Cnv Topography And Generators

Parkinson Disease Medications

Get Instant Access

18.3.1 CNV Topography

The primary source of evoked potentials is neuronal activity in the upper layers of the cortex. It is primarily the simultaneous occurrence of numerous excitatory postsynaptic potentials at apical dentrites of pyramidal neurons close to the recording electrode that causes changes in the evoked potentials at the surface of the head (Birbaumer et al., 1990). Most studies report that the CNV is mainly observed at fronto-central, central, and centro-parietal regions. However, it is sometimes difficult to disambiguate motor ERPs and cognitive ERPs in the CNV, whereas it has been suggested that cognitive preparation (reflected in portions of the CNV) is a psychological process that differs from motor preparation (Bereitschaftspotential (BP)). A study by Leynes et al. (1998) has provided separate records of brain activity during motor and cognitive preparations. The ERP topographies differed for the two types of tasks (cognitive vs. motor) used: a central CNV for the motor task and a frontal CNV for the cognitive task. These topographic effects suggest that the neural circuitries subserving the two preparatory processes differ. Hamano et al. (1997) pointed out that the CNV recorded from the scalp is the summation of several cortical potentials that have different origins and different functions. Besides, multiple cortical as well as subcortical regions have been suggested to participate in the generation of the CNV. Consequently, the exact generator or generators of the CNV in humans is still unclear.

18.3.2 Some Insights on CNV Generators

Using high-resolution spatiotemporal statistics and current source density, Cui et al. (2000) investigated the CNV topography. Their data suggest that the origin of the early CNV may rest in the frontal lobes. The authors argue that this is consistent with other previous results. For example, Lai et al. (1997) obtained the early CNV in the S1-S2 paradigm, which consisted of a frontal to fronto-polar midline negative potential, most likely associated with judgment and decision-making processes. Oishi and Mochizuki (1998) investigated the CNV paradigm with the rCBF method and found that there was a significant positive correlation between the amplitude of the early CNV and frontal blood flow. The generators of the CNV have been examined using subdural recordings in humans. They are said to include mesial frontal areas, the supplementary motor area, and the dorsolateral prefrontal cortex (Hamano et al., 1997; Ikeda et al., 1999; Lamarche et al., 1995). In sum, a growing body of evidence suggests that the frontal lobes appear to contain the main generators for the CNV. Nonetheless, surface recordings on the frontal part of the scalp may reflect the activity of a larger network. Results from studies recording the CNV of PD patients suggest that the cortical-basal ganglia-thalamo-cortical circuit plays a major role in the generation of the CNV. This possibility will be examined below.

18.3.3 Electrophysiological Study of the CNV in Parkinson's Disease

According to the model of basal ganglia-thalamo-cortical circuitry (Wichman and DeLong, 1996), a disease affecting basal ganglia function may lead to reduced outflow to the cortex. This view is supported by the analysis of the CNV recorded in patients with PD by Amabile et al. (1986). They found a correlation between CNV measures and pharmacological treatment after a washout period and 15 and 30 days after the start of treatment with L-DOPA. A small amount of CNV was observed in the nondrug condition, and an enhanced CNV was found during treatment with L-DOPA. This increase in the CNV following medication may reflect a functional recovery, perhaps partial, in the striato-thalamo-cortical connections that terminate in the upper frontal cortical layers, where they contact apical dentrites of pyramidal neurons (Amabile et al., 1986). Impaired activation of frontal cortical areas, including the supplementary motor area and prefrontal cortex, would result from impaired thalamo-cortical output of the basal ganglia.

Surgical interventions aimed at increasing basal ganglia-thalamic outflow to the cortex, such as electrical stimulation of the subthalamic nucleus with chronically implanted electrodes, have also been shown to improve cortical functioning, particularly within the frontal and premotor areas (Gerschlager et al., 1999). This is in accord with previous functional imaging studies showing increased activation of the supplementary motor area, cingulate cortex, and dorsolateral prefrontal cortex during effective stimulation of the subthalamic nucleus (Limousin et al., 1997).

Using EEG recordings, the results of Gerschlager et al. (1999) extend those of Limousin et al.'s (1997) positron emission tomography (PET) study. The latter study showed improved cortical activation only when the activity was averaged over a prolonged period during which the task was performed (i.e., a 90-sec acquisition period was used). The former study was able to show precisely when in time subthalamic nucleus stimulation improves the cortical activity, that is, during the preparatory period prior to the initiation of a response. Consequently, Gerschlager et al. (1999) assumed that subthalamic nucleus stimulation in PD improves the cortical activity that underlies cognitive processes associated with the preparation and organization of forthcoming responses. Combining EEG and PET data allows one to determine not only the areas involved in the processing of information, but also the time course of the activation of these areas.

18.3.4 Event-Related Potentials and Positron Emission Tomography Analysis of CNV Generators

In order to question the specificity of the spatiotemporal organization of cerebral areas subserving the processing of stimulus duration, we combined positron emission tomography and event-related potential data recorded from subjects performing two visual discrimination tasks, one based on the duration of a visual stimulus and the other on the intensity of the same stimulus. Results of the PET study showed that the same network was activated in both tasks — right prefrontal cortex, right inferior parietal lobule, anterior cingulate cortex, left fusiform gyrus, and vermis (Maquet et al., 1996). We could not unambiguously conclude, however, that this pattern of activation was specific to the perception of a stimulus duration (see Hinton, this volume).

