Transition of brain activation from frontal to parietal areas in visuomotor sequence learning (original) (raw)
Related papers
Presupplementary motor area activation during sequence learning reflects visuo-motor association
The Journal of neuroscience : the official journal of the Society for Neuroscience, 1999
In preceding studies (Hikosaka et al., 1996; Sakai et al., 1998) we have shown that the presupplementary motor area (pre-SMA), an anterior part of the medial premotor cortex, is active during visuo-motor sequence learning. However, the paradigm required the subjects first to acquire correct visuo-motor association and then to acquire correct sequence, and it was still unknown which of the two processes the pre-SMA is involved in. To further characterize the role of pre-SMA, we have conducted another series of functional magnetic resonance imaging experiments using three learning paradigms. The three were the same in that they involved a visuo-motor association component, but they differed in terms of the involvement of sequential components; one involved no sequence learning, whereas the other two involved learning of motor sequence or perceptual sequence. Comparison of the learning conditions with the any-order button press condition revealed pre-SMA activation in all three paradig...
The Time Course of Changes during Motor Sequence Learning: A Whole-Brain fMRI Study
NeuroImage, 1998
There is a discrepancy between the results of imaging studies in which subjects learn motor sequences. Some experiments have shown decreases in the activation of some areas as learning increased, whereas others have reported learning-related increases as learning progressed. We have exploited fMRI to measure changes in blood oxygen level-dependent (BOLD) signal throughout the course of learning. T2*-weighted echo-planar images were acquired over the whole brain for 40 min while the subjects learned a sequence eight moves long by trial and error. The movements were visually paced every 3.2 s and visual feedback was provided to the subjects. A baseline period followed each activation period. The effect due to the experimental conditions was modeled using a squarewave function, time locked to their occurrence. Changes over time in the difference between activation and baseline signal were modeled using a set of polynomial basis functions. This allowed us to take into account linear as well as nonlinear changes over time. Low-frequency changes over time common to both activation and baseline conditions (and thus not learning related) were modeled and removed. Linear and nonlinear changes of BOLD signal over time were found in prefrontal, premotor, and parietal cortex and in neostriatal and cerebellar areas. Single-unit recordings in nonhuman primates during the learning of motor tasks have clearly shown increased activity early in learning, followed by a decrease as learning progressed. Both phenomena can be observed at the population level in the present study. 1998 Academic Press
Prefrontal lesions impair the implicit and explicit learning of sequences on visuomotor tasks
Experimental Brain Research, 2002
Objective: (1) To verify whether the prefrontal cortex (PFC) is specifically involved in visuomotor sequence learning as opposed to other forms of motor learning and (2) to establish the role of executive functions in visuomotor sequence learning. Background: Visuomotor skill learning depends on the integrity of the premotor and parietal cortex; the prefrontal cortex, however, is essential when the learning of a sequence is required. Methods: We studied 25 patients with PFC lesions and 86 controls matched for age and educational level. Participants performed: (1) a Pursuit Tracking Task (PTT), composed of a random tracking task (perceptual learning) and a pattern tracking task (explicit motor sequence learning with learning indicated by the decrease in mean root square error across trial blocks), (2) a 12-item sequence version of a serial reaction time task (SRTT) with specific implicit motor sequence learning indicated by the rebound increase in response time when comparing the last sequence block with the next random block, and (3) a neuropsychological battery that assessed executive functions. Results: PFC patients were impaired in sequence learning on the pattern tracking task of the PTT and on the SRTT as compared to controls, but performed normally on the PTT random tracking task. Learning on the PTT did not correlate with learning on the SRTT. PTT performance correlated with planning functions while SRTT performance correlated with working memory capacity. Conclusions: The PFC is specifically involved in explicit and implicit motor sequence learning. Different PFC regions may be selectively involved in such learning depending on the cognitive demands of the sequential task.
Human brain activation accompanying explicitly directed movement sequence learning
Experimental Brain Research, 2001
We examined brain activation patterns occurring during the production and encoding of a motor sequence. Participants performed a variant of the serial reaction-time task under two conditions. The first condition was designed to foster the engagement of explicit mechanisms of knowledge acquisition. The second condition was intended to encourage the engagement of implicit learning mechanisms that would be more typical of the standard serial reaction-time task. In the first condition, the acquisition of explicit knowledge about an 8-element ordered sequence led to a significant and rapid decline in reaction time. By contrast, the second condition, the task in which a sequence was presented unbeknownst to participants, did not yield changes in reaction time. Several brain regions, including prefrontal cortex, superior and inferior parietal lobules, and cerebellum, exhibited explicit learning-related activation. The prefrontal cortex and inferior parietal lobules increased their levels of activation between the beginning and end of the experiment, while primary motor, primary sensory, and cerebellar cortex decreased their levels of activation from the beginning to the end of the experiment. We propose a model in which two processes, a learning-related increase and a habituation process might interact to produce the activation patterns observed during movement sequence acquisition. In short, the prefrontal cortex and inferior parietal lobule together direct and recruit superior parietal lobule and cerebellum to encode and perform the sequence. The increased activation in prefrontal cortex and inferior parietal lobule may represent the activity of a working memory circuit that functions in the acquisition and recall of sequence information.
