI] [PubMed] [Google Scholar]
  40. Foffani G, Priori A, Rohr M, Pesenti A, Locatelli M, Caputo E, et al. Event-related desynchronization (ERD) in the human subthalamus and internal globus pallidus. J Physiol. 2002;539.P:39P. [Google Scholar]
  41. Freeman WJ, Rogers LJ. Fine temporal resolution of analytic phase reveals episodic synchronization by state transitions in gamma EEGs. J Neurophysiol. 2002;87:937–945. doi: 10.1152/jn.00254.2001. [DOI] [PubMed] [Google Scholar]
  42. Frost JD., Jr EEG-intracellular potential relationships in isolated cerebral cortex. Electroencephalogr Clin Neurophysiol. 1968;24:434–443. doi: 10.1016/0013-4694(68)90103-x. [DOI] [PubMed] [Google Scholar]
  43. Goldberg JA, Rokni U, Boraud T, Vaadia E, Bergman H. Spike synchronization in the cortex/basal-ganglia networks of Parkinsonian primates reflects global dynamics of the local field potentials. J Neurosci. 2004;24:6003–6010. doi: 10.1523/JNEUROSCI.4848-03.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Gruber WR, Klimesch W, Sauseng P, Doppelmayr M. Alpha phase synchronization predicts P1 and N1 latency and amplitude size. Cereb Cortex. 2005;15:371–377. doi: 10.1093/cercor/bhh139. [DOI] [PubMed] [Google Scholar]
  45. Hamani C, Saint-Cyr JA, Fraser J, Kaplitt M, Lozano AM. The subthalamic nucleus in the context of movement disorders. Brain. 2004;127:4–20. doi: 10.1093/brain/awh029. [DOI] [PubMed] [Google Scholar]
  46. Haykin S. Communication Systems. 3. New York: John Wiley & Sons; 1994. [Google Scholar]
  47. Hoppensteadt FC, Izhikevich EM. Weakly Connected Neural Networks. New York: Springer-Verlag; 1997. [Google Scholar]
  48. Hoppensteadt FC, Izhikevich EM. Thalamo-cortical interactions modeled by weakly connected oscillators: could the brain use FM radio principles? Biosystems. 1998;48:85–94. doi: 10.1016/s0303-2647(98)00053-7. [DOI] [PubMed] [Google Scholar]
  49. Hurtado JM, Gray CM, Tamas LB, Sigvardt KA. Dynamics of tremor-related oscillations in the human globus pallidus: a single case study. Proc Natl Acad Sci U S A. 1999;96:1674–1679. doi: 10.1073/pnas.96.4.1674. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Hurtado JM, Lachaux JP, Beckley DJ, Gray CM, Sigvardt KA. Inter- and intralimb oscillator coupling in parkinsonian tremor. Mov Disord. 2000;15:683–691. doi: 10.1002/1531-8257(200007)15:4<683::aid-mds1013>3.0.co;2-#. [DOI] [PubMed] [Google Scholar]
  51. Hurtado JM, Rubchinsky LL, Sigvardt KA, Wheelock VL, Pappas CT. Temporal evolution of oscillations and synchrony in GPi/muscle pairs in Parkinson's disease. J Neurophysiol. 2005;93:1569–1584. doi: 10.1152/jn.00829.2004. [DOI] [PubMed] [Google Scholar]
  52. Hutchison WD, Dostrovsky JO, Walters JR, Courtemanche R, Boraud T, Goldberg J, Brown P. Neuronal oscillations in the basal ganglia and movement disorders: evidence from whole animal and human recordings. J Neurosci. 2004;24:9240–9243. doi: 10.1523/JNEUROSCI.3366-04.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Izhikevich EM. Weakly connected quasiperiodic oscillators, FM interactions, and multiplexing in the brain. SIAM J Appl Mathematics. 1999;59:2193–2223. [Google Scholar]
