ایجاد اختلال در کنترل توجه حرکات در طول یادگیری حرکتی با آسیب فرونتال

Abstract This study examined the effects of frontal lobe lesions on the control of movements during motor learning. We compared the performance of patients with unilateral frontal or temporal excisions and controls in two-dimensional aiming movements during adaptation to a transformed visuomotor mapping. Subjects tried to reach a fixed target on a graphics tablet using indirect visual control from a monitor in either: (1) the standard visuomotor mapping, (2) a full inversion of motor space preserving the axis of movement, or (3) a mirror-like inversion of one axis of motor space. In the standard mapping, all groups showed precise and rapid aiming movements. In the full inversion condition, frontal lobe patients showed a stronger tendency than others to initiate movements in the natural direction (capture errors) during adaptation. In the mirror-like inversion, frontal patients showed deficits in both movement initiation and movement corrections. These control deficits disappeared with practice. These data provide evidence for a critical role of frontal cortex in the attentional control of unpracticed movements in man.

Introduction The execution of simple visually guided movements is little affected by lesions to premotor or prefrontal cortex [16], [17], [47] and [55]. This contrasts with the significant effects of parietal lesions which often affect the precision of pointing, reaching and manipulation movements [3], [15], [26] and [42] This lack of effect of frontal lesions is also surprizing in light of the increased activity in these structures before and during simple visually guided movements as evidenced from single-cell recordings in monkeys [29] and [37] and functional imaging of regional activation in man [8], [43] and [49]. Frontal lesions can affect movement programming and selection in special conditions. For example, frontal lesions affect the programming and execution of sequential movements in monkeys and man [5], [12], [25], [31], [33], [34], [35], [39] and [57]. Frontal lesions can also affect movements which require inhibiting prepotent automatic movements such as antisaccades [22] or spatially-inverted choice responses [11]. In monkeys, premotor lesions can impair indirect reaching movements around a transparent obstacle which involves inhibiting direct reaching movements [36]. Finally, frontal lesions produce deficits in the selection of movements associated with arbitrary cues, which can be interpreted as a deficit in movement selection in novel stimulus-response mappings [11], [23], [24], [44] and [45]. A common characteristic of movements affected by frontal lesions is that their context or sensorimotor mapping is unfamiliar and that they thus require a more important contribution from voluntary control. Frontal cortex has long been implicated in the more voluntary or attentional aspects of actions as opposed to the more automatic or well-learned aspects [17], [40] and [53]. However, few studies have tested this dissociation directly in visually-guided movements by comparing the performance of unpracticed and practiced movements. In novel unpracticed situations, performance depends a great deal on attentional control, while after acquisition attentional control is much less necessary as well-learned programs take over control of many portions of movements [28], [46], [51] and [56]. One reason that frontal lesions produce so few problems in visually guided movements may be that most of these movements are well practiced. This study examined whether frontal lesions affect unpracticed movements during motor learning in man. Some evidence points to a role of frontal cortex in unpracticed movements. For example, cerebral activation in man appears to increase in frontal areas during the acquisition of motor skills [14], [19], [20], [21], [32], [41], [50] and [52]. If frontal cortex activity is necessary for the attentional control of movements, frontal lesions should impair unpracticed movements in a motor learning situation. Unpracticed movements can be observed during adaptation to new sensorimotor mappings and some studies have examined sensorimotor adaptation after frontal cortical damage. Movements during prism adaptation have shown mixed results, some showing no effects of frontal lesions [58]; others showing some adaptation deficits after frontal lesions but few initial performance problems [6]. Depending on the procedure used, prism adaptation may involve adaptation of several motor modalities including eye, head and or arm movements at different moments in the task and may therefore be too complex to directly address the question of attentional motor control problems. Some studies have examined mirror-reversed movements. Some case studies have shown adaptation problems after frontal lesions and some have not [1], [4] and [10]. We recently examined mirror tracing performance in patients with frontal excisions [7]. In this situation, frontal lobe patients were slower than temporal lobe patients as expected but also showed more frequent oscillatory movements, suggesting an impaired visuomotor control. However, mirror tracing provides a very coarse measure of sensorimotor control. The present study was designed to provide a more direct test of the effect of human frontal damage on inverted movements. We compared the performance of patients with unilateral frontal lobe lesions to that of patients with temporal lobe lesions on simple visually guided aiming movements during adaptation to a transformed visuomotor mapping. If frontal lesions affect attentional motor control, patients with frontal lesions should show problems in inverted movements during adaptation but not after learning when the newly formed programs can take over control of the performance.