Journal of neurophysiology, 1996
1. Using functional magnetic resonance imaging, we investigated the neural correlates of sequential procedural learning. During the test scans the subjects learned a new sequence (position or color) of button presses; during the control scans they pressed the buttons in any order. The comparison of the test and control scans was expected to reveal the neural activities related to learning, not sensory-motor processes. 2. We found that a localized area in what we regard to be the human homologue of the presupplementary motor area (pre-SMA) was particularly active for learning of new sequential procedures (either position or color sequences), not movements per se. 3. In contrast, the SMA proper (posterior to pre-SMA) was active for the performance of sequential movements, not learning. This was shown in another paradigm in which the subjects pressed the buttons in any order in the test scans and just watched the sequence in the control scans. 4. The learning-related pre-SMA region, wh...
A visuo-motor sequence can be learned as a series of visuo-spatial cues or as a sequence of effector movements. Earlier imaging studies have revealed that a network of brain areas is activated in the course of motor sequence learning. However, these studies do not address the question of the type of representation being established at various stages of visuo-motor sequence learning. In an earlier behavioral study, we demonstrated that acquisition of visuo-spatial sequence representation enables rapid learning in the early stage and progressive establishment of somato-motor representation helps speedier execution by the late stage.We conducted functional magnetic resonance imaging (fMRI) experiments wherein subjects learned and practiced the same sequence alternately in normal and rotated settings. In one rotated setting (visual), subjects learned a new motor sequence in response to an identical sequence of visual cues as in normal. In another rotated setting (motor), the display sequence was altered as compared to normal, but the same sequence of effector movements was used to perform the sequence. Comparison of different rotated settings revealed analogous transitions both in the cortical and subcortical sites during visuo-motor sequence learning—a transition of activity from parietal to parietal–premotor and then to premotor cortex and a concomitant shift was observed from anterior putamen to a combined activity in both anterior and posterior putamen and finally to posterior putamen. These results suggest a putative role for engagement of different cortical and subcortical networks at various stages of learning in supporting distinct sequence representations. Keywords: Sequence representation; Anterior striatum; Posterior striatum; DLPFC; Pre-SMA; SMA
An fMRI Study of the Role of the Medial Temporal Lobe in Implicit and Explicit Sequence Learning
Neuron, 2003
and Brain longer-term, explicit episodic retrieval of longer se-Department of Psychology quences, memory accounts implicate the MTL (Squire Boston University and Zola-Morgan, 1991). For implicit sequence learning, Boston, Massachusetts 02215 both frameworks implicate the striatum. While motor 2 MGH-NMR Center models also include the supplementary motor area Department of Radiology (SMA), parietal lobe, and cerebellum (Willingham, 1998; Harvard Medical School Middleton and Strick, 2000), some memory accounts Charlestown, Massachusetts 02129 implicate MTL structures in certain types of implicit learning (Curran, 1997; Cohen and Eichenbaum, 1993). We used fMRI to investigate the role of the human Summary MTL in implicit and explicit sequence learning. fMRI data were acquired while subjects performed a serial reaction fMRI was used to investigate the neural substrates time task (SRTT; Figure 1A), developed originally by Nissupporting implicit and explicit sequence learning, fosen and Bullemer (1987). In the SRTT, learning results in cusing especially upon the role of the medial temporal faster response times (RTs) for repeated than for random lobe. Participants performed a serial reaction time sequences of cued locations. task (SRTT). For implicit learning, they were naive For implicit SRTT learning, convergent evidence impliabout a repeating pattern, whereas for explicit learncates subcortical and cortical components of frontostriing, participants memorized another repeating seatal pathways. Patients with striatal dysfunction are imquence. fMRI analyses comparing repeating versus paired on implicit SRTTs (Knopman and Nissen, 1991; random sequence blocks demonstrated activation of Vakil et al., 2000; Jackson et al., 1995; Doyon et al., 1997, frontal, parietal, cingulate, and striatal regions impli-1998). Neuroimaging studies using an implicit SRTT with cated in previous SRTT studies. Importantly, mediohealthy adults have shown activation in the caudate, temporal lobe regions were active in both explicit and putamen (Rauch et al., 1995, 1997a; Hazeltine et al., implicit SRTT learning. Moreover, the results provide 1997; Grafton et al., 1995; Willingham et al., 2002; Peigevidence of a role for the hippocampus and related neux et al., 2000) and ventral striatum (Berns et al., 1997; cortices in the formation of higher order associations Doyon et al., 1996). The caudate has been proposed to under both implicit and explicit learning conditions, be important for stimulus-response association (Polregardless of conscious awareness of sequence drack et al., 2001) and cognitive abilities, such as workknowledge. ing memory (Owen et al., 1998). Activation has also been found in cortical components of frontostriatal circuits, Introduction including the DLPFC, parietal lobe, premotor cortex, anterior cingulate, and SMA (Rauch et al., 1995, 1997b; Sequence learning is used for behaviors like typing, mu-Berns et al., 1997; Grafton et al., 