  54. Jansen BH, Agarwal G, Hegde A, Boutros NN. Phase synchronization of the ongoing EEG and auditory EP generation. Clin Neurophysiol. 2003;114:79–85. doi: 10.1016/s1388-2457(02)00327-9. [DOI] [PubMed] [Google Scholar]
  55. Karmos G, Lakatos P, Pincze Z, Rajkai C, Ulbert I. Frequency of gamma activity is modulated by motivation in the auditory cortex of cat. Acta Biol Hung. 2002;53:473–483. doi: 10.1556/ABiol.53.2002.4.8. [DOI] [PubMed] [Google Scholar]
  56. Kimura T, Toda S, Sekiguchi T, Kirino Y. Behavioral modulation induced by food odor aversive conditioning and its influence on the olfactory responses of an oscillatory brain network in the slug Limax marginatus. Learn Mem. 1998;4:365–375. doi: 10.1101/lm.4.5.365. [DOI] [PubMed] [Google Scholar]
  57. Kirk IJ. Frequency modulation of hippocampal theta by the supramammillary nucleus, and other hypothalamo–hippocampal interactions: mechanisms and functional implications. Neurosci Biobehav Rev. 1998;22:291–302. doi: 10.1016/s0149-7634(97)00015-8. [DOI] [PubMed] [Google Scholar]
  58. Kühn AA, Trottenberg T, Kivi A, Kupsch A, Schneider G, Brown P. The relationship between local field potential and neuronal discharge in the subthalamic nucleus of patients with Parkinson's disease. Exp Neurol. 2005;194:212–220. doi: 10.1016/j.expneurol.2005.02.010. [DOI] [PubMed] [Google Scholar]
  59. Kühn AA, Williams D, Kupsch A, Limousin P, Hariz M, Schneider GH, Yarrow K, Brown P. Event-related beta desynchronization in human subthalamic nucleus correlates with motor performance. Brain. 2004;127:735–746. doi: 10.1093/brain/awh106. [DOI] [PubMed] [Google Scholar]
  60. L.I.M.P.E. Guidelines for the treatment of Parkinson's disease. Neurol Sci. 2003;24(S3):S203–S204. [Google Scholar]
  61. Lakatos P, Szilagyi N, Pincze Z, Rajkai C, Ulbert I, Karmos G. Attention and arousal related modulation of spontaneous gamma-activity in the auditory cortex of the cat. Brain Res Cogn Brain Res. 2004;19:1–9. doi: 10.1016/j.cogbrainres.2003.10.023. [DOI] [PubMed] [Google Scholar]
  62. Levy R, Ashby P, Hutchison WD, Lang AE, Lozano AM, Dostrovsky JO. Dependence of subthalamic nucleus oscillations on movement and dopamine in Parkinson's disease. Brain. 2002a;125:1196–1209. doi: 10.1093/brain/awf128. [DOI] [PubMed] [Google Scholar]
  63. Levy R, Hutchison WD, Lozano AM, Dostrovsky JO. High-frequency synchronization of neuronal activity in the subthalamic nucleus of parkinsonian patients with limb tremor. J Neurosci. 2000;20:7766–7775. doi: 10.1523/JNEUROSCI.20-20-07766.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Levy R, Hutchison WD, Lozano AM, Dostrovsky JO. Synchronized neuronal discharge in the basal ganglia of parkinsonian patients is limited to oscillatory activity. J Neurosci. 2002b;22:2855–2861. doi: 10.1523/JNEUROSCI.22-07-02855.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Magill PJ, Sharott A, Bevan MD, Brown P, Bolam JP. Synchronous unit activity and local field potentials evoked in the subthalamic nucleus by cortical stimulation. J Neurophysiol. 2004;92:700–714. doi: 10.1152/jn.00134.2004. [DOI] [PubMed] [Google Scholar]
  66. Magnani G, Cursi M, Leocani L, Volonte MA, Locatelli T, Elia A, Comi G. Event-related desynchronization to contingent negative variation and self-paced movement paradigms in Parkinson's disease. Mov Disord. 1998;13:653–660. doi: 10.1002/mds.870130408. [DOI] [PubMed] [Google Scholar]
