Could experience modulate imagery of limb movements? : a case in individuals with spinal cord injury

Abstract

The mechanisms underlying reorganization of the neural system due to paralysis of the lower limbs after spinal cord injury (SCI) remains unclear. This study aims to use functional imaging to investigate the neural changes brought by the loss of physical movements and sensory feedback in the lower limbs among a group of chronic SCI participants. The participants were 11 adult patients who suffered from SCI at the T7-T11 level, resulting in complete paralysis of the lower limbs. The control group was composed of 13 healthy participants with matched demographic characteristics. The experimental task used was visuomotor imagery tasks requiring participants to engage in visualization of repetitive tapping movements of the upper or lower limbs. The task processes involved retrieval of visuomotor images of the limbs, visualization of tapping of upper or lower limbs in working memory, and inspection of the direction of movements of the designated limb. The tapping movements had three rhythmic patterns of 0.8, 1.0, and 1.33 Hz, respectively. A typical trial began with three auditory cues presented at one of the three rhythmic patterns. The participant was to follow the rhythm of the tones and begin visualizing the tapping movement of upper or lower limbs, one after the other, for 2.1 to 4.4 s. After hearing a high-pitched tone, the participant was to pause the visualized movements and indicate which side of the limb was toward the platform at that instance by pressing a button on a keyboard. The duration of capturing blood oxygen-level dependent (BOLD) responses by the scanner was 2.0 s, beginning from the presentation of the third auditory cue. There were two conditions: upper and lower limb. Accuracy rate and mean response time were the behavioral parameters of the task. The participant received training on the tapping movements and gained an accuracy rate reaching at least 70% before proceeding to the scanning session. Clinical measures on cognitive functions and post-SCI impairments were administered to the participants. Behavioral data, including the visuomotor task and clinical measures, were compared between the SCI and healthy control group. Between-group effects on the BOLD responses elicited from the task conditions were tested. The relationships between the BOLD responses and the behavioral and clinical variables were explored. It was hypothesized that, in the lower limb condition, the SCI participants would display stronger BOLD responses than the healthy control participants in the motor-related subcortical structure such as the basal ganglion and other regions outside of the sensorimotor areas. This would reflect possible neural changes among the SCI participants due to the post-injury loss of sensorimotor experience from the lower limbs. It was also hypothesized that, when compared with the healthy control group, the SCI participants would have stronger BOLD responses elicited in the sensorimotor areas for imagery under the upper limb condition. This would reflect the post-SCI reorganization of the neural system as a result of the experience-dependent plastic changes of the motor system. No significant between-group differences were revealed in the accuracy rates and response times on the upper and lower limb task conditions. The SCI participants had significantly lower performances on the tests concerning working memory (Rey Verbal Auditory Learning Test) and executive functions (Trail Making Test). The main findings of this study were in the significantly stronger BOLD responses elicited in the left lingual gyrus among the SCI participants compared to the healthy control participants when engaging in imagining lower limb movements. The right external globus pallidus (GPe) also showed significantly stronger activations among the SCI participants. No between-group differences, however, were revealed in the BOLD responses in the sensorimotor areas. The stronger activations in the GPe suggested plausible increases in relaying input to and output from the basal ganglion during the visuomotor imagery processes. Stronger activations of the GPe suggested possible sub-cortical excitability among the SCI participants under the lower limb condition. The stronger activations in the left lingual gyrus indicated increases in the involvement of visual function during the visuomotor imagery for the SCI participants. This was supported by the stronger activations in the middle occipital gyrus in the lower limb versus the upper limb condition contrast among the SCI participants compared to the healthy control participants. These findings suggested possible compensatory strategies adopted by the SCI participants for the post-injury loss of sensorimotor inputs from the lower limbs. A similar strategy would have been used when the SCI participants visualized the repetitive movements of the lower limbs. The lack of increase in BOLD responses within the sensorimotor areas in the lower limb condition was likely to be attributable to the diminished movement feedbacks experienced by the SCI participants. For the upper limb condition, the results indicate that the SCI participants showed significantly stronger BOLD responses than the healthy control participants in extensive areas over the brain including the bilateral right precentral gyrus, the left postcentral gyrus, the right middle frontal gyrus, the bilateral superior temporal gyrus, the right superior and inferior parietal gyrus, the right external GPe, and the thalamus. The stronger BOLD responses in the widely distributed bilateral sensorimotor areas for the SCI participants in the upper limb condition suggested possible systematic post-injury changes of the motor control system. The stronger activations found in the GPe in the upper limb condition were comparable to those in the lower limb condition. This suggested that the post-injury neural changes were likely to be at a systemic level, influencing both upper and lower limbs. In contrast, the healthy control participants displayed significantly stronger BOLD responses than the SCI participants in the frontal areas including the middle frontal gyrus, the medial frontal gyrus, the inferior frontal gyrus, and the left anterior cingulate gyrus for both the upper and lower limb conditions. These neural substrates were by and large mediating visuomotor imagery processes such as working memory, motor inhibition, and set shifting. These corresponded to the declined working memory and executive functions among the SCI participants as reflected from the significantly lower performances on the clinical measures when compared with the healthy control participants. The present study supported the notion that the SCI participants probably underwent post-spinal-cord-injury plasticity in neural substrates, mediating visuomotor imagery of upper and lower limbs. The results highlighted the significant post-injury changes in the responses of the GPe within the basal ganglion and its involvement in the visuomotor imagery. Without sensorimotor inputs, as the lower limbs had been paralyzed, the SCI participants were found to rely more on the visual system to undergo the visuomotor imagery. These plastic changes may have other impacts on neural processes among the SCI participants. Future studies should investigate how the post-injury plastic changes among SCI individuals would impact the preparation and execution of upper and lower limb functions by comparing complete and incomplete lesions.