Vibrotactile imagery : neural correlates of older adults and patients suffered from stroke

Abstract

The present study examined the neural processes associated with the imagery of vibrotactile sensation of older adults. It was hypothesized that vibrotactile imagery was composed of an image generation process and an image maintenance process. These two processes would elicit event-related potential in the frontal parietal and central areas of the brain. It was further hypothesized that the image maintenance process was task and modality specific, and hence would elicit event-related potential in the sensorimotor areas. To further verify these hypotheses, event-related potentials of a few patients who suffered from a stroke and resulted in different brain lesions were recruited. The results obtained from these patients were compared with those from the older adult group. Twelve normal older adults and three post-stroke patients participated in both the imagery and control tasks. The imagery task required the subjects to imagine a vibrotactile sensation for 4000 ms primed by a 250 ms vibrotactile stimulus. Another stimulus was then delivered right after the imagery of the first stimulus, and the subjects were required to differentiate between the two stimuli. Under the control condition, the subjects were required to passively detect the vibrotactile sensation without imagining it. The event related potential was recorded using a 32-array of silver-silver chloride electrodes placed in an extended International 10-20 system. The accuracy and response time was also recording during the recording. Among the normal older adult, an ERP component was observed within the 150 - 200 ms time window (N170) which had its maximum in the frontal central area for both the imagery and control conditions. This suggests probable attention and detection of the vibrotactile stimuli a short while after the stimulus was delivered to the subjects. The second ERP component was found to be within the 280 - 390 ms time window (P300) of which the amplitudes for the imagery condition were less positive-going than the control condition in all sites, especially the frontal and central area. There were two ERPs appeared around the 480 - 620 ms and 640 - 970 ms which suggest possible N400 and P600 components. The N400 component was found less negative-going at the parietal and occipital sites, whilst the P600 component was less positive-going at the frontal and central sites in the control condition. These P300, N400 and P600 components, with the dipoles modelled and located at the frontal and parietal areas, indicate a continual processes from generation of vibrotactile images which was dominated by the P300 and N400 components to maintenance of the images in the working memory which was dominated by the P600 component. These observations and processes are consistent with the "imagery network" described in the literature. The lesions of in the cerebellum, thalamus, and frontal and temporal lobes in the three post-stroke patients were found to separately modulate different neural processes associated with imagery of vibrotactile sensation. Stronger modulation effects P600 were observed in the patient who had lesions over the frontal and temporal areas, which are consistent from the findings based on normal older adults. Future research should use neuroimaging techniques to identify the neural substrates associated with vibrotactile imagery and test the efficacy of using mental imagery to promote recovery of somatosensory deficits in post-stroke patients.