Implicit Body Representation of the Hand Enlarged by Repetitive Peripheral Magnetic Stimulation within the Boundary of a Real Hand

Featured Application

Our work provided a simpler paradigm for measuring implicit body representation of the hand and highlighted its potential biomarker function for clinical practice in neurorehabilitation as it could be enlarged by repetitive peripheral magnetic stimulation within the boundary of a real hand.

Abstract

Deafferentation induced by local anesthesia causes a larger perceived area than the real area of the mouth, which, in the perspective of body representation, belongs to implicit body representation. In this study, we applied repetitive peripheral magnetic stimulation (rPMS) on the motor branch of the radial nerve of participants’ non-dominant-side forearm to induce extension movements of wrist and fingers. This intervention was supposed to increase proprioception to the brain and had an enlargement effect on implicit body representation of the hand in our hypothesis. A total of 39 participants were randomly allocated to the real rPMS group (n = 19) or the sham rPMS group (n = 20). Implicit representation of the hand was measured by a simplified paradigm based on the proposal of Longo and Haggard that depicted perceived locations of fingertips and metacarpophalangeal joints of participants’ occluded hand, in which they showed that implicit body representation of the hand was smaller than the real hand. We compare the main effect of real rPMS vs. sham rPMS and its interaction effect with time by setting four timepoints—before stimulation, right after stimulation, 10 min after stimulation and 20 min after stimulation—to demonstrate the possible short-lasting effect. Results showed that real rPMS had a short-lasting enlargement effect on implicit representation of the hand in general, which was significant especially on the ulnar side of fingers. What is more, the enlarged implicit body representation of the hand was still within the boundary of a real hand, which might indicate the identification role of a real body part.

1. Introduction

Robust hand function is indispensable for our daily life and work. It is also of crucial significance in neurorehabilitation, for the impairment of hand function that is required for performing delicate movements makes it difficult to complete daily activities, such as eating, dressing, and return to work [1,2]. During these activities, the human body is a specific mental category within which interoceptive and exteroceptive signals are integrated and processed faster comparing surrounding objects of similar complexity [3,4], and the human mind creates and stores distinct representations of the body and peri-personal objects [5,6,7]. Body representation was not only found to be misperceived in healthy adults [8], but also distorted in numerous pathological conditions, including spinal cord injury [9], amputation [10], cerebral palsy [11], and stroke [12,13]. It has been assumed that the plasticity of body representations is important concerning functional motor improvement by its integration with the motor control system, possibly through the activation of the mirror neuron system secondary to the visuomotor feedback provision [13]. Although body representation deficits are not rare in stroke patients [6,14,15], evidence that shows the plastic change of body representations in stroke patients is so far scarce—even pilot studies in healthy people are insufficient. Recent evidence suggested that there are at least three distinct types of body representation: body schema, body structural descriptions, and body semantics [14,16,17]. Body schema encompass a dynamic representation of the relative position of body parts derived from multiple sensory and motor inputs (e.g., proprioceptive, vestibular, tactile, visual, and efference copy) that interacts with the motor system in the genesis of actions. Body structural descriptions encompass topological maps of locations derived primarily from visual input that define body part boundaries and proximity relationships. Body semantics are lexical-semantic representations of the body including body part names, functions, and relations with artifacts [14,16,17]. Despite this triadic taxonomy and previously used dyadic taxonomy, which distinguishes between body schema and body image (probably the combination of body structural description and body semantics), Longo and Haggard developed an implicit body representation underlying position sense by isolating and measuring the perceived location of landmarks in healthy people [18]. Participants in their research were asked to use a baton to indicate with their right hand ten the landmarks of their occluded left hand: the knuckle at the base of each finger and the tip of each finger. The perceived locations of landmarks were captured as still JPEG images that depicted the implicit perceptual maps of hand structure that were constructed and compared to the actual hand structure. These maps were highly distorted in a stereotyped way across participants, with the perceived upper part of the hand represented as wider and shorter than the real hand [18], neglecting the part between knuckles and wrist joint.

