ناتوانی در ادراک بیماری برای همی پلژی به عنوان یک کسری حرکتی در آگاهی موتور: شواهد از اندام غیر فلج
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|38890||2010||8 صفحه PDF||سفارش دهید||6614 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Neuropsychologia, Volume 48, Issue 12, October 2010, Pages 3443–3450
Abstract The current study adds to the growing empirical research into the mechanisms underlying unawareness of paralysis following stroke (anosognosia for hemiplegia or AHP) by investigating action awareness for the non-paralysed limb in a single AHP patient. Visual feedback representing patient GG's goal-directed reaching movements was either modified by a computer or left unperturbed. Unlike healthy and brain-damaged controls, GG was unable to detect computer-generated visual perturbations as large as 20°. GG also failed to report awareness of the large on-line corrective movements that he made when compensating (often unsuccessfully) for the visual perturbations. These results suggest that the motor comparators implicated in AHP are functioning, but not at optimum levels. Moreover, because the current findings reveal a deficit in awareness for reaches with the unimpaired limb, it is suggestive of common right hemisphere networks for motor awareness in both limbs and that AHP may be a global deficit in motor awareness as opposed to a specific lack of awareness for a particular motor deficit
Introduction The past decade has seen remarkable growth in our understanding of the way in which we move and interact with the world around us. This subject is not only central to identifying the mechanisms responsible for our own actions, but provides considerable insight into how we interpret the intentions and desires of other people (Blakemore and Decety, 2001 and Blakemore et al., 2004). These abilities are fundamental to human social interaction. A driving force behind recent discoveries has been the development of a computational ‘forward’ or ‘comparator’ model of motor control, based on well-established engineering principles (Miall & Wolpert, 1996). According to this model, the control and awareness of movement relies on the comparison of information derived from various sources (see Fig. 1). It is generally accepted that the motor control system contains at least three ... Fig. 1. It is generally accepted that the motor control system contains at least three comparators comparing and responding to differences between the desired and predicted state (C1) the desired and estimated actual state (C2) and the estimated actual and predicted states of the movement effector (C3). Frith's version argues that gross discrepancies in C2 and C3 are ignored or are unavailable whereas Berti and Pia argue that comparator C3 itself is damaged. Crucially, for comparator models of agency, it is discrepancies in comparator C3 that indirectly evokes the evaluation that a movement or event has been externally produced (Synofzik, Vosgerau & Newen, 2008). Figure options Provided that we reach our desired state, such comparisons typically occur with limited conscious awareness (Fourneret & Jeannerod, 1998). The comparisons are thought to serve several crucial functions, such as allowing the continual fine-tuning of movements during their planning and execution, detecting when errors in movement occur, and correctly discriminating our own actions (self) from those of another person (other) (Wolpert, 1997). These proposals are supported by the evidence from healthy and brain-injured populations using behavioural experiments (e.g. Blakemore, Frith, & Wolpert, 1999), functional neuroimaging (e.g.m), cortical stimulation/disruption (e.g. Preston & Newport, 2008a) and neuropsychological data (Fotopoulou et al., 2008 and Synofzik et al., 2010). However, the exact number, functions, and underlying brain mechanisms of the comparators remains equivocal. Patients with abnormal motor awareness provide a unique opportunity to cast light on these unresolved issues. In particular, stroke patients who are not aware of severe motor impairments (i.e. patients with anosognosia for hemiplegia, or AHP) can help us to better understand the functional neuroanatomy of motor awareness. Recent forward/comparator model accounts of AHP suggest that the underlying cause of the disorder is a failure to detect discrepancies between patients’ predicted and (estimated) actual state (Berti and Pia, 2006 and Frith et al., 2000). Erroneous claims that the patient can move their paralysed limb are believed to occur because awareness in AHP is constructed entirely from intact predictions of intended movement (Fotopoulou et al., 2008 and Jenkinson et al., 2009). Such discrepancies usually trigger the mechanism responsible for conscious awareness and self-correction of the error (Fourneret & Jeannerod, 1998); however, AHP seems to represent an instance of pathological (lesion-induced) unawareness of large discrepancies between the predicted and actual state of the body, such that normal monitoring is impaired and awareness is deceived (Jenkinson & Fotopoulou, 2010). Accounts of AHP disagree on the source of this failure to detect errors. Frith et al. (2000) propose impaired input to the comparators responsible for monitoring concordance of the estimated actual state of the limb with the predicted and desired states. Alternatively, Berti and Pia (2006) suggest that the actual vs. predicted states comparator itself is damaged with the main distinction between the two theories being the certainty with which the comparator is held responsible (see Fig. 1 and Jenkinson & Fotopoulou, 2010 for further discussion of this point). However, no study to date has directly assessed the functionality of the comparator(s) in AHP. What is more, the suggestion of an impaired motor comparator may not be so straightforward. The observed specificity of AHP (e.g. differential perceived abilities for the arm and leg within patients; Marcel, Tegner, & Nimmo-Smith, 2004), and apparent independence of verbal/behavioural (Jehkonen, Laihosalo, & Kettunen, 2006) or implicit/explicit awareness (Cocchini, Beschin, Fotopoulou, & Della Sala, 2010) suggests that certain comparators may be impaired in AHP, while others remain intact. Furthermore, reports of AHP following unilateral left-hemisphere lesions (Cocchini, Beschin, Cameron, Fotopoulou, & Della Sala, 2009) argue against explanations that have located the mechanisms responsible for conscious error detection exclusively in the right hemisphere (e.g. Preston and Newport, 2008b and Ramachandran, 1995). More experiments are needed to further investigate these unresolved issues. Observations in patients with AHP, such as those described above, have been useful in providing insight into the workings of the healthy motor system; however, they present only indirect evidence regarding the possible operation of comparator mechanisms. The current experiment directly explores the functioning of the motor comparator (s) in a case of chronic AHP, primarily using a movement agency task to assess the patient's ability to detect and correct movement errors, by examining the performance using the unimpaired limb. AHP performance was compared with that of a group of hemiplegic patients without anosognosia (patient controls), and young healthy controls. The task involved making self-other judgements about observed reaching movements involving the intact (non-hemiplegic) hand, during which the participants received visual feedback of a visually coincident cursor that was spatially perturbed or unperturbed. The ability to detect discrepancies was assessed by asking the participants to state whether the movement seen was self (unperturbed) or other (perturbed). Furthermore, kinematic data regarding the participants’ movement trajectory was used to objectively assess reach accuracy and on-line correction. If damage to a general right-hemisphere motor comparator is responsible for impaired awareness in AHP, then self-other judgements should be selectively disrupted in the AHP patient regardless of which arm was involved in reaching. That is, if awareness of motor actions for both the left and right hands predominantly involves a right hemisphere network (e.g. Preston and Newport, 2008a, Preston and Newport, 2008b and Ramachandran, 1995), then damage to this network should also have implications for awareness of movements performed by the non-paralysed limb and should not be restricted to the paralysed contralesional limb for which the patient exhibits anosognosia. Furthermore, selective impairment of the comparator that provides information for high-level monitoring of self-other judgements might spare other low-level comparators responsible for the automatic updating and correction of movement errors, in which case subjective self-other judgements and reach accuracy measures should dissociate. The current study considered both of these measures in order to tease apart the different components of motor awareness.