As stressed above, the PET method integrates radioactive tracer activity over a period of many seconds; therefore, the temporal resolution of this method is insuf ficient to determine when the different areas are active and then to determine in which processing stages these areas are involved. ERPs, on the other hand, reflect the rapidly changing electrical activity in the brain evoked by a stimulus or a cognitive event, providing the means to differentiate the electrical activity pertaining to the different stages of information processing (see Sakata and Onoda, this volume). It is difficult, however, to determine the neural generators of ERP components, the solution of the inverse problem not being unique. Therefore, we first carried a dipole modeling using a PET-seeded model (right prefrontal, right parietal, anterior cingu-late, left and right fusiforms). Then, to obtain a better fit, two sources (cuneus and left prefrontal area) had to be added. This dipole modeling showed that the proportion of accounted variance was equivalent in both tasks (Pouthas et al., 2000). This indicates, consistent with the earlier PET findings, that duration and intensity dimensions of a visual stimulus are processed in the same cerebral areas.

However, ERPs revealed prominent differences between the time courses of the dipole activations for each task, particularly that of the probable generators of the late-latency ERP components. The magnitude waveforms of three dipoles — right frontal, anterior cingulate, and cuneus, observed when the intensity or the duration of a test stimulus equivalent to the memorized standard stimulus (i.e., 700 msec — 15 cd/m2) has to be judged — are superimposed on Figure 18.5. The timing of activation of these dipoles largely differed between the two tasks. It must be stressed that the stimulus was physically the same and that only the instructions received by the subjects were different; i.e., evaluate either intensity or duration. We will focus on the duration task and the putative generator of the CNV. Importantly, in the duration task, the right prefrontal dipole was the most active during the CNV (from 400 to 1000 msec), whereas in the intensity task, it showed a low level of activity during this time range. Moreover, in the duration task, although the CNV began earlier (at 300 msec), the magnitude waveform of this dipole appeared to be very similar to that of the CNV, resolving its activity when the CNV resolved, as shown in Figure 18.6. This suggests that in such a matching-to-sample task, the right frontal area has an essential role in making a decision about the stimulus duration. Because the dipole located in this area was not very active in the intensity task, only following LED switch off, we assume that this role is specific to the temporal dimension of the stimulus.

This assumption is in accord with other neuroimaging data (e.g., Hinton and Meck, 1997; Hinton et al., 1996; Meck et al., 1998). In addition, neuropsychological data of Harrington et al. (1998) on patients with focal left or right lesions revealed that only patients with right lesions show time perception deficits. In a study designed to investigate how brain activity is lateralized during the encoding and recognition of a visual stimulus duration, results showed that the right frontal cortex was involved in both operations, suggesting that the involvement of right frontal structures is critical for time perception (Monfort et al., 2000). Another example of a right hemispheric bias for processing temporal information can be found in Damen and Brunia's (1987) experiment. In a typical warning-imperative stimulus paradigm, the CNV reflects both motor preparation and stimulus anticipation. In order to separate temporally motor and cognitive preparations, the authors used a

FIGURE 18.5 Time course activation of the dipoles located in the right frontal cortex, anterior cingulate cortex, and cuneus in the 100- to 1500-msec window after stimulus onset for the standard stimulus in both duration (thick line) and intensity (thin line) tasks. (Adapted from Pouthas, V., Garnero, L., Ferrandez, A.M., and Renault, B., Hum. Brain Mapping, 10, 49-60, 2000.)

FIGURE 18.5 Time course activation of the dipoles located in the right frontal cortex, anterior cingulate cortex, and cuneus in the 100- to 1500-msec window after stimulus onset for the standard stimulus in both duration (thick line) and intensity (thin line) tasks. (Adapted from Pouthas, V., Garnero, L., Ferrandez, A.M., and Renault, B., Hum. Brain Mapping, 10, 49-60, 2000.)

Stimulus Preceding Negativity
FIGURE 18.6 Time course activation of the right frontal dipole in the 100- to 1500-msec window following stimulus onset for the five stimulus durations. (Adapted from Pouthas, V., Garnero, L., Ferrandez, A.M., and Renault, B., Hum. Brain Mapping, 10, 49-60, 2000.)

feedback paradigm. In this paradigm participants press a key when they believe that a target interval has elapsed, and a few seconds later they receive feedback on the accuracy of their temporal estimation. The potential that occurs prior to the feedback stimulus, named stimulus preceding negativity (SPN), is very similar to the nonmotor CNV observed in S1-S2 paradigms that do not require a motor response. It would reflect a similar cognitive preparatory process. The results of Damen and Brunia (1987) showed that the SPN was larger over the right hemisphere, irrespective of the movement side. This pointed to a right hemisphere preponderance of the SPN source.

Was this article helpful?

0 0

Post a comment