Repetitive peripheral magnetic stimulation (rPMS) is a unique non-invasive treatment method that was developed for therapeutic neuromodulation in movement disorders by activating proprioceptive afferents [19,20,21,22]. rPMS can be performed to safely stimulate deeper regions of tissue with little activation of cutaneous receptors and has been proven as a feasible approach for improving activities of daily living and functional abilities in people with hemiplegic paralysis [21,23]. A recent study found that single-session rPMS applied to stroke patients’ upper limbs could alleviate spasticity in both elbow flexors and wrist flexors, and upper limb function measured according to the Fugl–Meyer assessment was improved [24]. What is more, the power of event-related desynchronization after rPMS decreased in the mu rhythm band (8–12 Hz), which may suggest certain favorable change in terms of its neurological effects on the central nervous system [24]. Gallasch et al. observed changes in blood oxygen-dependent contrast after rPMS, revealing its short-lasting modulation effect on the sensorimotor cortex [25].

Body representations are formed and maintained based on the input of multiple somatosensory information [26]; in terms of implicit body representation, perceiving the body part that cannot be seen is mainly relied on proprioception [27,28]. There is a similar research paradigm that movements of the mouth cannot be guided by vision, while after the acute somatosensory deafferentation induced by local anesthesia, implicit body representations of the mouth were larger than the real size [29,30]. In summary of previous research, implicit body representation of the fingers is shorter than real fingers; however, local anesthesia-mediated loss of somatosensory information makes implicit body representations of the mouth overshoot the boundary of the real mouth. Based on these interesting phenomena, we aimed to study whether adding proprioception from the peripheral body part could induce plastic change in implicit body representation. Therefore, we hypothesized that rPMS applied to the motor branch of the radial nerve that induced extensions of the wrist and fingers could increase proprioception input to the corresponding cortex area in the brain, which might enlarge implicit body representation of the hand in return.

2. Materials and Methods

2.1. Experimental Design

We performed this cross-sectional study in Tohoku University, which was mainly to compare the effects of single-session of real rPMS and single-session of sham rPMS on implicit body representation of the hand in two matched groups in healthy adults, adopting and simplifying the paradigm of a localization task developed by Longo and Haggard [18,31]. To avoid neglecting the part between the knuckles and the wrist joint of the hand, we measured the perceived length of the wrist and perceived lengths from each fingertip to wrist joint, and the percentage of each perceived length to its corresponding real length, as indices of implicit body representation of the hand. In both groups, we measured implicit body representation of the non-dominant hand before stimulation, right after stimulation, 10 min after stimulation and 20 min after stimulation. Previous research showed that implicit body representation of the hand between two hands had no difference [18]; however, where rPMS has an enlargement effect on implicit body representation of the hand, we chose the non-dominant hand as it might have greater potential to be improved by rPMS.

2.2. Participants

A total of 40 healthy participants (1 participant in the real stimulation group dropped out due to oversensitive to rPMS) were recruited through posters in the campus of Tohoku University and randomly divided into the real stimulation group and the sham stimulation group, with all participants blinded to this information. A total of 39 participants were included (age = 25.7 ± 3.1 years, 23 males)—19 participants in the real stimulation group (age = 24.8 ± 2.6 years, 13 males) and 20 participants in the sham stimulation group (age = 26.6 ± 3.3 years, 10 males). Student’s t test and the chi-square test showed no statistical differences in age and gender between groups, respectively (p < 0.05).

2.3. Apparatus

The repetitive peripheral magnetic stimulation machine used in this study, (Institute of Field Generation, Sendai, Japan, Japan Medical Device Number (Type II): 36902000, Japan Medical Device Identification Code: 227AFBZX00021000) consisted of a generator and a round coil that had a cube appearance. The diameter of the coil was 8 cm. The hand perception device for measuring implicit body representation of the hand was self-made, consisting of a wooden hollow box and a wooden base; see Figure 1a The plane of the hollow box was at an angle of 60° to the base. The hollow box measured 20 cm in height, 10 cm in width, and 5 cm in depth, with an opaque board A facing the participants and a board B back to the participants. An A4-sized white paper was pasted on board A; see Figure 1b There was a thin line and a big dot on the paper. The given line was parallel to the short edges of the paper, and the given dot was located on that line. The distance of the given line to the near short edge of the paper and the distance of the given dot to the near endpoint of the given line were both 6 cm. The lower short edge of the paper overlapped the bottom of board A. The left and right positions of the paper relative to board A were fixed.