نتیجه گیری انگلیسی
Results 5.1. Judgement data Self/other judgments were converted into percent self-scores for each participant at each perturbation size. The overall pattern of responses for the healthy control participants exhibited a relatively normal distribution with the peak of self-responses being centred on 0° perturbations. This pattern of results corresponds well with the results of similar paradigms using this measurement method (Farrer, Bouchereau, Jeannerod, & Franck, 2008). GG, on the other hand, recorded a totally flat response profile, responding at 100% self for each perturbation size and falling outside the 95% confidence intervals calculated for both the healthy and neglect control groups for each perturbation (Fig. 4). For reaches with the larger perturbation sizes (only performed by GG), again, GG's performance is clearly abnormal, responding with self-judgments on 100% of the trials whereas the patient control group were exhibiting improved detection of the perturbations at only 8°. Patient GG's mean % self-responses (crosses) plotted against the mean (and 95% ... Fig. 4. Patient GG's mean % self-responses (crosses) plotted against the mean (and 95% confidence intervals) for young healthy controls (open circles) and elderly brain-damaged controls (filled squares) at each of the 0°, 4° and 8° perturbations. Figure options It is notable that the performance of the neglect patients appears to be skewed and shifted such that the peak for self-responses falls to the left of centre. This is not simply a group effect as each of the patients produced a similar performance. The pattern is consistent with theories of neglect based on a compression of space from left to right (Halligan & Marshall, 1991). If the visual representation of the limb in left space is remapped as being further rightward than its veridical position, then it might be perceived as being closer to the proprioceptively felt position of the limb than it really is–giving rise to an increase in self-judgements. Cursors appearing in central and right space, on the other hand, might be perceived as more rightward than the veridical position, potentially leading to a reduction in self-responses to zero and rightward perturbations. 5.2. Accuracy and correction judgements GG was very accurate when reaching in unperturbed trials. End-point errors (calculated in the lateral × dimension measured from the seen cursor position, not the actual hand position) in this condition were very small (median absolute end point error = 1.9°), and smaller than all of the other patients (range of medians = 3.5–6.2°) including the reaches of the most closely matched patient control, KM (median = 3.5°). This demonstrates that his ability to integrate sensorimotor information in order to plan and execute reaches remained intact. Inspection of GG's handpaths in the perturbed conditions, however, revealed that many of his reaches exhibited large and late on-line corrections. His errors across all perturbation sizes ranged from 0.03° to 24.94° with a median of 3.1° (the median error increased to 12.6° for the largest 20° perturbations). Note that an error of ∼12° for the cursor in the 20° perturbation condition represents an error of 8° for the true position of the hand). Despite these substantial inaccuracies, sometimes accompanied by very large corrective movements (see Fig. 5), GG rated his perception of reach accuracy as always being accurate (100% accurate responses) for all perturbations while his reporting of motor corrections remained fixed at 0% (never having to correct). Although it was not tested directly, these data would argue against an account positing that GG's deficit was caused by a failure to successfully integrate visual and proprioceptive information regarding the limb. In any case, such impairment could only be in addition to, and not instead of, a failure to be aware of the consequences of his own reaching movements. Hand paths (left panels) and velocity profiles (right panels) of typical reaches ... Fig. 5. Hand paths (left panels) and velocity profiles (right panels) of typical reaches from the last two blocks of GG's reaches. For some reaches, GG made on-line movement corrections (left hand panel a), while for other reaches his motor corrections were secondary movements (right hand panel b). Note the re-accelerations (secondary increases in velocity late in the movement) in panel b. Regardless of the error or magnitude of correction, GG always reported his reaches as being accurate and uncorrected. Dashed lines represent the trajectory of the actual limb and solid lines represent the cursor trajectory. Figure options 5.3. Kinematic data The first two blocks of GG's reaches were compared to those of control patient KM. Only the first 36 of KM's reaches were used because 4 trials were lost from GG's data due to GG letting go of the vBOt handle, causing early termination of the trial. The first two blocks were chosen because they could be directly matched for perturbation size and KM was chosen as the most closely matched of the control patients (in regards to age, cognitive score, physical disability). Paired samples t-tests demonstrated that the overall profile of reach kinematics were very similar between the two, with no significant differences found for peak velocity (t(35) = 0.26, p = 0.8) (GG mean = 336.2 mm/s, S.D = 79.5 mm/s; KM mean = 341.5 mm/s, S.D = 73.9 mm/s) or movement time (t(35) = 1.44, p = 0.16) (GG mean = 1280 ms, S.D = 210 ms; KM mean = 1200 ms, S.D = 220 ms).