  67. Makeig S, Delorme A, Westerfield M, Jung TP, Townsend J, Courchesne E, Sejnowski TJ. Electroencephalographic brain dynamics following manually responded visual targets. PLoS Biol. 2004;2:e176. doi: 10.1371/journal.pbio.0020176. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Makeig S, Westerfield M, Jung TP, Enghoff S, Townsend J, Courchesne E, Sejnowski TJ. Dynamic brain sources of visual evoked responses. Science. 2002;295:690–694. doi: 10.1126/science.1066168. [DOI] [PubMed] [Google Scholar]
  69. Murthy VN, Fetz EE. Coherent 25- to 35-Hz oscillations in the sensorimotor cortex of awake behaving monkeys. Proc Natl Acad Sci U S A. 1992;89:5670–5674. doi: 10.1073/pnas.89.12.5670. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Murthy VN, Fetz EE. Synchronization of neurons during local field potential oscillations in sensorimotor cortex of awake monkeys. J Neurophysiol. 1996a;76:3968–3982. doi: 10.1152/jn.1996.76.6.3968. [DOI] [PubMed] [Google Scholar]
  71. Murthy VN, Fetz EE. Oscillatory activity in sensorimotor cortex of awake monkeys: synchronization of local field potentials and relation to behavior. J Neurophysiol. 1996b;76:3949–3967. doi: 10.1152/jn.1996.76.6.3949. [DOI] [PubMed] [Google Scholar]
  72. Ohara S, Ikeda A, Kunieda T, Yazawa S, Baba K, Nagamine T, Taki W, Hashimoto N, Mihara T, Shibasaki H. Movement-related change of electrocorticographic activity in human supplementary motor area proper. Brain. 2000;123:1203–1215. doi: 10.1093/brain/123.6.1203. [DOI] [PubMed] [Google Scholar]
  73. Paradiso G, Saint-Cyr JA, Lozano AM, Lang AE, Chen R. Involvement of the human subthalamic nucleus in movement preparation. Neurology. 2003;61:1538–1545. doi: 10.1212/01.wnl.0000096021.28967.57. [DOI] [PubMed] [Google Scholar]
  74. Penny WD, Kiebel SJ, Kilner JM, Rugg MD. Event-related brain dynamics. Trends Neurosci. 2002;25:387–389. doi: 10.1016/s0166-2236(02)02202-6. [DOI] [PubMed] [Google Scholar]
  75. Pfurtscheller G, Aranibar A. Evaluation of event-related desynchronization (ERD) preceding and following voluntary self-paced movement. Electroencephalogr Clin Neurophysiol. 1979;46:138–146. doi: 10.1016/0013-4694(79)90063-4. [DOI] [PubMed] [Google Scholar]
  76. Pfurtscheller G, Graimann B, Huggins JE, Levine SP, Schuh LA. Spatiotemporal patterns of beta desynchronization and gamma synchronization in corticographic data during self-paced movement. Clin Neurophysiol. 2003;114:1226–1236. doi: 10.1016/s1388-2457(03)00067-1. [DOI] [PubMed] [Google Scholar]
  77. Pfurtscheller G, Lopes da Silva FH. Event-related EEG/MEG synchronization and desynchronization: basic principles. Clin Neurophysiol. 1999;110:1842–1857. doi: 10.1016/s1388-2457(99)00141-8. [DOI] [PubMed] [Google Scholar]
  78. Plenz D, Katai ST. A basal ganglia pacemaker formed by the subthalamic nucleus and external globus pallidus. Nature. 1999;400:677–682. doi: 10.1038/23281. [DOI] [PubMed] [Google Scholar]
  79. Priori A, Egidi M, Pesenti A, Rohr M, Rampini P, Locatelli M, Tamma F, Caputo E, Chiesa V, Barbieri S. Do intraoperative microrecordings improve subthalamic nucleus targeting in stereotactic neurosurgery for Parkinson's disease? J Neurosurg Sci. 2003;47:56–60. [PubMed] [Google Scholar]