2.4. Experimental Procedures

Participants sat in front of a table in a quiet and bright room with the hand perception device placed flat on the table. Before the experiment began, participants fully understood all the procedures with detailed explanation given by the experimenter and practiced on how to use the device. Participants were asked to put their non-dominant hand into the hollow box and adjust the height of their elbows until they felt the given line and the given dot on the paper overlapped the baseline and the far-ulnar edge of their wrist joint, respectively; see Figure 1a. A towel was used to support the elbow if necessary with the position recorded for use in the following experiment. After participants were instructed to place their hand correctly, the experimenter instructed participants to sequentially draw six landmarks on the paper with a marker pen. The first landmark is the perceived far-radial edge of the wrist joint on the given line, followed by the perceived fingertips of the thumb, index, middle, ring, and little fingers above the given line; see Figure 2. Participants were instructed to perceive implicit body representation of the hand twice (suspend condition and touch condition) four times (before, right after, 10 min after and 20 min after stimulation) in total. In the real stimulation group, the center of the round coil was placed anterior or lateral to the head of the radius, which was the corresponding skin position of the motor branch of the radial nerve. The site of coil could be adjusted properly to obtain maximum extension movement of the wrist joint during the practice stage. Parameters of the rPMS machine were set at 50 Hz, and the maximum intensity each participant could tolerate (ranging from 50% to 80% of the highest intensity of the machine across participants). Each trial had a 2 s stimulation time, and there was a 1 s interval time between trials. Single-session of 36 trials in total was performed without interruption. In the sham stimulation group, the cube-shaped coil was rotated 90°, while other experimental procedures and parameters of rPMS machine were the same as in the real stimulation group to produce noise. All participants were asked to close their eyes during stimulations.

2.5. Statistical Analysis

The Shapiro–Wilk test showed that all the data were normally distributed (p > 0.05). To elucidate the statistical differences in perceived lengths (i.e., perceived wrist length and perceived hand lengths of where five fingers locate) between groups, multivariate analysis of variance (MANOVA) with repeated measures was performed. The within-subject factor was time (i.e., before stimulation, right after stimulation, 10 min after stimulation and 20 min after stimulation). The between-subject factor was group (i.e., the real rPMS group and the shamrPMS group). MANOVA was also performed to determine whether there were significant differences in the percentage of perceived lengths compared to real lengths. Where Mauchley’s test indicated a violation of the sphericity assumption, Greenhouse–Geisser correction was applied. Bonferroni correction was applied for multiple comparisons. If there was an interaction effect of time and stimulation identified through MANOVA, then a paired t-test was adopted for post hoc analysis to compare results for before stimulation with right after stimulation, 10 min after stimulation, and 20 min after stimulation, respectively, in both the real rPMS group and the sham rPMS group.

Analyses were conducted using SPSS version 26.0 (IBM Inc., Chicago, IL, USA). Continuous variables are presented as the mean ± standard deviation. A p-value of <0.05 (two-sided) was considered to indicate statistical significance.

3. Results

3.1. Perceived Lengths (cm)

Statistical analysis results showed that single-session real rPMS compared with single-session sham rPMS had an enlargement effect on perceived lengths (i.e., perceived wrist length and perceived hand lengths of where the thumb, index, middle, ring and little finger locate) in general. MANOVA showed that the within-subject effect of time (Pillai’s Trace) was statistically significant, F (18, 21) = 1.706, p = 0.037, ηp² = 0.087. There was an interaction effect between time and group (Pillai’s Trace), F (18, 21) = 1.940, p = 0.013, ηp² = 0.097. We looked further into the interaction effect between time and group in the univariate test, as shown in

Table 1. Perceived wrist length and perceived hand lengths of where fingers locate (cm).