Results 3.1. Initial direction of movement The initial direction of aiming movements was quantified by the direction of the third sampling point of the movement. This point was chosen because it occurred 240 ms into the movement before any feedback-based correction of the movement is possible. Initial directions were highly variable both across trials and across subjects and they were thus analyzed using three categories: (1) adequate initiations, defined as initial directions within 30° of the perfect direction; (2) capture errors, defined as initial directions within 30° of the direction of the visually presented target; and (3) other erroneous directions. The 30° criterion was used because all control subjects showed initial directions within this interval after adaptation in all conditions. In the baseline condition, all subjects showed an adequate initiation in all 24 trials. In the full inversion condition, capture errors were observed mostly in the first trials. Capture errors were averaged for trial quadruplets 2–4 so as not to include initial directions that were not based on previous exposure to the transformed space (quadruplet 1) and because of the very low frequency of these errors after trial quadruplet 4. The frontal group produced significantly more frequent capture errors than the other two groups in the full inversion condition (frontals vs temporals: Mann–Whitney U=30.0, P<0.02; temporals vs controls: U=49.5, n.s.). No other types of errors were observed in this condition. In the single-axis inversion condition, both capture errors and other directional errors were observed (see Table 1). These two types of errors were averaged over quadruplets 2–8 because of their very low frequency after quadruplet 8. The analyses indicated that the frontal group made significantly more capture errors than other groups (F vs T: U=32.5, P<.03, T vs C: U=37.5, n.s.) but not more of other types of errors (F vs T: U=42.0, n.s.; T vs C: U=50.0, n.s.). Table 1. Mean error rates (%) in the initial direction of aiming movements in the full inversion condition and in the single-axis inversion condition in patients with frontal lesions, patients with temporal lesions and controls Frontal Temporal Control Full inversion Capture errors 25.0 8.3 8.3 Single-axis inversion Capture errors 25.0 10.7 3.6 Other errors 39.3 28.6 14.3 Table options 3.2. Movement precision The precision of aiming movements was measured using the total length of the movement. Fig. 2 shows group averages of movement length for consecutive trial quadruplets in the three conditions. In the baseline condition, movements were very close to a straight line in all groups. An ANOVA on the first and last trial quadruplets showed no significant group differences [F(2,30)=1.8, n.s.], no significant effect of practice [F(1,30)<1.0, n.s.], and no interaction [F(2,30)<1.0, n.s.]. Average length of aiming movements as a function of practice in the normal ... Fig. 2. Average length of aiming movements as a function of practice in the normal visuomotor space, in a full inversion of visual feedback, and in a mirror-like inversion of feedback. Figure options In the full inversion condition, an ANOVA on the first and last trial quadruplets showed a significant effect of practice [F(1,30)=7.7, P<0.01], but no significant group effect [F(2,30)=1.6, n.s.], and no significant group-by-trial interaction [F(2,30)=1.6, n.s.]. To quantify the aftereffect of the full inversion condition, an ANOVA was performed on the first and last trials of baseline performance which followed the full-inversion condition. In this analysis, there was a significant effect of practice [F(1,30)=3.4, P=0.02], indicating that a measurable after effect was produced by the full-inversion condition, but there was no group effect [F(2,30)<1.0, n.s.] and no interaction [F(2,30)=1.4, n.s.]. In the single-axis inversion condition, frontal lobe patients appear to show abnormal movement lengths in the early phase of adaptation but not after practice. An ANOVA was performed on the first, middle, and last trial quadruplets to determine whether the performance of the groups became similar early or late in adaptation. The analysis showed that there was a significant group effect [F(2,30)=5.6, P<0.01; F vs T: t(20)=2.9, P<0.01; T vs C: t(20)=0.01, n.s.] and that all groups improved with practice [F(2,60)=48.4, P<0.0001]. There was also a significant Group-by-Trial interaction [F(4,60)=3.6, P=0.01] and simple effects showed that the performance of the frontal lobe patients had reached a normal level after the first half of practice [F vs T: t(20)=1.5, n.s.; T vs C: t(20)=1.0, n.s.]. 3.3. Movement speed Fig. 3 shows the measures of average movement speed obtained in the three conditions of the task. In the baseline condition, movement speed was very similar in the three groups. An ANOVA on the first and last trial quadruplet showed that movement speed increased with practice [F(1,30)=9.0, P<0.005], but there was no significant group effect [F(2,30)<1.0, n.s.] and no significant Group-by-Trial interaction [F(2,30)<1.0, n.s.]. In the full inversion condition, the same pattern of results was obtained: a significant practice effect [F(2,30)=34.1, P<0.0001], but no significant group effect [F(2,30)=1.0, n.s.] and no significant interaction [F(2,30)<1.0, n.s.]. Average speed of aiming movements as a function of practice in the normal ... Fig. 3. Average speed of aiming movements as a function of practice in the normal visuomotor space, in a full inversion of visual feedback and in a mirror-like inversion of feedback. Figure options In the single-axis inversion condition, average movement speed showed a relatively high variability throughout acquisition in all three groups but the temporal and control groups appeared to produce faster movements with practice. An ANOVA on the first, middle, and last trial quadruplets showed a significant practice effect [F(2,60)=16.6, P<0.0001], no significant group effect [F(2,30)=1.1, n.s.], and a significant interaction [F(4,60)=4.4, P<0.005]. Analyses of the simple effects suggest that both the temporal group and the control group showed a significant improvement of speed with practice [Temporal: t(10)=4.2, P<0.0002; Control: t(10)=4.6, P<0.0001] but that the frontal group did not [t(10)=0.02, n.s.]. We examined the individual performances of frontal lobe patients and found no differences between patients with right-sided lesions and patients with left-sided lesions nor between patients with sparing of premotor regions or medial frontal regions and other patients on any measure.