  80. Priori A, Foffani G, Pesenti A, Bianchi A, Chiesa V, Baselli G, Caputo E, Tamma F, Rampini P, Egidi M, Locatelli M, Barbieri S, Scarlato G. Movement-related modulation of neural activity in human basal ganglia and its L-DOPA dependency: recordings from deep brain stimulation electrodes in patients with Parkinson's disease. Neurol Sci. 2002;23(Suppl. 2):S101–S102. doi: 10.1007/s100720200089. [DOI] [PubMed] [Google Scholar]
  81. Priori A, Foffani G, Pesenti A, Tamma F, Bianchi AM, Pellegrini M, Locatelli M, Moxon KA, Villani RM. Rhythm-specific pharmacological modulation of subthalamic activity in Parkinson's disease. Exp Neurol. 2004;189:369–379. doi: 10.1016/j.expneurol.2004.06.001. [DOI] [PubMed] [Google Scholar]
  82. Purpura KP, Kalik SF, Schiff ND. Analysis of perisaccadic field potentials in the occipitotemporal pathway during active vision. J Neurophysiol. 2003;90:3455–3478. doi: 10.1152/jn.00011.2003. [DOI] [PubMed] [Google Scholar]
  83. Rampini PM, Locatelli M, Alimehmeti R, Tamma F, Caputo E, Priori A, Pesenti A, Rohr M, Egidi M. Multiple sequential image-fusion and direct MRI localisation of the subthalamic nucleus for deep brain stimulation. J Neurosurg Sci. 2003;47:33–39. [PubMed] [Google Scholar]
  84. Riehle A, Grun S, Diesmann M, Aertsen A. Spike synchronization and rate modulation differentially involved in motor cortical function. Science. 1997;278:1950–1953. doi: 10.1126/science.278.5345.1950. [DOI] [PubMed] [Google Scholar]
  85. Rodriguez-Oroz MC, Rodriguez M, Guridi J, Mewes K, Chockkman V, Vitek J, DeLong MR, Obeso JA. The subthalamic nucleus in Parkinson's disease: somatotopic organization and physiological characteristics. Brain. 2001;124:1777–1790. doi: 10.1093/brain/124.9.1777. [DOI] [PubMed] [Google Scholar]
  86. Romanelli P, Esposito V, Schaal DW, Heit G. Somatotopy in the basal ganglia: experimental and clinical evidence for segregated sensorimotor channels. Brain Res Brain Res Rev. 2005;48:112–128. doi: 10.1016/j.brainresrev.2004.09.008. [DOI] [PubMed] [Google Scholar]
  87. Salinas E, Sejnowski T. Correlated neuronal activity and the flow of neural information. Nat Rev Neurosci. 2001;2:539–550. doi: 10.1038/35086012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  88. Schnitzler A, Gross J. Normal and pathological oscillatory communication in the brain. Nat Rev Neurosci. 2005;6:285–296. doi: 10.1038/nrn1650. [DOI] [PubMed] [Google Scholar]
  89. Sharott A, Magill PJ, Harnack D, Kupsch A, Meissner W, Brown P. Dopamine depletion increases the power and coherence of beta-oscillations in the cerebral cortex and subthalamic nucleus of the awake rat. Eur J Neurosci. 2005;21:1413–1422. doi: 10.1111/j.1460-9568.2005.03973.x. [DOI] [PubMed] [Google Scholar]
  90. Silberstein P, Kühn AA, Kupsch A, Trottenberg T, Krauss JK, Wohrle JC, Mazzone P, Insola A, Di Lazzaro V, Oliviero A, Aziz T, Brown P. Patterning of globus pallidus local field potentials differs between Parkinson's disease and dystonia. Brain. 2003;126:2597–2608. doi: 10.1093/brain/awg267. [DOI] [PubMed] [Google Scholar]