	Sham rPMS				Real rPMS
	before	Right after	10 Min after	20 Min after	before	Right after	10 Min after	20 Min after	p	ηp2
PL0	4.55 ± 0.66	4.57 ± 0.86	4.69 ± 0.79	4.48 ± 0.69	4.18 ± 1.04	4.39 ± 0.81	4.39 ± 0.78	4.42 ± 0.72	0.151	0.046
PL1	7.40 ± 2.14	7.31 ± 1.99	7.41 ± 2.07	7.21 ± 1.87	7.26 ± 2.08	7.52 ± 2.14	7.61 ± 2.35	7.67 ± 2.13	0.309	0.032
PL2	11.70 ± 2.77	11.91 ± 3.02	11.78 ± 2.99	11.56 ± 2.68	10.71 ± 2.58	11.17 ± 2.57	11.05 ± 2.93	11.35 ± 2.60	0.155	0.046
PL3	12.80 ± 2.80	12.89 ± 2.94	12.71 ± 2.91	12.61 ± 2.77	11.79 ± 2.71	12.37 ± 2.60	12.11 ± 3.02	12.43 ± 2.77	0.051	0.067
PL4	11.68 ± 2.69	11.70 ± 2.86	11.53 ± 2.85	11.25 ± 2.75	10.46 ± 2.53	11.19 ± 2.47	10.98 ± 2.92	11.28 ± 2.68	0.004	0.111
PL5	10.13 ± 2.15	9.86 ± 2.70	9.68 ± 2.75	9.43 ± 2.53	8.69 ± 2.54	9.17 ± 2.51	8.76 ± 2.90	9.08 ± 2.63	0.027	0.079

This table shows perceived length. PL0 represents perceived wrist length, PL1 to PL5 represent perceived hand lengths of where the thumb, index, middle, ring and little finger locate, respectively. p represents the p-value of the interaction between time and group, and Bonferroni correction was applied for multiple comparisons. ηp² represents partial eta square.

. Enlargement of perceived wrist length and hand lengths of where the thumb, index, and middle finger locate was not statistically significant; perceived wrist length, real rPMS (4.35 ± 0.16) compared with sham rPMS (4.54 ± 0.16), F (3, 36) = 1.803, p = 0.151, ηp² = 0.046; perceived hand length of where the thumb locates, real rPMS (7.48 ± 0.46) compared with sham rPMS (7.3 ± 0.45), F (3, 36) = 1.212, p = 0.309, ηp² = 0.032; perceived hand length of where the index finger locates, real rPMS (11.05 ± 0.62) compared with sham rPMS (11.71 ± 0.60), F (3, 36) = 1.779, p = 0.155, ηp² = 0.046; perceived hand length of where the middle finger locates, real rPMS (12.14 ± 0.63) compared with sham rPMS (12.72 ± 0.61), F (3, 36) = 2.675, p = 0.051, ηp² = 0.067. However, perceived hand lengths of where the ring and little finger locate were enlarged by real rPMS with statistical significance; perceived hand length of where the ring finger locates, real rPMS (10.95 ± 0.60) compared with sham rPMS (11.48 ± 0.59), F (3, 36) = 4.611, p = 0.004, ηp² = 0.111; perceived hand length of where the little finger locates, real rPMS (8.88 ± 0.59) compared with sham rPMS (9.73 ± 0.57), F (3, 36) = 3.189, p = 0.027, ηp² = 0.079.

3.2. The Percentage of Perceived Lengths to Corresponding Real Lengths (%)

Percentages were calculated by comparing each perceived length to its corresponding real length, which indicates the relationship between implicit body representation and a real body part. Results showed that real rPMS compared with sham rPMS also had an enlargement effect on the percentage of perceived lengths to corresponding real lengths in general. What is more, all the perceived lengths were shorter than real lengths, before stimulation, right after stimulation, 10 min after stimulation, and 20 min after stimulation. MANOVA showed that the within-subject effect of time (Pillai’s Trace) was statistically significant, F (18, 21) = 1.713, p = 0.036, ηp² = 0.087). There was an interaction effect between time and group F (18, 21) = 1.922, p = 0.014, ηp² = 0.096). We looked further into the interaction effect between time and group in the univariate test of percentage, as shown in

Table 2. Percentage of perceived lengths to corresponding real lengths (%).

	Sham rPMS				Real rPMS
	before	Right after	10 Min after	20 Min after	before	Right after	10 Min after	20 Min after	p	ηp2
PP0	89.2 ± 16.1	89.3 ± 18.6	92.0 ± 18.9	87.8 ± 16.5	82.8 ± 19.4	87.3 ± 15.7	87.4 ± 15.8	88.0 ± 14.5	0.123	0.051
PP1	64.3 ± 17.8	63.5 ± 16.5	64.4 ± 17.4	62.5 ± 15.0	60.9 ± 17.4	63.2 ± 17.7	63.9 ± 19.3	64.4 ± 17.8	0.321	0.031
PP2	69.7 ± 14.8	70.9 ± 15.7	70.3 ± 16.3	69.0 ± 14.6	62.6 ± 14.9	65.2 ± 14.5	64.6 ± 16.8	66.4 ± 15.3	0.145	0.047
PP3	72.0 ± 13.9	72.4 ± 14.1	71.6 ± 14.9	71.0 ± 14.0	64.9 ± 14.5	68.0 ± 13.6	66.6 ± 16.0	68.5 ± 15.0	0.044	0.070
PP4	70.3 ± 14.1	70.3 ± 14.8	69.5 ± 15.6	67.7 ± 15.1	60.9 ± 13.7	65.2 ± 13.2	64.0 ± 16.1	65.8 ± 15.0	0.004	0.114
PP5	73.2 ± 19.6	71.3 ± 16.7	70.2 ± 17.9	68.4 ± 16.4	60.2 ± 16.5	63.4 ± 16.0	60.6 ± 18.8	62.9 ± 17.2	0.024	0.081

This table shows the percentages of perceived length to their real lengths; PP0 represents the percentage of perceived wrist length compared to real wrist length, PP1 to PP5 represent the percentages of perceived hand lengths of where the thumb, index, middle, ring and little finger locate compared to corresponding real lengths, respectively. p represents the p-value of the interaction between time and group, and Bonferroni correction was applied for multiple comparisons. ηp² represents partial eta square.

. Enlargement of the percentage of perceived wrist length compared to real wrist length, the percentage of perceived hand lengths of where the thumb, and index finger locate compared to corresponding real hand lengths was not statistically significant; perceived wrist length, real rPMS (86.5 ± 3.4) compared with sham rPMS (89.0 ± 3.3), F (3, 36) = 1.971, p = 0.123, ηp² = 0.051; perceived hand length of where the thumb locates, real rPMS (62.8 ± 3.8) compared with sham rPMS (63.3 ± 3.7), F (3, 36) = 1.179, p = 0.321, ηp² = 0.031; the perceived hand length of where the index finger locates, real rPMS (64.6 ± 3.4) compared with sham rPMS (69.8 ± 3.3), F (3, 36) = 1.837, p = 0.145, ηp² = 0.047. However, the percentages of perceived hand lengths of where the middle, ring, and little finger locate were statistically enlarged by real rPMS; the perceived hand length of where the middle finger locates, real rPMS (66.8 ± 3.2) compared with sham rPMS (71.6 ± 3.2), F (3, 36) = 2.793, p = 0.044, ηp² = 0.070; the perceived hand length of where the ring finger locates, real rPMS (63.8 ± 3.2) compared with sham rPMS (69.1 ± 3.1), F (3, 36) = 4.755, p = 0.004, ηp² = 0.114; the perceived hand length of where the little finger locates, real rPMS (61.5 ± 3.7) compared with sham rPMS (70.5 ± 3.6), F (3, 36) = 3.268, p = 0.024, ηp² = 0.081.

3.3. Short-Lasting Effect of Real rPMS

Because of the interaction effect of groups and time found through MANOVA, we performed a paired t-test as a post hoc analysis to identify at which timepoint (i.e., before stimulation vs. right after stimulation, before stimulation vs. 10 min after stimulation, before stimulation vs. 20 min after stimulation) the rPMS effect was statistically significant, multiple comparisons was corrected by Bonferroni, and α level was set to 0.167. In the real rPMS group, for perceived hand length of where the ring finger locates, before stimulation (10.46 ± 2.53) compared with right after stimulation (11.19 ± 2.47), p = 0.017, compared with 10 min after stimulation (10.98 ± 2.92), p = 0.115, compared with 20 min after stimulation (11.28 ± 2.68), p = 0.015; for perceived hand length of where the little finger locates, before stimulation (8.69 ± 2.54) compared with right after stimulation (9.17 ± 2.51), p = 0.074, compared with 10 min after stimulation (8.76 ± 2.90), p = 0.848, compared with 20 min after stimulation (9.08 ± 2.63), p = 0.208; for the percentage of perceived hand length of where the middle finger locates to corresponding real length, before stimulation (64.89 ± 14.54) compared with right after stimulation (68.03 ± 13.59), p = 0.075, compared with 10 min after stimulation (66.63 ± 16.02), p = 0.310, compared with 20 min after stimulation (68.47 ± 14.54), p = 0.035; for the percentage of perceived hand length of where the ring finger locates to corresponding real length, before stimulation (60.93 ± 13.70) compared with right after stimulation (65.17 ± 13.19), p = 0.017, compared with 10 min after stimulation (63.99 ± 16.12), p = 0.110, compared with 20 min after stimulation (65.80 ± 15.00), p = 0.014; for the percentage of perceived hand length of where the little finger locates to corresponding real length, before stimulation (60.16 ± 16.47) compared with right after stimulation (63.38 ± 15.95), p = 0.085, compared with 10 min after stimulation (60.55 ± 18.78), p = 0.868, compared with 20 min after stimulation (62.92 ± 17.23), p = 0.207. See Figure 3. In the sham rPMS group, paired t-test of before stimulation compared with right after stimulation, 10 min after stimulation, and 20 min after stimulation, none of the comparisons were of statistical significance.

4. Discussion

The main finding of this study was that single-session real rPMS compared with sham rPMS had an enlargement effect on implicit body representation of the hand in general, which might be attributed to the proprioception input induced directly and indirectly by real rPMS. When the motor branch of the radial nerve was activated by magnetic stimulation, there were two possible proprioception inputs from the peripheral body to the central nervous system; one was generated by the radial nerve and conveyed directly from the stimulation site through the radial nerve; the other was produced by the extension movements of wrist joint and fingers secondary to magnetic stimulation, then uploaded to the brain [32,33]. Therefore, our hypothesis was validated that by inducing the extension movements of the wrist joint and fingers, proprioception was inputted contributed to the enlargement of implicit body representation of the hand. Further, this enlarged implicit body representation of the hand was still within the boundary of a real hand. As mentioned in the introduction, local anesthesia can also enlarge implicit body representation of the mouth, beyond the boundary of a real mouth. Combined with these findings, we propose that intact and even enhanced implicit body representation is conservative within the boundary of a real body part, while loss of real-time somatosensory information input (mainly the proprioception) makes the conservative implicit body representation hard to maintain, thus overshooting the boundary of a real body part. Recent research also supports the above interpretation that perceptual distortion of the mouth was larger after local anesthesia, and after repetitive transcranial magnetic stimulation applied to the mouth representation area of somatosensory cortex, the overshooting perceptual distortion decreased to approaching the original perceived size, even though this Intervention was given directly to the brain [34].

We noticed that this enlargement effect was significant in the ulnar-side fingers. In this study, perceived hand lengths of where the ring and little finger locate, and the percentages of perceived hand lengths of where the middle, ring and little finger locate to their corresponding real lengths were enlarged by rPMS. This result was in accordance with previous research [18,35] that the percent overestimation of perceived finger length was observed in healthy adults, decreasing from the radial to the ulnar side of the hand. This enlargement effect was especially significant in the ring finger rather than in the little finger, which might be because the thumb and index finger are used more frequently in people’s daily life and work and have a relatively bigger somatosensory cortex areas in the brain, while the little finger is the least used finger and occupies a relatively smaller somatosensory cortex area; therefore, a bigger somatosensory cortex might be able to deal with a higher resolution of input signals that form more accurate body representation of the corresponding body part [36].

