The Lateral Occipito-temporal Cortex Is Involved in the Mental Manipulation of Body Part Imagery

Abstract

The lateral occipito-temporal cortex (LOTC), including the extrastriate body area, is known to be involved in the perception of body parts. Although still controversial, recent studies have demonstrated the role of the LOTC in higher-level body-related cognition in humans. This study consisted of two experiments (E1 and E2). The first (E1) was an exploratory experiment to find the neural correlate of the mental manipulation of body part imagery, in which brain cerebral glucose metabolic rates and the performance of mental rotation of the hand were measured in 100 subjects who exhibited a range of symptoms of cognitive decline. In E1, we found that the level of glucose metabolism in the right LOTC was significantly correlated with performance in a task involving mental manipulation of the hand. Next, in E2, we performed a randomized, double-blind, controlled intervention study (clinical trial number: UMIN 000018310) in younger healthy adults to test whether right occipital (corresponding to the right LOTC) anodal stimulation using transcranial direct current stimulation (tDCS) could enhance the mental manipulation of the hand. In E2, we demonstrated a significant effect of tDCS on the accuracy rate in a task involving mental manipulation of the hand. Although further study is necessary to answer the question of whether these results are specific for the mental manipulation of body parts but not non-body parts, E1 demonstrated a possible role of the LOTC in carrying out the body mental manipulation task in patients with dementia, and E2 suggested the possible effect of tDCS on this task in healthy subjects.

Keywords: extrastriate body area; hand imagery; lateral occipito-temporal cortex; mental rotation; transcranial direct current stimulation.

FIGURE 1

Task paradigms. (A) Choice reaction time (CRT) task. The participants were instructed to select the right or left picture that included a circle (i.e., a target) by pressing a button with their right or left hand as soon as possible after a circle appeared on the screen. (B) During the recognition phase of the simple visual working memory task (WM task), the participants were instructed to select the picture that depicted the same hand as that presented during the acquisition phase (non-flipped images). (C) During the recognition phase of the visual working memory with mental rotation task (RWM task), the participants were instructed to select the picture that depicted the same hand, which was flipped from the palm side (acquisition) to the back side (recognition). Each condition consisted of 10 trials with different hand shape pictures and lasted for a duration of 3 min.

FIGURE 2

(A) Stimulation targets (i.e., LOTC and DLPFC] marked on the reconstructed scalp and brain surface of one subject in the active condition. (B) Schematic representation of the experimental design. Each participant began to perform the CRT, WM, and RWM tasks after receiving active or control stimulation for 10 min. The stimulation continued throughout the duration of the three tasks and for an additional 10 min. CRT, choice reaction time. WM, working memory. RWM, working memory with hand mental rotation. LOTC, lateral occipito-temporal cortex. DLPFC, dorsolateral prefrontal cortex.

FIGURE 3

SPM analyses with multiple regression models in which the accuracy rate (A) or the inverse efficiency (IE) scores (B–D) for the WM conditions were used as independent variables. (A) A decrease in the [¹⁸F]FDG-SUVR in the right frontal cortices was associated with poorer performance (i.e., lower accuracy rate) in the WM task. However, this significant association disappeared if we employed a conservative statistical threshold set at p = 0.05 (corrected with FWE). (B,D) A decrease in the [¹⁸F]FDG-SUVR in the frontal cortices was associated with poorer performance (i.e., higher IE) in the WM task. (C) This significant correlation was still observed in the right frontal cortex if we employed a conservative statistical threshold set at p = 0.05 (corrected with FWE). The yellow color bar indicates the T-value. L, left hemisphere. R, right hemisphere. A, anterior. P, posterior.

FIGURE 4

SPM analyses with multiple regression models in which the accuracy rate (A) or the IE scores (B–D) for the RWM conditions were used as independent variables. (A) A decrease in the [¹⁸F]FDG-SUVR in the occipital and parietal cortices was associated with poorer performance (i.e., lower accuracy rate) in the RWM task. However, this significant association disappeared if we employed a conservative statistical threshold set at p = 0.05 (corrected with FWE). (B,D) A decrease in the [¹⁸F]FDG-SUVR in the occipital and parietal cortices was associated with poorer performance (i.e., higher IE) in the RWM task. (C) This significant correlation was still observed in the bilateral occipito-temporal cortices if we employed a conservative statistical threshold set at p = 0.05 (corrected with FWE). The yellow color bar indicates the T-value. L, left hemisphere. R, right hemisphere. A, anterior. P, posterior.

FIGURE 5

(A) Accuracy scores in the three tasks (i.e., CRT, WM, and RWM) with the application of active (red circles) and control (blue circles) transcranial direct current stimulation (tDCS). The RWM task scores during right occipital anodal stimulation (i.e., the active condition, n = 20) were significantly higher than those during the control stimulation (n = 20). (B) If the control condition was divided into two conditions, i.e., the reversed condition (frontal anodal/occipital cathodal stimulation; n = 10) (blue circles) and sham condition (n = 10) (green circles), the RWM task scores during the active condition (n = 20) were significantly higher than those during the reversed condition (n = 10) but not than those during the sham condition (n = 10). The error bar indicates 1 standard error. CRT, choice reaction time. WM, working memory. RWM, working memory with hand mental rotation.