  91. Sochurkova D, Rektor I. Event-related desynchronization/synchronization in the putamen. An SEEG case study. Exp Brain Res. 2003;149:401–404. doi: 10.1007/s00221-003-1371-2. [DOI] [PubMed] [Google Scholar]
  92. Stanford IM. Independent neuronal oscillators of the rat globus pallidus. J Neurophysiol. 2003;89:1713–1717. doi: 10.1152/jn.00864.2002. [DOI] [PubMed] [Google Scholar]
  93. Szurhaj W, Derambure P, Labyt E, Cassim F, Bourriez JL, Isnard J, Guieu JD, Mauguiere F. Basic mechanisms of central rhythms reactivity to preparation and execution of a voluntary movement: a stereoelectroencephalographic study. Clin Neurophysiol. 2003;114:107–119. doi: 10.1016/s1388-2457(02)00333-4. [DOI] [PubMed] [Google Scholar]
  94. Terman D, Rubin JE, Yew AC, Wilson CJ. Activity patterns in a model for the subthalamopallidal network of the basal ganglia. J Neurosci. 2002;22:2963–2976. doi: 10.1523/JNEUROSCI.22-07-02963.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  95. Toro C, Deuschl G, Thatcher R, Sato S, Kufta C, Hallett M. Event-related desynchronization and movement-related cortical potentials on the ECoG and EEG. Electroencephalogr Clin Neurophysiol. 1994;93:380–389. doi: 10.1016/0168-5597(94)90126-0. [DOI] [PubMed] [Google Scholar]
  96. Varela F, Lachaux JP, Rodriguez E, Martinerie J. The brainweb: phase synchronization and large-scale integration. Nat Rev Neurosci. 2001;2:229–239. doi: 10.1038/35067550. [DOI] [PubMed] [Google Scholar]
  97. Wang HC, Lees AJ, Brown P. Impairment of EEG desynchronisation before and during movement and its relation to bradykinesia in Parkinson's disease. J Neurol Neurosurg Psychiatry. 1999;66:442–446. doi: 10.1136/jnnp.66.4.442. [DOI] [PMC free article] [PubMed] [Google Scholar]
  98. Wichmann T, DeLong MR. Functional and pathophysiological models of the basal ganglia. Curr Opin Neurobiol. 1996;6:751–758. doi: 10.1016/s0959-4388(96)80024-9. [DOI] [PubMed] [Google Scholar]
  99. Wichmann T, DeLong MR. Oscillations in the basal ganglia. Nature. 1999;400:621–622. doi: 10.1038/23148. [DOI] [PubMed] [Google Scholar]
  100. Wichmann T, DeLong MR. Functional neuroanatomy of the basal ganglia in Parkinson's disease. Adv Neurol. 2003;91:9–18. [PubMed] [Google Scholar]
  101. Williams D, Kühn A, Kupsch A, Tijssen M, van Bruggen G, Speelman H, Hotton G, Loukas C, Brown P. The relationship between oscillatory activity and motor reaction time in the parkinsonian subthalamic nucleus. Eur J Neurosci. 2005;21:249–258. doi: 10.1111/j.1460-9568.2004.03817.x. [DOI] [PubMed] [Google Scholar]
  102. Williams D, Kühn A, Kupsch A, Tijssen M, van Bruggen G, Speelman H, Hotton G, Yarrow K, Brown P. Behavioural cues are associated with modulations of synchronous oscillations in the human subthalamic nucleus. Brain. 2003;126:1975–1985. doi: 10.1093/brain/awg194. [DOI] [PubMed] [Google Scholar]
  103. Williams D, Tijssen M, van Bruggen G, Bosch A, Insola A, Di Lazzaro V, Mazzone P, Oliviero A, Quartarone A, Speelman H, Brown P. Dopamine-dependent changes in the functional connectivity between basal ganglia and cerebral cortex in humans. Brain. 2002;125:1558–1569. doi: 10.1093/brain/awf156. [DOI] [PubMed] [Google Scholar]

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES