Abstract

Subthalamic nucleus (STN) deep brain stimulation (DBS) has become an accepted treatment for the motor manifestations of Parkinson disease (PD). The beneficial motor effects of STN DBS are likely due to modulation of BG output to frontal cortical regions associated with motor control, but the underlying neurophysiology of STN DBS effects, especially at the level of the cortex, is not well understood. In this study, we examined the effects of STN DBS on motor disability and visual working memory, a cognitive process supported by pFC. We tested 10 PD participants off medications, ON and OFF stimulation, along with 20 normal controls on a visual working memory task while simultaneously recording cortical EEG. In the OFF state, PD patients had poor motor function, were slower and less accurate in performing the working memory task, and had greater amplitudes and shorter latencies of the N200 ERP response. DBS improved clinical motor function, reduced N200 amplitudes, and increased N200 latencies but had little effect on working memory performance. We conclude that STN DBS normalizes neurophysiological activity in fronto striatal circuits and this may independently affect motor and cognitive function.

INTRODUCTION

Working memory is the temporary storage and manipulation of information necessary for many higher order cognitive tasks (Baddeley, 1986). Working memory consists of several component processes including initial sensory processing, short-term storage, continuous upgrade of information, and an executive control system for manipulation or retrieval (Baddeley, 2000). Executive components of working memory, especially those related to response selection, appear to rely critically on the pFC (Rowe, Toni, Josephs, Frackowiak, & Passingham, 2000), which are linked, perhaps by attentional control (Postle, 2006), to distributed cortical regions in a functional network (Mesulam, 2000). Most of pFC participates in mixed parallel and integrative BG-cortical circuits in which information from functionally distinct regions of pFC (Haber, Kim, Mailly, & Calzavara, 2006; Haber, Fudge, & McFarland, 2000) and other regions of cortex converge to inform such behaviors as response selection (Schroll, Vitay, & Hamker, 2012; Aron et al., 2007) and working memory (McNab & Klingberg, 2008). The subthalamic nucleus (STN) is part of the hyperdirect pathway, which connects regions of frontal cortex to the BG and may be critical in supporting integrative functions of BG-thalamo-cortical circuits through rapid inhibition of thalamo-cortical pathways (Mink, 1996).

Parkinson disease (PD) is a disorder of BG circuit function, and cognitive impairments, including working memory deficits in PD, are thought to result from loss of ascending dopaminergic projections to pFC terminal fields, particularly those in lateral pFC (Alexander, DeLong, & Strick, 1986). The first line approach to treating the cardinal motor manifestations of PD involves dopaminergic pharmacotherapy. Dopaminergic medications can also ameliorate cognitive and affective deficits (e.g., Cools, Barker, Sahakian, & Robbins, 2001; Owen, Iddon, Hodges, Summers, & Robbins, 1997). However, the effects of dopamine replacement on non-motor features are not consistent across studies (Lange et al., 1992; Gotham, Brown, & Marsden, 1988; Girotti et al., 1986) and the reasons for this are not fully understood. Similarly, modulation of cortical-BG circuits by chronic deep brain electrical stimulation (DBS) of the STN can significantly improve motor function and reduce motor disability in PD patients (Moro et al., 2010; Weaver et al., 2009; Deuschl et al., 2006; Rodriguez-Oroz et al., 2005), but again, effects on cognitive function, particularly working memory, are mixed. Whereas some investigators have reported working memory improvements (Mollion, Dominey, Broussolle, & Ventre-Dominey, 2011; Jahanshahi et al., 2000), others have shown either no change (Heo et al., 2008; Ardouin et al., 1999) or worsening of working memory (Hershey et al., 2004, 2008; Saint-Cyr, Trépanier, Kumar, Lozano, & Lang, 2000). These contradictory behavioral findings may be partially explained by PET studies of pFC following STN DBS (see Ballanger, Jahanshahi, Broussolle, & Thobois, 2009, for a review). Several PET studies have demonstrated increased blood flow to dorsolateral and motor regions of pFC in addition to the anterior cingulate following STN stimulation (Sestini et al., 2002; Limousin et al., 1997), which may account for the common motor benefits and also observations of working memory improvement following DBS. On the other hand, a more recent study (Hershey et al., 2003) showed that STN stimulation increased blood flow to midbrain regions, but this was associated with reduced blood flow to regions of bilateral frontal, parietal, and temporal cortex. These findings suggest that STN DBS may increase inhibitory output from BG to thalamo-cortical circuits and may underlie working memory impairments following STN DBS.

An alternate approach to understanding DBS effects on working memory is to measure the electrophysiology of cortical regions known to support this behavior. During visual working memory tasks ERP activity beginning around 200 msec post stimulus may reflect early, post sensory processes associated with stimulus unfamiliarity (Daffner et al., 2000) and may be important for encoding to and retrieval from the working memory buffer (Palva, Kulashekhar, Hamalainen, & Palva, 2011). The negative deflection occurring around 200 msec following the visual stimulus (N200) may be selectively modulated with the earlier P200 or later P300 deflection (Missonnier et al., 2003; Daffner et al., 2000) or may represent the earliest negativity associated with a more diffuse negative slow wave, a relative suppression of positive deflections, which starts around 200 msec post stimulus (Ruchkin, Johnson, Grafman, Canoune, & Ritter, 1997). Recent work has shown that these negativities, including the well-known contralateral delay activity, may reflect the maintenance of working memory in posterior cortical regions (Vogel & Machizawa, 2004). In these studies, the N200 peak observed during encoding is also seen at retrieval (Vogel & Machizawa, 2004) and the amplitude of the N200 is related to the working memory load (Palva et al., 2011). These observations suggest that early components of the visual working memory ERP, particularly the N200, may be strong markers of the underlying neurophysiological processes involved in working memory response selection.

In this study, we examined neurophysiological and behavioral effects of DBS on motor function and working memory performance. We simultaneously recorded cortical EEG while 10 PD and 20 control participants performed two versions of a visual serial probe recognition task. We hypothesized that putative behavioral effects of DBS would be reflected in the N200 response. We found that N200 responses of PD patients without DBS occurred significantly later and were of greater amplitude than the controls and these were related to poorer overall working memory performance. With DBS, clinical ratings of motor disability improved and N200 responses became significantly slower and the amplitude was significantly reduced such that they no longer differed from the age-matched control group on the more difficult working memory task. We conclude that STN DBS normalizes neurophysiological characteristics of BG-thalamo-cortical circuits with dissociable effects on motor and cognitive function.

METHODS

Participants

The study participants included 10 PD patients receiving bilateral STN stimulation and 20 healthy control participants. The control participants were volunteers from the local community or from the University of Rochester and included 10 younger control (YCS; mean age = 25 ± 3.9 years) and 10 older control participants (OCS; mean age = 68.8 ± 6.9 years) who did not differ in mean age from the PD group (mean age = 63.7 ± 8.1 years). All participants were right-handed except for one PD participant, and all had normal or corrected-to-normal vision. PD participants were all at least 3 months post stimulator activation and were on stable medication and stimulation parameters for at least two weeks before participation. All PD participants were programmed for optimum motor benefit before enrollment in the study. This typically involves targeting therapy for best motor symptom control with stimulator settings that minimize acutely evoked side effects (e.g., paresthesias, weakness of the limb or face, or visual changes). All participants provided informed consent and all study procedures were approved by the University of Rochester Research Subjects Review Board. Participant characteristics can be found in Table 1, PD clinical characteristics in Table 2, and PD stimulation parameters in Table 3.

Table 1. 

Participant Characteristics (Means and SDs)


Younger Control
Older Control
PD
n (sex) 10 (8 m/2 f) 10 (8 m/2 f) 10 (9 m/1 f) 
Age (years) 25 (3.9) 68.8 (6.9) 63.1 (8.7) 
Handedness Right (10) Right (10) Right (9), Left (1) 

Younger Control
Older Control
PD
n (sex) 10 (8 m/2 f) 10 (8 m/2 f) 10 (9 m/1 f) 
Age (years) 25 (3.9) 68.8 (6.9) 63.1 (8.7) 
Handedness Right (10) Right (10) Right (9), Left (1) 

m = male; f = female.

Table 2. 

PD Participant Clinical Characteristics

Participant
Sex
Age
Hand
Time since Initial Programming (years)
Duration of PD (years)
UPDRS ON
UPDRS OFF
Levodopa Equivalent Daily Dose
55 0.4 12.0 40 45 550 
59 1.2 4.5 21 36 
81 1.1 16.0 19 36 350 
63 7.2 15.5 16 21 850 
65 5.3 16.0 34 40 875 
63 0.3 16.5 19 29 320 
68 9.1 24.5 48 55 750 
59 9.3 20.5 20 30 1560 
49 3.9 12.0 16 29 1255 
10 69 0.4 4.5 22 36 780 
Participant
Sex
Age
Hand
Time since Initial Programming (years)
Duration of PD (years)
UPDRS ON
UPDRS OFF
Levodopa Equivalent Daily Dose
55 0.4 12.0 40 45 550 
59 1.2 4.5 21 36 
81 1.1 16.0 19 36 350 
63 7.2 15.5 16 21 850 
65 5.3 16.0 34 40 875 
63 0.3 16.5 19 29 320 
68 9.1 24.5 48 55 750 
59 9.3 20.5 20 30 1560 
49 3.9 12.0 16 29 1255 
10 69 0.4 4.5 22 36 780 

M = male; F = female; R = right; L = left.

Table 3. 

PD DBS Characteristics

Participant
Side
Voltage (V)
Pulse Width (μs)
Frequency (Hz)
Electrode Configuration
2.6 60 130 C+, 10− 
2.4 90 130 C+, 0− 
60 125 C+, 8− 
L 1 90 125 C+, 0− 
L 2a 0.5 60 125 C+, 3− 
2.3 60 135 C+, 4−, 5− 
3.1 60 135 C+, 0−, 1− 
120 185 6+, 4− 
3.4 90 185 3+, 1− 
3.5 60 145 C+, 0− 
2.9 90 145 C+, 2− 
2.2 60 125 C+, 9− 
L 1 1.8 60 125 C+, 0− 
L 2 0.8 60 125 C+, 3− 
3.3 60 130 C+, 1− 
2.6 60 130 C+, 1− 
1.6 60 130 2+, 3− 
1.3 60 135 1+, 0− 
1.8 90 130 C+, 1−, 2− 
2.6 90 130 C+, 2− 
10 R L 5.0 90 125 C+, 10− 
L 1 3.6 90 125 C+, 2− 
L 2 1.6 90 125 C+, 0− 
Participant
Side
Voltage (V)
Pulse Width (μs)
Frequency (Hz)
Electrode Configuration
2.6 60 130 C+, 10− 
2.4 90 130 C+, 0− 
60 125 C+, 8− 
L 1 90 125 C+, 0− 
L 2a 0.5 60 125 C+, 3− 
2.3 60 135 C+, 4−, 5− 
3.1 60 135 C+, 0−, 1− 
120 185 6+, 4− 
3.4 90 185 3+, 1− 
3.5 60 145 C+, 0− 
2.9 90 145 C+, 2− 
2.2 60 125 C+, 9− 
L 1 1.8 60 125 C+, 0− 
L 2 0.8 60 125 C+, 3− 
3.3 60 130 C+, 1− 
2.6 60 130 C+, 1− 
1.6 60 130 2+, 3− 
1.3 60 135 1+, 0− 
1.8 90 130 C+, 1−, 2− 
2.6 90 130 C+, 2− 
10 R L 5.0 90 125 C+, 10− 
L 1 3.6 90 125 C+, 2− 
L 2 1.6 90 125 C+, 0− 

R = right; L = left.

aTwo settings per side indicate interleaved stimulation via Activa PC implantable pulse generator.

Procedure

All participants completed the experimental paradigm in a single visit to the laboratory. Control participants completed testing in one session, whereas PD participants required two sessions separated by a 30-min rest period. PD participants were tested after withholding PD medications for at least 12 hr to nominally dissociate effects of dopamine replacement from direct effects of stimulation. PD participants were tested with bilateral stimulators ON in one session and OFF in the other. Stimulator activation order was randomized across the PD participants, with a 30-min wash-in/wash-out period after stimulation change before the second session. We measured motor disability in the PD patients using the clinician administered motor subscale of the Unified Parkinson Disease Rating Scale (UPDRS; Fahn, Elton, & UPDRS Program Members, 1987).

We used a two-alternative forced-choice modification of the serial probe recognition (SPR) task described by Sternberg (1966) to measure working memory. Participants viewed a series of two, four, six, or eight stimuli followed by a probe stimulus, which was indicated by a red perimeter (Figure 1). Using their dominant hand, participants indicated if the probe stimulus was in the just-viewed series by pressing the right button on a two-button response box and the left button if it was not. RT (msec) and accuracy were recorded. Stimuli were presented using the NeuroScan STIM (Compumedics USA, Charlotte, NC) platform and displayed on an LCD monitor 24 in. in front of the participants. SPR stimuli were displayed for 1000 msec with an ISI of 1000 msec. To minimize the effects of slowed RTs in the PD patients, all participants were given as much time as needed to respond to the probe and were required to initiate each trial with a button press.

Figure 1. 

Example of four-item set of the SPR task used to assess working memory. The object set included easily named objects (A) and the fractal set used fractal images (B). Stimuli remained on the screen for 1000 msec and the interstimulus interval was 1000 msec. The probe item was indicated with a red box in the task, and subjects responded to whether the probe was in the shown series with a two-alternative forced-choice button press.

Figure 1. 

Example of four-item set of the SPR task used to assess working memory. The object set included easily named objects (A) and the fractal set used fractal images (B). Stimuli remained on the screen for 1000 msec and the interstimulus interval was 1000 msec. The probe item was indicated with a red box in the task, and subjects responded to whether the probe was in the shown series with a two-alternative forced-choice button press.

We administered two versions of the SPR task that differed primarily in how easily the stimuli could be named. One version included line drawings of common objects (450 × 360 pixels in size and 300 dpi resolution) and a second version contained grayscale fractal images (512 × 384 pixels in size and 72 dpi resolution) that are more difficult to verbally encode (Ragland et al., 2002; Miyashita, Higuchi, Sakai, & Masui, 1991). We anticipated that the fractal SPR task would be more difficult and that by using both tasks we would cover the range of performance for the three participant groups. Participants completed a total of 32 trials in each task (eight trials each at the two-, four-, six-, and eight-item list lengths). The list length for each trial was randomized and the order of the images in each task was randomized. Two equivalent versions of each task (object and fractal) were used for the two testing sessions of the PD participants, and the order of these was counterbalanced across subjects.

During the SPR tasks, we simultaneously recorded cortical EEG using a SynAmps system (NeuroScan Compumedics USA, Charlotte, NC). The EEG was recorded using a 32-channel QuikCap with 30 nonpolarizable Ag/AgCl electrodes. Electrodes were located at the standard International 10–20 sites and included two referentially linked mastoids and a ground electrode near midline of the forehead. In addition, four electrodes were attached separately around the eyes to identify vertical and horizontal eye movement artifacts. Continuous EEG was recorded at a rate of 1 kHz with all impedances ≤10 kΩ (for controls and PD participants off stimulation) and a band-pass filter between 0.3 and 100 Hz. We chose to sample at 1 kHz to limit harmonic artifacts of stimulation in the lower frequency ranges of interest (Jeck et al., 2006). All data were collected in an enclosed laboratory with minimal ambient light and extraneous noise. EEG data were analyzed off-line in the MATLAB environment (Mathworks, Natick, MA) using the EEGLAB toolbox (Delorme & Makeig, 2004). Continuous EEG data were low-pass filtered (short infinite impulse response filter) at 50 Hz and then high-pass filtered (basic finite impulse response filter) at 0.5 Hz to further limit potential harmonic artifacts of the stimulating electrodes. Other artifacts including eye blinks, muscle contamination, and cardiac signals were removed using independent component analysis (Jung et al., 2000). All components were individually inspected and rejected if necessary. Although we collected data throughout the trial, we were most interested in the retrieval phase of the SPR task. The retrieval phase was defined from the onset of the probe stimulus and lasted 1 sec. Baseline activity included the 200 msec before probe onset.

From the EEG data, we first created time–frequency plots for each group to visually examine frequency-specific neurophysiological responses during the retrieval phase of the SPR tasks. Time–frequency plots were created by synchronizing neurophysiological activity from the EEG using a trigger pulse coincident with the onset of the probe stimulus at each trial. We collapsed these ERPs across list length for each participant and averaged these to create group time–frequency plots.

We used the ERPLAB toolbox (www.erpinfo.org/erplab/) to create averaged ERP data sets for each SPR task. EEG activity was synchronized to the probe onset trigger pulse and lasted for 1 sec. We filtered the ERP data sets at 10 Hz to minimize PD related pathological frequency artifacts that are known to be present in the low beta (12–18 Hz) range (Sagliocco, Meistrowitz, Schwendemann, Herrmann, & Basar-Eroglu, 2005; Green et al., 1996). All filtered ERP data sets were grand averaged and appended for group comparisons. From the ERP data, we extracted N200 peak amplitude, N200 latency, and N200 mean amplitude from midline channels (Fz, FCz, Cz, CPz, and Pz) in the time frame 200–350 msec following probe onset (after Nobre, Griffin, & Rao, 2007; Kusak, Grune, Hagendorf, & Metz, 2000) for analysis.

Data Analyses

We used the nonparametric sign test to examine the effects of DBS (ON, OFF) for the UPDRS. We conducted four separate analyses of variance (ANOVAs) to examine effects of Group (YCS, OCS, PD OFF) and Stimulus Load (two, four, six, or eight items) for the dependent variables (RT, accuracy) for each of the two versions of the SPR task (object, fractal). We used only correct trials for RT analyses. We also conducted four repeated-measures ANOVAs to examine the effect of Stimulation (OFF, ON) and Stimulus Load (two, four, six, or eight items) on RT and accuracy in the PD group for both SPR tasks. Finally, we repeated the RT and accuracy univariate ANOVAs to examine DBS effects relative to the other Groups (YCS, OCS, PD ON) by Stimulus Load (two, four, six, or eight items).

In a preliminary analysis, we found no significant differences between the five midline electrodes (Fz, FCz, Cz, Cpz, and Pz) for any dependent measure on either SPR task and there were no significant interactions between Group and Channel for any of the three dependent measures, so all five midline channels were included in the analysis of the neurophysiological data to increase overall power. For the neurophysiological data, we conducted two separate multivariate ANOVAs, one for each version of the SPR task (object, fractal) to compare Groups (YCS, OCS, PD OFF) for each dependent variable (N200 peak amplitude, N200 latency, and N200 mean amplitude). List length was not included as a within-subject factor because data across all list lengths were combined to derive the ERP-dependent measures. We used repeated-measures ANOVA to examine the effects of Stimulation (OFF, ON) on peak amplitude, latency, and mean amplitude in the PD group for the object and fractal SPR tasks. We then repeated the multivariate ANOVAs to examine DBS effects relative to the other groups (YCS, OCS, PD ON).

Finally, we used step-wise linear regression to examine the relationships between the behavioral and neurophysiological data. The dependent measures were accuracy and RT for the object and fractal tasks. Independent variables were the corresponding N200 peak amplitude, N200 latency, and N200 mean amplitude. All post hoc comparisons used Tukey's procedure and significance for all tests was set to p < .05. All statistical analyses were performed in SPSS v.17 (SPSS, Inc., Chicago, IL).

RESULTS

Behavioral Measures

The YCS, OCS, and PD OFF groups differed significantly in both accuracy, F(2, 948) = 11.347, p < .001, and RT, F(2, 861) = 77.479, p < .001, for the object version of the task (Figure 1A). The three groups also differed significantly in both accuracy, F(2, 948) = 20.249, p < .001, and RT, F(2, 948) = 45.180, p < .001, on the fractal version of the task (Figure 1B). In both versions of the task, the control groups were significantly more accurate and faster to respond than the PD OFF group (all ps < .05). Whereas the control groups did not differ from each other in accuracy or RT for the object task, the YCS group was significantly more accurate (p < .05) and faster to respond (p < .001) than the OCS group in the fractal task.

Length of the SPR list significantly affected RT for both the object, F(3, 861) = 3.359, p < .05, and fractal, F(3, 651) = 2.729, p < .05, tasks; however, the effects were in opposite directions. All participants were slower to the two-item list compared with the four- and six-item lists in the object task (p < .05 each) and faster to the two-item list compared with the four-item list in the fractal task (p < .05). List length significantly affected accuracy in the fractal task only, F(3, 948) = 6.418, p < .001; all groups were more accurate on the less demanding two-item list compared with the four-, six-, and eight-item lists (p < .05 for all). There were no interactions between Group and List Length for either accuracy or RT for either task.

DBS Effects on Behavior

As expected, DBS improved motor function in the PD participants. There was DBS-related improvement in clinical ratings of motor disability with lower UPDRS scores (reduced motor disability) for all PD participants (nonparametric sign test, p < .01). DBS also improved motor RTs on the most difficult eight-item list length of the fractal task, F(3, 119) = 2.879, p < .05. There were no significant effects of DBS on accuracy in either working memory task. However, the statistically significant differences between the PD group without stimulation and their age-matched controls on measures of accuracy were no longer present with stimulation (Figure 1, bottom). The PD group remained significantly less accurate and slower than the YCS group on both tasks even after stimulation (Figure 2).

Figure 2. 

Behavioral results from the object (A) and fractal (B) SPR tasks. RT (top) and accuracy for each of the four different list lengths (2, 4, 6 and 8) are plotted. Error bars represent SEMs. Note the overall poorer accuracy of the PD OFF group relative to the controls on both tasks improves with stimulation (PD ON). Also note that RTs for the PD group did not significantly improve with stimulation.

Figure 2. 

Behavioral results from the object (A) and fractal (B) SPR tasks. RT (top) and accuracy for each of the four different list lengths (2, 4, 6 and 8) are plotted. Error bars represent SEMs. Note the overall poorer accuracy of the PD OFF group relative to the controls on both tasks improves with stimulation (PD ON). Also note that RTs for the PD group did not significantly improve with stimulation.

Neurophysiological Measures

In both SPR tasks, a strong band of low frequency power, primarily theta (4–8 Hz) appeared, and there was relative suppression of higher frequency power (>8 Hz) immediately after the probe onset. This pattern of activity was sustained for approximately 1 sec in the YCS group and 600 msec in the OCS group. The PD OFF group showed a similar pattern of low-frequency activation and higher-frequency suppression initially, but there was less overall power in the theta range and it was of much shorter duration, diminishing by approximately 400 msec. There was also a small band of low beta power (12–20 Hz) present immediately after the probe appeared in the PD OFF group that was not present in the control groups (Figure 3). ERP waveforms of both SPR tasks showed peak amplitude of the N200 within a time window of 200–350 msec after the appearance of the probe for all groups confirming this time window was appropriate for analyses (Figure 4).

Figure 3. 

Group-averaged time–frequency plots during the object and fractal SPR tasks at channel Fz. All plots have a frequency scale (y axis) from 3 to 25 Hz, a color bar scale from −2.5 to 2.5 μV, and a timescale from −200 to 1200 msec. The dotted red line represents the onset of the probe stimulus for all list lengths (2, 4, 6, and 8 items). Note the relatively greater amplitude of power in the theta (4–8 Hz) band with relative suppression of alpha activity (8–12 Hz) in the two control groups and increased beta activity (12–25 Hz) in the PD groups.

Figure 3. 

Group-averaged time–frequency plots during the object and fractal SPR tasks at channel Fz. All plots have a frequency scale (y axis) from 3 to 25 Hz, a color bar scale from −2.5 to 2.5 μV, and a timescale from −200 to 1200 msec. The dotted red line represents the onset of the probe stimulus for all list lengths (2, 4, 6, and 8 items). Note the relatively greater amplitude of power in the theta (4–8 Hz) band with relative suppression of alpha activity (8–12 Hz) in the two control groups and increased beta activity (12–25 Hz) in the PD groups.

Figure 4. 

ERP waveforms averaged across all working memory loads (2–8) and all five channels (Fz, FCz, Cz, CPz, Pz) for all four groups on the object (A) and fractal (B) tasks. The x axis represents the time from −200 msec before probe onset (0 msec) to 1000 msec. The y axis represents voltage in microvolts. The gray box from 200 to 350 msec represents the N200 time frame used in the analyses. To the right of each averaged ERP plot are topographic head plots showing the mean N200 amplitude for each group during the N200 time window. The head plot color bar scale is −7.0 to 7.0 μV. Note the larger negative deflection in the N200 time frame for the PD group compared with the controls. Also note the head plots showing greater posterior power during the time window for the control groups and greater frontal suppression in the PD group.

Figure 4. 

ERP waveforms averaged across all working memory loads (2–8) and all five channels (Fz, FCz, Cz, CPz, Pz) for all four groups on the object (A) and fractal (B) tasks. The x axis represents the time from −200 msec before probe onset (0 msec) to 1000 msec. The y axis represents voltage in microvolts. The gray box from 200 to 350 msec represents the N200 time frame used in the analyses. To the right of each averaged ERP plot are topographic head plots showing the mean N200 amplitude for each group during the N200 time window. The head plot color bar scale is −7.0 to 7.0 μV. Note the larger negative deflection in the N200 time frame for the PD group compared with the controls. Also note the head plots showing greater posterior power during the time window for the control groups and greater frontal suppression in the PD group.

In the object task, the YCS, OCS, and PD OFF groups differed significantly on N200 peak latency, F(2, 147) = 7.845, p < .001, N200 peak amplitude, F(2, 147) = 6.034, p < .05, and N200 mean amplitude, F(2, 147) = 29.872, p < .001 (Figure 4A). In the fractal task, the groups differed on N200 peak amplitude, F(2, 146) = 8.030, p < .001, and N200 mean amplitude, F(2, 146) = 19.261, p < .001, but not peak latency (Figure 4B). In the object task, the N200 response of the YCS group occurred significantly sooner after the probe onset than both the OCS (p < .001) and PD OFF groups (p < .05), but the OCS and PD OFF groups did not differ significantly in N200 response latency. In the fractal task, the PD OFF participants had a faster N200 peak latency than the OCS group, but the three groups did not differ significantly on this measure. On both SPR tasks, the PD OFF peak N200 amplitude was significantly greater than both control groups (all ps < .05). In addition, the PD OFF group had a larger mean N200 amplitude than both control groups on the object (both p < .001) and fractal (p < .001 each) tasks. The control groups did not differ in the size of the peak or mean N200 amplitude.

DBS Effects on Neurophysiology

In general, DBS led to neurophysiological activity that was more similar to that seen in the age-matched OCS group. In the N200 time window, there was qualitatively less overall suppression of activity in frontal regions and greater activation in posterior regions (Figure 4, right). In addition, DBS led to significant reductions in the PD group's N200 peak amplitude for both the object, F(1, 49) = 16.239, p < .001, and fractal, F(1, 49) = 59.259, p < .001, tasks and reduced N200 mean amplitude in the object, F(1, 49) = 15.071, p < .001, and fractal, F(1, 49) = 26.398, p < .001, tasks (Figure 4). Furthermore, DBS-related reduction in N200 peak amplitude made these responses equivalent to both control groups in the fractal task. With regard to the latency of responses, DBS increased N200 latency on the fractal task by an average of 25 msec, F(1, 49) = 14.692, p < .001. DBS-related reductions in N200 peak latency in the object task made these responses faster than the OCS group and increased N200 peak latency in the fractal task made these responses equivalent to the OCS group (Figure 4).

Relationships between Neurophysiology and Behavior

Collapsed across all groups, response latency was significantly related to the mean N200 amplitude, F(1, 38) = 16.19, p < .001, R2 = .28, in the object task alone. There were no significant relationships between accuracy and any neurophysiological measures for either task. Linear regressions performed at the group level were more informative. For the OCS group mean N200 amplitude was related to accuracy in the object task, F(1, 38) = 6.21, p < .05, R2 = .37. In the PD OFF group, response speed was significantly related to mean N200 amplitude, F(1, 38) = 6.55, p < .05, R2 = .38, in the object task. In the more difficult fractal task, accuracy in the PD OFF group was significantly related to N200 peak latency, F(1, 38) = 64.30, p < .001, R2 = .88, such that faster latencies were related to poorer accuracy (Figure 5A). Also for the PD OFF group, RT was related to both peak and mean N200 amplitude, F(1, 38) = 9.75, p < .01, R2 = .66, with larger peak and mean amplitudes related to slower RTs (Figure 5B).

Figure 5. 

Results of the linear regressions for the PD OFF group on the more difficult fractal SPR task on accuracy (A) and RT (B). Linear regression model for PD OFF accuracy selected the N200 latency as the sole independent variable (Beta = .943, t = 8.02, p < .001) with an adjusted R2 of .876. Linear regression revealed two neurophysiological variables, N200 peak and mean amplitudes combining to predict RT with an adjusted R2 of .660.

Figure 5. 

Results of the linear regressions for the PD OFF group on the more difficult fractal SPR task on accuracy (A) and RT (B). Linear regression model for PD OFF accuracy selected the N200 latency as the sole independent variable (Beta = .943, t = 8.02, p < .001) with an adjusted R2 of .876. Linear regression revealed two neurophysiological variables, N200 peak and mean amplitudes combining to predict RT with an adjusted R2 of .660.

DISCUSSION

In this study, we examined the neurophysiological effects of STN DBS on motor disability and working memory. Because DBS is used to treat the severe motor disability of PD, we first sought to determine the effects of DBS on motor function. DBS significantly improved motor disability and also improved response speed on the most difficult condition of a working memory task. We then sought to determine the effects of DBS on working memory, a nonmotor behavior which also relies on frontal cortex. Here, DBS did not significantly alter working memory performance. From the neurophysiological data collected during the working memory task, we found that without stimulation, PD participants with worse accuracy had faster N200 responses (peak latencies), and slower RTs were related to greater N200 peak and mean amplitudes. Although DBS did not have a significant effect on working memory performance, it did significantly alter the N200 response; significantly slowing N200 latencies during both working memory tasks and reducing N200 peak and mean amplitudes in the fractal task. Our data suggest that DBS normalizes task-related neurophysiological function with dissociable effects on motor function and working memory.

The main finding from this study is that DBS altered PD task-related N200 latencies and amplitudes to be more like those of age-matched normal controls. Normalizing effects of DBS on cortical EEG have been reported previously, however, earlier studies have generally examined motor related behavior and none have looked specifically at the neural effects of stimulation on working memory. In a previous EEG study, Gerschlager and colleagues (1999) reported beneficial effects of STN DBS on the contingent negative variation, an early cognitive process associated with planning a motor response. They found increased contingent negative variation amplitudes with DBS over frontal and fronto-central regions. However, a more recent study by the same group (Gerschlager et al., 2001) found no evidence that DBS improved pathologically slowed P300 latencies in PD subjects during an auditory oddball paradigm. Another study showed no effects of STN DBS on the N2/PC ERP response during a stop suppression task (Swann et al., 2011).

In our study, we found that the PD OFF N200 peak latencies were significantly related to accuracy in the more difficult fractal SPR task with longer latencies associated with better accuracy. We found that DBS increased N200 latencies, which might suggest a corresponding improvement in working memory performance that we did not find. Overall slower ERP latencies are commonly reported in PD and other neurological and psychiatric disease states, likely reflecting poorer neural efficiency, and are generally associated with diminished behavioral performance (e.g., Eusebio et al., 2009; Caviness et al., 2007). We found that accuracy was numerically better, but not statistically better in subjects with DBS-induced slower N200 responses, which suggests that these PD participants were able to make more thoughtful and accurate decisions about the contents of working memory. This notion is supported by results from a recent study demonstrating that STN DBS improves the fidelity of executive processes related to stopping a planned motor action in a Stop Signal Suppression Task (Swann et al., 2011). In this study, the authors also report that oscillatory activity in the beta band (16–20 Hz) over the right pFC increased during stimulation. The disruption of PD pathological beta synchrony by STN DBS (Brown & Eusebio, 2008; Brown, 2003) may allow more flexible deployment of task-specific oscillatory activity. This normalization of neural activity may be the mechanism by which STN DBS allows for the recruitment of fronto-striatal networks for specific tasks.

Recent imaging studies would seem to support this mechanistic explanation for DBS effects on pFC. One study found decreases in rCBF in motor regions of frontal cortex following STN DBS (Ceballos-Baumann et al., 1999) and others have reported reductions in FDG PET activity in pFC and other regions of heteromodal cortex (Hershey et al., 2003). Reduction in pathological hypersynchrony by DBS could account for the reduced metabolic activity seen in functional imaging studies and the observation of normalized electrophysiological activity we observed with cortical EEG in the current study may link these processes.

In this study, we observed effects of DBS on motor function with faster RTs in the most difficult 8-item condition of the fractal task and reduced motor disability as measured by the UPDRS. These findings are not unexpected. For the majority of carefully selected PD patients, STN DBS greatly improves motor function and reduces disability (Moro et al., 2010; Weaver et al., 2009; Deuschl et al., 2006). The underlying mechanisms of DBS on motor function are not entirely clear but are similar to those seen in dopamine replacement. Indeed, DBS can extend therapeutic efficacy of dopamine replacement, reducing the amount of medication required for motor benefit and in a minority of patients, dopamine replacement can be eliminated altogether (Rodriguez-Oroz et al., 2005). In our nonmedicated subjects, the observed motor effects are attributable to direct effects of DBS. The beneficial motor effects of STN DBS are greatest when the stimulation field is restricted to the dorsal sensorimotor region of the STN, which sends efferent signals to primary and supplementary motor cortex. The encroachment of the stimulation field on the more ventrally located cognitive subregion of the STN may underlie the frequently reported cognitive effects of STN DBS.

Despite the clear effects of DBS on cortical neurophysiology and motor function, we found little evidence to suggest that DBS meaningfully alters working memory performance. Reports of DBS effects on executive functions including working memory are inconsistent. Most studies report a negative impact on working memory (see Tröster, McTaggart, & Heber, 2008, for a review) and fewer report beneficial effects (Mollion et al., 2011; Jahanshahi et al., 2000; Pillon et al., 2000). Our results are somewhat similar to those reported by Mollion and colleagues (2011), in which PD participants without stimulation were found to be significantly impaired on a visual working memory task compared with controls. Their results suggest that, although DBS significantly improved working memory performance relative to the OFF state, their subjects remained significantly impaired relative to the control group. We also find that PD participants OFF stimulation differed significantly from controls, but in the ON comparison with controls, they no longer differed. However, we did not find a significant main effect of DBS on working memory accuracy within the PD group; thus, we should not conclude that the elimination of this significant difference was due to DBS.

There are a few possible reasons why we did not observe an effect of DBS on working memory performance (either improvement or worsening), whereas other groups have. One reason may be related to our working memory task and the use of rehearsal. For example, the spatial delayed response task used by Hershey et al. (2004, 2008) involves remembering a spatial location on a computer screen over a filled delay period, during which the subject must complete a continuous performance task intended to prevent rehearsal. In the SPR task used in the current study, no such attempts were made to prevent rehearsal. We attempted to minimize the use of verbal strategies for encoding by using the fractal designs, but it is possible that subjects still used verbal approaches to encode and rehearse during the task (cf. Postle, D'Esposito, & Corkin, 2005). Alternately, a lack of observable effect of STN DBS on working memory in our study may be explained in part by alterations of other cognitive functions such as complex attention, selective attention, processing speed, conceptualization, and response inhibition; all of which have been reported to be improved by STN DBS (e.g., Swann et al., 2011; Frank, Samanta, Moustafa, & Sherman, 2007; Alegret et al., 2001; Jahanshahi et al., 2000; Saint-Cyr et al., 2000). We find the lack of an effect of DBS on working memory interesting given the clear effect on motor function and the alteration of underlying neurophysiology. These dissociable effects underscore the complexity of these mixed parallel and integrative striato-frontal networks (Haber et al., 2006).

In this study, we attempted to address several possible confounds often not accounted for in previous studies to increase the specificity of our results. All PD participants were tested after overnight PD medication withdrawal, so we would be more confident that any observed effects were due to DBS alone. We also attempted to minimize practice or familiarity effects that can result from administering the same tests in a short span of time commonly used in DBS ON versus OFF studies by randomizing the order of stimulator condition (ON/OFF) and using parallel versions of the SPR task. However, a major limitation to our study and interpretation of the results is the fact that we do not know the exact location of the electrodes and stimulation fields in our patients. This would appear to be a critical piece of information, because the STN is a neuroanatomically dense hub of information processing and between-subject variation in location of active stimulation could significantly impact outcomes. Valid and reliable methods for determining lead placement and stimulation fields postoperatively are paramount for determining STN DBS effects at the individual level and for truly understanding the role of the STN in fronto-striatal circuitry and cognition. We anticipate that future work will attempt to more rigorously link the location of stimulation field in the STN with nonmotor outcomes.

Reprint requests should be sent to Mark Mapstone, Department of Neurology, University of Rochester, 601 Elmwood Avenue, Box 673, Rochester, NY 14642, or via e-mail: mark_mapstone@urmc.rochester.edu.

REFERENCES

REFERENCES
Alegret
,
M.
,
Junqué
,
C.
,
Valldeoriola
,
F.
,
Vendrell
,
P.
,
Pilleri
,
M.
,
Rumià
,
J.
,
et al
(
2001
).
Effects of bilateral subthalamic stimulation on cognitive function in Parkinson disease.
Archives of Neurology
,
58
,
1223
1227
.
Alexander
,
G. E.
,
DeLong
,
M. R.
, &
Strick
,
P. L.
(
1986
).
Parallel organization of functionally segregated circuits linking basal ganglia and cortex.
Annual Review of Neuroscience
,
9
,
357
381
.
Ardouin
,
C.
,
Pillon
,
B.
,
Peiffer
,
E.
,
Bejjani
,
P.
,
Limousin
,
P.
,
Damier
,
P.
,
et al
(
1999
).
Bilateral subthalamic or pallidal stimulation for Parkinson's disease affects neither memory nor executive functions: A consecutive series of 62 patients.
Annals of Neurology
,
46
,
217
223
.
Aron
,
A. R.
,
Durston
,
S.
,
Eagle
,
D. M.
,
Logan
,
G. D.
,
Stinear
,
C. M.
, &
Stuphorn
,
V.
(
2007
).
Converging evidence for a fronto-basal-ganglia network for inhibitory control of action and cognition.
Journal of Neuroscience
,
27
,
11860
11864
.
Baddeley
,
A. D.
(
1986
).
Working memory.
New York
:
Oxford University Press
.
Baddeley
,
A. D.
(
2000
).
The episodic buffer: A new component of working memory?
Trends in Cognitive Science
,
4
,
417
423
.
Ballanger
,
B.
,
Jahanshahi
,
M.
,
Broussolle
,
E.
, &
Thobois
,
S.
(
2009
).
PET functional imaging of deep brain stimulation in movement disorders and psychiatry.
Journal of Cerebral Blood Flow & Metabolism
,
29
,
1743
1754
.
Brown
,
P.
(
2003
).
Oscillatory nature of human basal ganglia activity: Relationship to the pathophysiology of Parkinson's disease.
Movement Disorders
,
18
,
357
363
.
Brown
,
P.
, &
Eusebio
,
A.
(
2008
).
Paradoxes of functional neurosurgery: Clues from basal ganglia recordings.
Movement Disorders
,
23
,
12
20
.
Caviness
,
J. N.
,
Driver-Dunckley
,
E.
,
Connor
,
D. J.
,
Sabbagh
,
M. N.
,
Hentz
,
J. G.
,
Noble
,
B.
,
et al
(
2007
).
Defining mild cognitive impairment in Parkinson's disease.
Movement Disorders
,
22
,
1272
1277
.
Ceballos-Baumann
,
A. O.
,
Boecker
,
H.
,
Bartenstein
,
P.
,
von Falkenhayn
,
I.
,
Riescher
,
H.
,
Conrad
,
B.
,
et al
(
1999
).
A positron emission tomographic study of subthalamic nucleus stimulation in Parkinson disease: Enhanced movement-related activity of motor-association cortex and decreased motor cortex resting activity.
Archives of Neurology
,
56
,
997
1003
.
Cools
,
R.
,
Barker
,
R. A.
,
Sahakian
,
B. J.
, &
Robbins
,
T. W.
(
2001
).
Enhanced or impaired cognitive function in Parkinson's disease as a function of dopaminergic medication and task demands.
Cerebral Cortex
,
11
,
1136
1143
.
Daffner
,
K. R.
,
Mesulam
,
M. M.
,
Scinto
,
L. F.
,
Calvo
,
V.
,
Faust
,
R.
, &
Holcomb
,
P. J.
(
2000
).
An electrophysiological index of stimulus unfamiliarity.
Psychophysiology
,
37
,
737
747
.
Delorme
,
A.
, &
Makeig
,
S.
(
2004
).
EEGLAB: An open source toolbox for analysis of single-trial EEG dynamics including independent component analysis.
Journal of Neuroscience Methods
,
134
,
9
21
.
Deuschl
,
G.
,
Schade-Brittinger
,
C.
,
Krack
,
P.
,
Volkmann
,
J.
,
Schafer
,
H.
,
Botzel
,
K.
,
et al
(
2006
).
A randomized trial of deep-brain stimulation for Parkinson's disease.
New England Journal of Medicine
,
355
,
896
908
.
Eusebio
,
A.
,
Pogosyan
,
A.
,
Wang
,
S.
,
Averbeck
,
B.
,
Gaynor
,
L. D.
,
Cantiniaux
,
S.
,
et al
(
2009
).
Resonance in subthalamo-cortical circuits in Parkinson's disease.
Brain
,
132
,
2139
2150
.
Fahn
,
S.
,
Elton
,
R. L.
, &
UPDRS Program Members
. (
1987
).
Unified Parkinson's Disease Rating Scale.
In S. Fahn, C. D. Marsden, M. Goldstein, & D. B. Calne (Eds.)
,
Recent developments in Parkinson's disease
(
Vol. 2
, pp.
153
163
).
Florham Park, NJ
:
Macmillan Healthcare Information
.
Frank
,
M. J.
,
Samanta
,
J.
,
Moustafa
,
A. A.
, &
Sherman
,
S. J.
(
2007
).
Hold your horses: Impulsivity, deep brain stimulation, and medication in parkinsonism.
Science
,
318
,
1309
1312
.
Gerschlager
,
W.
,
Alesch
,
F.
,
Cunnington
,
R.
,
Deecke
,
L.
,
Dirnberger
,
G.
,
Endl
,
W.
,
et al
(
1999
).
Bilateral subthalamic nucleus stimulation improves frontal cortex function in Parkinson's disease. An electrophysiological study of the contingent negative variation.
Brain
,
122
,
2365
2373
.
Gerschlager
,
W.
,
Bloem
,
B. R.
,
Alesch
,
F.
,
Lang
,
W.
,
Deecke
,
L.
, &
Cunnington
,
R.
(
2001
).
Bilateral subthalamic nucleus stimulation does not improve prolonged P300 latencies in Parkinson's disease.
Journal of Neurology
,
248
,
285
289
.
Girotti
,
F.
,
Carella
,
F.
,
Grassi
,
M. P.
,
Soliveri
,
P.
,
Marano
,
R.
, &
Caraceni
,
T.
(
1986
).
Motor and cognitive performances of parkinsonian patients in the on and off phases of the disease.
Journal of Neurology, Neurosurgery, and Psychiatry
,
49
,
657
660
.
Gotham
,
A. M.
,
Brown
,
R. G.
, &
Marsden
,
C. D.
(
1988
).
“Frontal” cognitive function in patients with Parkinson's disease “on” and “off” levodopa.
Brain
,
111
,
299
321
.
Green
,
J.
,
Woodard
,
J. L.
,
Sirockman
,
B. E.
,
Zakers
,
G.
,
Maier
,
C. L.
,
Green
,
R. C.
,
et al
(
1996
).
Event-related potential P3 change in mild Parkinson's disease.
Movement Disorders
,
11
,
32
42
.
Haber
,
S.
,
Fudge
,
J.
, &
McFarland
,
N.
(
2000
).
Striato-nigrostriatal pathways in primates form an ascending spiral from the shell to the dorsolateral striatum.
The Journal of Neuroscience
,
20
,
2369
2382
.
Haber
,
S.
,
Kim
,
K.
,
Mailly
,
P.
, &
Calzavara
,
R.
(
2006
).
Reward-related cortical inputs define a large striatal region in primates that interface with associative cortical connections providing a substrate for incentive-based learning.
The Journal of Neuroscience
,
26
,
8368
8376
.
Heo
,
J.-H.
,
Lee
,
K.-M.
,
Paek
,
S. H.
,
Kim
,
M.-J.
,
Lee
,
J.-Y.
,
Kim
,
J.-Y.
,
et al
(
2008
).
The effects of bilateral subthalamic nucleus deep brain stimulation (STN DBS) on cognition in Parkinson disease.
Journal of Neurological Science
,
273
,
19
24
.
Hershey
,
T.
,
Revilla
,
F. J.
,
Wernle
,
A.
,
McGee-Minnich
,
L.
,
Antenor
,
J. V.
,
Videen
,
T. O.
,
et al
(
2003
).
Cortical and subcortical blood flow effects of subthalamic nucleus stimulation in PD.
Neurology
,
61
,
816
821
.
Hershey
,
T.
,
Revilla
,
F.
,
Wernle
,
A.
,
Schneider-Gibson
,
P.
,
Dowling
,
J.
, &
Perlmutter
,
J. S.
(
2004
).
Stimulation of STN impairs aspects of cognitive control in PD.
Neurology
,
62
,
1110
1114
.
Hershey
,
T.
,
Wu
,
J.
,
Weaver
,
P. M.
,
Perantie
,
D. C.
,
Karimi
,
M.
,
Tabbal
,
S. D.
,
et al
(
2008
).
Unilateral vs. bilateral STN DBS effects on working memory and motor function in Parkinson disease.
Experimental Neurology
,
210
,
402
408
.
Jahanshahi
,
M.
,
Ardouin
,
C. M. A.
,
Brown
,
R. G.
,
Rothwell
,
J. C.
,
Obeso
,
J.
,
Albanese
,
A.
,
et al
(
2000
).
The impact of deep brain stimulation on executive function in Parkinson's disease.
Brain
,
123
,
1142
1154
.
Jeck
,
R.
,
Rùzicka
,
E.
,
Urgosík
,
D.
,
Serranová
,
T.
,
Volfová
,
M.
,
Nováková
,
O.
,
et al
(
2006
).
Deep brain stimulation of the subthalamic nucleus affects resting EEG and visual evoked potential in Parkinson's disease.
Clinical Neurophysiology
,
117
,
1017
1028
.
Jung
,
T. P.
,
Makeig
,
S.
,
Humphries
,
C.
,
Lee
,
T. W.
,
McKeown
,
M. J.
,
Iragui
,
V.
,
et al
(
2000
).
Removing electroencephalographic artifacts by blind sources separation.
Psychophysiology
,
37
,
163
178
.
Kusak
,
G.
,
Grune
,
K.
,
Hagendorf
,
H.
, &
Metz
,
A. M.
(
2000
).
Updating of working memory in a running memory task: An event-related potential study.
International Journal of Psychophysiology
,
39
,
51
65
.
Lange
,
K. W.
,
Robbins
,
T. W.
,
Marsden
,
C. D.
,
James
,
M.
,
Owen
,
A. M.
, &
Paul
,
G. M.
(
1992
).
L-dopa withdrawal in Parkinson's disease selectively impairs cognitive performance in tests sensitive to frontal lobe dysfunction.
Psychopharmacology
,
107
,
394
404
.
Limousin
,
P.
,
Greene
,
J.
,
Pollak
,
P.
,
Rothwell
,
J.
,
Benabid
,
A. L.
, &
Frackowiak
,
R.
(
1997
).
Changes in cerebral activity pattern due to subthalamic nucleus or internal pallidum stimulation in Parkinson's disease.
Annals of Neurology
,
42
,
283
291
.
McNab
,
F.
, &
Klingberg
,
T.
(
2008
).
Prefrontal cortex and basal ganglia control access to working memory.
Nature Neuroscience
,
11
,
103
107
.
Mesulam
,
M.
(
2000
).
Brain, mind, and the evolution of connectivity.
Brain and Cognition
,
42
,
4
6
.
Mink
,
J. W.
(
1996
).
The basal ganglia: Focused selection and inhibition of competing motor programs.
Progress in Neurobiology
,
50
,
381
425
.
Missonnier
,
P.
,
Leonards
,
U.
,
Gold
,
G.
,
Palix
,
J.
,
Ibanez
,
V.
, &
Giannakopoulos
,
P.
(
2003
).
A new electrophysiological index for working memory load in humans.
NeuroReport
,
14
,
1451
1455
.
Miyashita
,
Y.
,
Higuchi
,
S.
,
Sakai
,
K.
, &
Masui
,
N.
(
1991
).
Generation of fractal patterns for probing the visual memory.
Neuroscience Research
,
12
,
307
311
.
Mollion
,
H.
,
Dominey
,
P. F.
,
Broussolle
,
E.
, &
Ventre-Dominey
,
J.
(
2011
).
Subthalamic nucleus stimulation selectively improves motor and visual memory performance in Parkinson's disease.
Movement Disorders
,
26
,
2019
2025
.
Moro
,
E.
,
Lozano
,
A. M.
,
Pollak
,
P.
,
Agid
,
Y.
,
Rehncrona
,
S.
,
Volkmann
,
J.
,
et al
(
2010
).
Long-term results of a multicenter study on subthalamic and pallidal stimulation in Parkinson's disease.
Movement Disorders
,
25
,
578
586
.
Nobre
,
A. C.
,
Griffin
,
I. C.
, &
Rao
,
A.
(
2007
).
Spatial attention can bias search in visual short-term memory.
Frontiers in Human Neuroscience
,
1
,
4
.
Owen
,
A. M.
,
Iddon
,
J. L.
,
Hodges
,
J. R.
,
Summers
,
B. A.
, &
Robbins
,
T. W.
(
1997
).
Spatial and non-spatial working memory at different stages of Parkinson's disease.
Neuropsychologia
,
35
,
519
532
.
Palva
,
S.
,
Kulashekhar
,
S.
,
Hamalainen
,
M.
, &
Palva
,
J. M.
(
2011
).
Localization of cortical phase and amplitude dynamics during visual working memory encoding and retention.
Journal of Neuroscience
,
31
,
5013
5025
.
Pillon
,
B.
,
Ardouin
,
C.
,
Damier
,
P.
,
Krack
,
P.
,
Houeto
,
J. L.
,
Klinger
,
H.
,
et al
(
2000
).
Neuropsychological changes between “off” and “on” STN or GPi stimulation in Parkinson's disease.
Neurology
,
55
,
411
418
.
Postle
,
B. R.
(
2006
).
Working memory as an emergent property of the mind and brain.
Neuroscience
,
139
,
23
38
.
Postle
,
B. R.
,
D'Esposito
,
M.
, &
Corkin
,
S.
(
2005
).
Effects of verbal and nonverbal interference on spatial and object visual working memory.
Memory & Cognition
,
33
,
203
212
.
Ragland
,
J. D.
,
Turetsky
,
B. I.
,
Gur
,
R. C.
,
Gunning-Dixon
,
F.
,
Turner
,
T.
,
Schroeder
,
L.
,
et al
(
2002
).
Working memory for complex figures: An fMRI comparison of letter and fractal n-back tasks.
Neuropsychology
,
16
,
370
379
.
Rodriguez-Oroz
,
M. C.
,
Obeso
,
J. A.
,
Lang
,
A. E.
,
Houeto
,
J. L.
,
Pollak
,
P.
,
Rehncrona
,
S.
,
et al
(
2005
).
Bilateral deep brain stimulation in Parkinson's disease: A multicentre study with 4 years follow-up.
Brain
,
128
,
2240
2249
.
Rowe
,
J. B.
,
Toni
,
I.
,
Josephs
,
O.
,
Frackowiak
,
R. S.
, &
Passingham
,
R. E.
(
2000
).
The prefrontal cortex: Response selection or maintenance within working memory?
Science
,
288
,
1656
1660
.
Ruchkin
,
D.
,
Johnson
,
R.
,
Grafman
,
J.
,
Canoune
,
H.
, &
Ritter
,
W.
(
1997
).
Multiple visuospatial working memory buffers: Evidence from spatiotemporal patterns of brain activity.
Neuropsychologia
,
35
,
195
209
.
Sagliocco
,
L.
,
Meistrowitz
,
A.
,
Schwendemann
,
G.
,
Herrmann
,
M.
, &
Basar-Eroglu
,
C.
(
2005
).
Theta and alpha oscillations reflect differences in memory strategy and visual discrimination in patients with Parkinson's disease.
Neuroscience Letters
,
388
,
138
143
.
Saint-Cyr
,
J. A.
,
Trépanier
,
L. L.
,
Kumar
,
R.
,
Lozano
,
A. M.
, &
Lang
,
A. E.
(
2000
).
Neuropsychological consequences of chronic bilateral stimulation of the subthalamic nucleus in Parkinson's disease.
Brain
,
123
,
2091
2108
.
Schroll
,
H.
,
Vitay
,
J.
, &
Hamker
,
F. H.
(
2012
).
Working memory and response selection: A computational account of interactions among cortico-basalganglio-thalamic loops.
Neural Networks
,
26
,
59
74
.
Sestini
,
S.
,
Scotto di Luzio
,
A.
,
Ammannati
,
F.
,
De Cristofaro
,
M. T.
,
Passeri
,
A.
,
Martini
,
S.
,
et al
(
2002
).
Changes in regional cerebral blood flow caused by deep-brain stimulation of the subthalamic nucleus in Parkinson's disease.
Journal of Nuclear Medicine
,
43
,
725
732
.
Sternberg
,
S.
(
1966
).
High-speed scanning in human memory.
Science
,
153
,
652
654
.
Swann
,
N.
,
Poizner
,
H.
,
Houser
,
M.
,
Gould
,
S.
,
Greenhouse
,
I.
,
Cai
,
W.
,
et al
(
2011
).
Deep brain stimulation of the subthalamic nucleus alters the cortical profile of response inhibition in the beta frequency band: A scalp EEG study in Parkinson's disease.
Journal of Neuroscience
,
31
,
5721
5729
.
Tröster
,
A. I.
,
McTaggart
,
A. B.
, &
Heber
,
I. A.
(
2008
).
Neuropsychological issues in deep brain stimulation of neurological and psychiatric disorders.
In D. Tarsey, J. L. Vitek, P. A. Starr, & M. S. Okun (Eds.)
,
Deep brain stimulation in neurological and psychiatric disorders
(pp.
399
452
).
New York
:
Humana Press
.
Vogel
,
E. K.
, &
Machizawa
,
M. G.
(
2004
).
Neural activity predicts individual differences in visual working memory capacity.
Nature
,
428
,
748
751
.
Weaver
,
F. M.
,
Follett
,
K.
,
Stern
,
M.
,
Hur
,
K.
,
Harris
,
C.
, &
Marks
,
W. J.
(
2009
).
Bilateral deep brain stimulation vs best medical therapy for patients with advanced Parkinson disease: A randomized controlled trial.
Journal of the American Medical Association
,
301
,
63
73
.