Working memory for vibrotactile frequencies: comparison of cortical activity in blind and sighted individuals

Abstract

In blind, occipital cortex showed robust activation to nonvisual stimuli in many prior functional neuroimaging studies. The cognitive processes represented by these activations are not fully determined, although a verbal recognition memory role has been demonstrated. In congenitally blind and sighted (10 per group), we contrasted responses to a vibrotactile one-back frequency retention task with 5-s delays and a vibrotactile amplitude-change task; both tasks involved the same vibration parameters. The one-back paradigm required continuous updating for working memory (WM). Findings in both groups confirmed roles in WM for right hemisphere dorsolateral prefrontal (DLPFC) and dorsal/ventral attention components of posterior parietal cortex. Negative findings in bilateral ventrolateral prefrontal cortex suggested task performance without subvocalization. In bilateral occipital cortex, blind showed comparable positive responses to both tasks, whereas WM evoked large negative responses in sighted. Greater utilization of attention resources in blind were suggested as causing larger responses in dorsal and ventral attention systems, right DLPFC, and persistent responses across delays between trials in somatosensory and premotor cortex. In sighted, responses in somatosensory and premotor areas showed iterated peaks matched to stimulation trial intervals. The findings in occipital cortex of blind suggest that tactile activations do not represent cognitive operations for nonverbal WM task. However, these data suggest a role in sensory processing for tactile information in blind that parallels a similar contribution for visual stimuli in occipital cortex of sighted.

Figure 1

Timing for image sequences and task paradigms. A. Six sequential vibration stimulation trials were presented during 32‐s ON cycles followed by 24‐s OFF cycles. Each On and 20 s during the following OFF cycle was analyzed as a single epoch. Five On/Off cycles were presented during each ∼5.3 min run. B. One‐back vibration frequency task. The question for the one‐back task was whether the frequency in two successive trials matched. C. Amplitude‐change task. A baseline vibration amplitude was maintained (A1, A4, or A6), increased (A2 or A5), or decreased (A3) during an On cycle. The question for the amplitude‐change task was whether vibration amplitude changed during a trial. A button response was required for stimulation trials 2–6. The trial 1 stimulation was attended but not responded to. The button pressed with the middle finger meant “yes” and with the index finger indicated “no” for the questions posed by each task.

Figure 2

Group t‐maps for each group and task were computed after registering volumetric uncorrected F‐statistic z‐scores to the surface nodes of the inflated and flattened average PALS‐B12 surfaces [Van Essen, 2005; Van Essen and Dierker, 2007]. The distribution of cortical activity per task for each group was determined by averaging the z‐scores per PALS‐B12 atlas node. A t‐test assessed whether average z‐scores per node significantly differed from a z‐score population mean of zero [Bosch, 2000]. Scale shows t‐values for 9 df per node (P values 0.0007–0.0001). IT, inferior temporal cortex; ITS, inferior temporal sulcus; IPS, intraparietal sulcus; MOg, middle occipital gyrus; PH, parahippocampal gyrus; PO, parietal operculum; PCG, postcentral gyrus; SFS, superior frontal sulcus.

Figure 3

Regions of interest (ROI) indicated on average PALS‐B12 flattened and inflated surfaces. Each ROI contains a unique population of surface nodes. Cortical regions were previously defined and included Brodmann areas [Drury et al., 1999; Van Essen, 2005], parietal opercular subdivisions [Burton et al., 2008a; Eickhoff et al., 2006], selected functional partitions of several Brodmann areas [Burton et al., 2008b], and occipital cortex visuotopic subdivisions [Van Essen, 2005; Van Essen and Dierker, 2007]. Letter tags are identified in Table III.

Figure 4

Response time courses during the working memory (Freq) and amplitude‐change (Amp) tasks from defined right hemisphere (RH) ROI in early blind (EB) and normally sighted (NS). Data points show mean and standard error of the mean (SEM) from 10 participants per group. Significant P values for main effects or interactions are shown below graphs.

Figure 5

Response time courses during the working memory task (Freq) from defined ROI in both hemispheres and from early blind (EB) and normally sighted (NS). Data points show mean and standard error of the mean (SEM) from 10 participants per group.

Figure 6

Response time courses during the working memory and amplitude‐change detection tasks from selected parietal cortex ROI in both hemispheres and from early blind (EB) and normally sighted (NS). Data points show mean and standard error of the mean (SEM) from 10 participants per group. PCG, postcentral gyrus; OP 1, parietal operculum subdivision 1; BA 40, Brodmann area 40.