Former research demonstrated that the perceived lengths of the five fingers were shorter than for real fingers [18,37,38], findings which were consistent with our results. Moreover, in another research paradigm of a knuckle localization task, perceived positions of knuckles were farther forward than real positions of participants’ hands [39,40]. Therefore, we might be able to conclude from previous research that: first, perceived lengths between fingertips and metacarpophalangeal joints were shorter than real lengths; second, perceived lengths between metacarpophalangeal joints and the wrist were longer than real lengths. However, the perceived lengths of the whole hand remained unclear. Our result showed for the first time that the perceived length of a hand was shorter than the real hand. In addition, perceived wrist length was not influenced by rPMS and was close to real wrist length, which might be attributed to the experimental settings where the forearm could be seen during the experiment. However, the function of perceiving the wrist joint was that it provided a baseline and reference for perceiving the fingertips in a systematic order, which could contribute to more accurate judgements [36]. To sum up, we adopted six landmarks rather than ten landmarks in this study as a modified paradigm for measuring the perceived lengths of the whole hand, which might facilitate future experimental or practical applications.

Our study observed a short-lasting effect of rPMS on implicit body representation of the hand. In the real rPMS group, before stimulation compared with right after stimulation and 10 min after stimulation, the enlargements of implicit body representation of the hand were not so obvious; however, before stimulation compared with 20 min after stimulation, the enlargement effect was statistically significant. Gallasch et al. found that with transcranial magnetic stimulation as a detection method, single and paired pulse motor evoked potential (MEP) recruitment curves were increased post-stimulation up to 1 h and intracortical facilitation was increased up to 30 min [25]; however, they adopted 20 min of 25 Hz rPMS application which was supposed to have longer-lasting effect than in this study. Nito et al. also demonstrated that a significant increase in MEP was observed up to 60 min following 15 min of 50 or 25 Hz rPMS [41]. Jia, Y et al. found that single-session 20 Hz rPMS could modulate corticomotor excitability by measuring MEP, together with a possible lasting improvement in hand dexterity [42]. We observed a 20 min lasting effect of single-session 50 Hz rPMS on implicit body representation of the hand, which is in accordance with the above research.

This study researched the implications of implicit body representation of the hand in the progress of neurorehabilitation practice. It has been reported that goggles maximize or minimize vision, which can affect the perceived size of certain objects by healthy participants; however, this effect was reduced when participants placed their hands beside the object, even though the visual size of their own hands was also enlarged or shrunk by the goggles [43]. This finding indicates that representation of the hand in the mind plays a calibration role when humans interact with the surrounding environment. The visual hand size being maximized or minimized by the goggle did not affect the calibration role of the participants’ real hand. In other words, it might be the implicit body representation of the hand that was not affected by the goggles, which undertook the important function of calibration. Since implicit body representation of the hand was enlarged by rPMS within the boundary of a real hand, we wonder whether the calibration function of the hand could be strengthened as well. Our next study project is to explore the correlations between hand functions and implicit body representation, both in healthy adults and chronic stroke patients, in order to develop new neurorehabilitation therapies based on implicit body representation.

This study had some limitations. First, we adopted only single-session rPMS with 36 trials, and because this is a pilot study to our knowledge, it is not clear whether there is a difference in the plastic change in implicit body representation of the hand after multi-session or long-term application of rPMS from what we observed in this study. Second, we did not explore whether there is a difference in implicit body representation of the hand when using the dominant side as opposed to the non-dominant side. This needs to be verified scientifically.

5. Conclusions

This study demonstrated the enlargement effect of single-session rPMS on implicit representation of the hand in healthy adults. Implicit body representation was conservative when intact or even after being enhanced by rPMS; however, acute loss of somatosensory information induced by local anesthesia can lead to overshooting the boundary of a real body. This finding provided a new understanding of the facilitative effect of rPMS and the identification role of the boundary of a real body. We observed that the short-lasting effect of rPMS is consistent with previous research, which should be noticed when applying this result to further research or practical applications. We also proposed a simpler paradigm for measuring implicit body representation of the whole hand based on the Longo and Haggard protocol. The perspective of implicit body representation might shed light on developing new methods for future neurorehabilitation.