1995, 1998; Hazeltine sical performance, and route navigation. Researchers et al., 1997; Willingham et al., 2002; Peigneux et al., have described the acquisition of perceptuomotor se-2000). quencing skills using either motor control (e.g., Hazeltine For explicit SRTT learning, while striatal activation has et al., 1997) or learning and memory frameworks (e.g., rarely been noted, neuroimaging studies consistently Reber and Squire, 1998). Both explanations agree that find activation in cortical components of frontostriatal distinct brain processes support explicit learning, which circuits, including the DLPFC, ventrolateral prefrontal, occurs with awareness, and implicit learning, which ocpremotor, anterior cingulate, and dorsal and inferior pacurs without awareness. However, the two accounts rietal cortices (Hazeltine et al., 1997; Jenkins et al., 1994; diverge over which brain systems are important. Most Grafton et al., 1995; Rauch et al., 1995; Willingham et critically, only memory frameworks posit a role for the al., 2002). Motor accounts posit the DLPFC controls mediotemporal lobe (MTL) in sequence learning. strategic processes throughout the explicit SRTT, and For explicit sequence learning, both motor control and it can be recruited during an implicit SRTT, if participants memory accounts implicate the dorsolateral prefrontal become aware of a sequence (Willingham, 1998). This cortex (DLPFC). This region, by motor accounts, supidea resembles an explicit-implicit variety of memory ports conscious executive motor control to select goals account but with the DLPFC, not MTL, being necessary or to select and maintain a spatial sequence in working for conscious sequence acquisition. memory (Willingham, 1998; Grafton et al., 1995; Ha-Memory, but not motor, accounts consider the MTL zeltine et al., 1997) or, by memory accounts, supports system to be necessary for learning sequences, espethe manipulation and monitoring functions of working cially those beyond the capacity of DLPFC processes memory (Smith and Jonides, 1999). Memory models, of working memory. Each account, however, posits a however, also posit that the role of the DLPFC is limited somewhat different role for MTL. For explicit learning, explicit-implicit (or declarative-nondeclarative) memory accounts state that the MTL is necessary for long-term
Learning hierarchically structured action sequences is unaffected by prefrontal-cortex lesion
Experimental Brain Research, 2006
This study tested the impact of prefrontalcortex lesion on learning hierarchically structured action sequences. Using a visual-manual serial reaction time task, we had subjects Wrst perform Wve blocks of trials with a hierarchically structured 14-element action sequence and then tested for sequence-speciWc learning by introducing a pseudo-random transfer sequence. Relative to control subjects (N = 39), we found that both lateral frontal (N = 16) and medial frontal (N = 18) patients showed reduced overall performance beneWts across the training phase. In contrast, the negative transfer test showed signiWcantly increased reaction times in all patient groups, indicating robust sequence-speciWc learning. This learning was not sig-niWcantly diVerent from that of the control group. Taken together, the data suggest that learning hierar-chically structured action sequences is unimpaired in patients with prefrontal-cortex lesion.
Dynamic cortical involvement in implicit and explicit motor sequence learning. A PET study
Brain, 1998
We examined the dynamic involvement of different brain regions in implicit and explicit motor sequence learning using PET. In a serial reaction time task, subjects pressed each of four buttons with a different finger of the right hand in response to a visually presented number. Test sessions consisted of 10 cycles of the same 10-item sequence. The effects of explicit and implicit learning were assessed separately using a different behavioural parameter for each type of learning: correct recall of the test sequence for explicit learning and improvement of reaction time before the successful recall of any component of the test sequence for implicit learning. Regional cerebral blood flow was measured repeatedly during the task, and a parametric analysis was performed to identify brain regions in which activity was significantly correlated with subjects' performances: i.e. with correct recall of the test sequence or with reaction time. Explicit learning, shown as a positive correlation with the correct recall of the sequence, was associated with increased activity in Abbreviations: PRE ϭ training block; rCBF ϭ regional CBF; RND ϭ random condition block; SEQ ϭ sequence condition block; SMA ϭ supplementary motor area; SM1 ϭ primary sensorimotor cortex; SRTT ϭ serial reaction time task; VIS ϭ visual control condition block © Oxford University Press 1998 the posterior parietal cortex, precuneus and premotor cortex bilaterally, also in the supplementary motor area (SMA) predominantly in the left anterior part, left thalamus, and right dorsolateral prefrontal cortex. In contrast, the reaction time showed a different pattern of correlation during different learning phases. During the implicit learning phase, when the subjects were not aware of the sequence, improvement of the reaction time was associated with increased activity in the contralateral primary sensorimotor cortex (SM1). During the explicit learning phase, the reaction time was significantly correlated with activity in a part of the frontoparietal network. During the post-learning phase, when the subjects achieved all components of the sequence explicitly, the reaction time was correlated with the activity in the ipsilateral SM1 and posterior part of the SMA. These results show that different sets of cortical regions are dynamically involved in implicit and explicit motor sequence learning.