Influence of Static Magnetic Field Stimulation on the Accuracy of Tachystoscopically Presented Line Bisection

Transcranial static magnetic stimulation (tSMS) has been known to reduce human cortical excitability. Here, we investigated whether tSMS would modulate visuo-spatial cognition in healthy humans. Subjects performed a visuo-spatial task requiring judgements about the symmetry of pre-bisected lines. Visual stimuli consisted of symmetrically or asymmetrically transected lines, tachystoscopically presented for 150 ms on a computer monitor. Task performance was examined before, immediately after, and 10 min after tSMS/sham stimulation of 20 min over the posterior parietal cortex (PPC: P4 from the international 10–20 system) or superior temporal gyrus (STG: C6). Nine out of 16 subjects misjudged pre-bisected lines by consistently underestimating the length of the right-side segment (judging lines to be exactly pre-bisected when the transector was located to the left of the midpoint, or judging the left-side segment to be longer when the transector was located at the midpoint). In these subjects showing a leftward bias, tSMS over the right STG reduced the magnitude of the leftward bias. This did not occur with tSMS over the right PPC or sham stimulation. In the remaining right-biased subjects, no intervention effect was observed with any stimulation. Our findings indicate that application of tSMS over the right STG modulates visuo-spatial cognition in healthy adults.


Introduction
In a last decade, a number of studies have demonstrated that the excitability of cerebral cortices is suppressed by "transcranial static magnetic field stimulation (tSMS)" [1], in which the scalp is exposed to The Subject was seated comfortably in front of a computer monitor with a distance of 50 cm. The subject's eye level was adjusted to be at the middle of the computer monitor that was centred on his/her sagittal midplane. The subject performed a line length judgement task requiring forced-choice decision about the symmetry of pre-bisected lines. Visual stimuli consisted of symmetrically or asymmetrically transected lines, which were tachystoscopically presented on the computer monitor. Task performance was examined before (pre), immediately after (post), and 10 min after (post 10) tSMS or sham stimulation of 20 min.

Preliminary Experiment
Prior to the intervention experiment using tSMS/sham, 11 of 16 subjects (6 males and 5 females) performed the task with the stimulus presentation time of 50, 150, and 300 ms twice in a random order (6 sessions in total), to examine learning effects and performance reproducibility, and also to determine the appropriate stimulus presentation time. The stimulus presentation time of 150 ms, which showed the highest reproducibility, was used in the intervention experiment. This preliminary experiment was conducted more than a week before the intervention experiment.

Transcranial Static Magnetic Stimulation (tSMS)
For tSMS, we used a cylindrical NdFeB neodymium magnet (diameter, 50 mm; height, 30 mm) with a surface magnetic flux density of 534 mT, maximum energy density of 49 MGOe, and a strength of 862 N (88 kgf) nominal value (NeoMag, Ichikawa, Japan). We previously showed that this magnet generates a magnetic field that accesses most cortical regions (strength 110-90 mT at 2-3 cm from the surface of the magnet) and elicits biological effects [13]. For sham stimulation, we utilized a non-magnetic stainless-steel cylinder of the same size and weight. With the aid of an arm-type lightning stand (C-stand, Avenger, Cassola, Italy), the magnet or sham device was placed over P4 to stimulate the right PPC (Brodmann area 39) [27,29], and over C6 to stimulate the right STG (Brodmann areas 22 and 42) [36,37] according to the International 10-20 system for electrode placement, similar to previous rTMS and tDCS studies. tSMS effects are polarity independent [1] and, thus, we used only south polarity for all experiments. Sham stimulation was performed on P4 in eight participants and on C6 in another eight participants. The real and sham stimulations were performed in a random order. As reported in our previous study, the static magnetic field does not interfere with biological tissue at all; thus, the subjects cannot determine whether the object placed on the scalp is a magnet (real) or a non-magnetic material (sham) [1]. Since the visual stimulus presentation system created in this study provided the final score automatically, examiner bias was considered to be minimal.

Line Length Judgement Task
The visual stimuli and their presentation method were based on a previous study [34]. 1 mm thick, 145 to 155 mm long horizontal lines transected by a 1 mm thick, 10 mm long vertical line was presented on the center of the monitor (12-inch FHD (1920 × 1080 dots), Refresh rate of 59 Hz). At each trial, one of five lines was presented, differing in the position of the vertical transector (at midpoint, rightward, or leftward) and in the overall length of the line and of its right and left segments ( Figure 1). presented using Psychopy 2.0.1 (Open Science Tools, Nottingham, UK) [38], an open source application for building a psychological experiment environment based on the Python language.

Scoring of Task Performance
The performance of the subject in each trial was scored as in Table 1. Line number see caption of Figure 1.

Preliminary Experiment
A single measure of the interclass correlation coefficient, ICC (2,1), was used to measure the reproducibility of the task score (response) between two sessions for each stimulus presentation time. We examined the effect of stimulus presentation time (50 ms, 150 ms, and 300 ms) on the task score using a one-way repeated-measures analysis of variance (ANOVA). We also examine the effect of task session (1st to 6th) on the task score using a one-way repeated-measures ANOVA.

Intervention Experiment
All data were expressed as the mean ± standard error of the mean (SEM). The normal distribution was confirmed using the Kolmogorov-Smirnov test. The effects of stimulation site (P4, C6, and Sham) and time (pre, post, and post 10) on the task score were analyzed using a two-way repeated-measures ANOVA. Post hoc differences were further analyzed with Bonferroni's test. All analyses were performed with IBM SPSS Statistics software version 21 (SPSS; IBM, Armonk, NY, USA), and the significance level was set at 5%.

Inter-Trial Reproducibility of the Score in Line Length Judgement Task
Seven out of 11 subjects (4 males and 3 females) consistently underestimated the length of the right-side segment (judging lines to be exactly pre-bisected when the vertical transector was located To prevent eye scanning, the visual stimulus was presented for 150 ms (tachistoscopic presentation). Prior to the stimulus presentation, the subject was required to look at a fixation (an upward pointing arrow), which disappeared after appearance of the visual stimulus. After the stimulus presentation, the subject was required to make decision orally about the respective length of the two segments. There were three possible responses: equal, longer right, or longer left. In all task sessions, the subject performed 30 trials in a random order (10 trials with line #1 and five trials with lines #2-5). The interstimulus interval varied randomly between 8-12 s. The visual stimuli were presented using Psychopy 2.0.1 (Open Science Tools, Nottingham, UK) [38], an open source application for building a psychological experiment environment based on the Python language.

Scoring of Task Performance
The performance of the subject in each trial was scored as in Table 1.

Preliminary Experiment
A single measure of the interclass correlation coefficient, ICC (2,1), was used to measure the reproducibility of the task score (response) between two sessions for each stimulus presentation time. We examined the effect of stimulus presentation time (50 ms, 150 ms, and 300 ms) on the task score using a one-way repeated-measures analysis of variance (ANOVA). We also examine the effect of task session (1st to 6th) on the task score using a one-way repeated-measures ANOVA.

Intervention Experiment
All data were expressed as the mean ± standard error of the mean (SEM). The normal distribution was confirmed using the Kolmogorov-Smirnov test. The effects of stimulation site (P4, C6, and Sham) and time (pre, post, and post 10) on the task score were analyzed using a two-way repeated-measures ANOVA. Post hoc differences were further analyzed with Bonferroni's test. All analyses were performed Brain Sci. 2020, 10, 1006 5 of 14 with IBM SPSS Statistics software version 21 (SPSS; IBM, Armonk, NY, USA), and the significance level was set at 5%.

Inter-Trial Reproducibility of the Score in Line Length Judgement Task
Seven out of 11 subjects (4 males and 3 females) consistently underestimated the length of the right-side segment (judging lines to be exactly pre-bisected when the vertical transector was located to the left of the midpoint, or judging left-side segment to be longer when the vertical transector was located at the midpoint) (Leftward bias). In contrast, the remaining four subjects (2 males and 2 females) showed rightward bias. Inter-trial reproducibility of the score in a line length judgement task was ICC (2,1) = 0.909 for a stimulus presentation time of 50 ms, 0.946 for 150 ms, and 0.771 for 300 ms (p < 0.01) ( Figure 2). The highest reproducibility was demonstrated with a stimulus presentation time of 150 ms. There was no significant effect of stimulus presentation time or session order on the score ( Figure 3). The tendency of a leftward or rightward bias was consistent across a stimulus presentation times and sessions.
Brain Sci. 2020, 10, x FOR PEER REVIEW 5 of 14 task was ICC (2,1) = 0.909 for a stimulus presentation time of 50 ms, 0.946 for 150 ms, and 0.771 for 300 ms (p < 0.01) ( Figure 2). The highest reproducibility was demonstrated with a stimulus presentation time of 150 ms. There was no significant effect of stimulus presentation time or session order on the score ( Figure 3). The tendency of a leftward or rightward bias was consistent across a stimulus presentation times and sessions.   Figure 4 shows serial changes in individual total scores of line length judgement task before (pre), immediately after (post), and 10 min (post 10) after tSMS/sham stimulation for a period of 20 min (tSMS over P4 and C6 and sham). In line with the preliminary experiment, we found two subgroups showing either a leftward (9 of 16, 6 males and 3 females) or rightward bias (7 of 16, 4    Figure 4 shows serial changes in individual total scores of line length judgement task before (pre), immediately after (post), and 10 min (post 10) after tSMS/sham stimulation for a period of 20 min (tSMS over P4 and C6 and sham). In line with the preliminary experiment, we found two subgroups showing either a leftward (9 of 16, 6 males and 3 females) or rightward bias (7 of 16, 4 males and 3 females). The total score and the number of errors in total and in each score (−1, −2, +1, and +2) are summarized in Table 2 for the left-biased group and in Table 3 for the right-biased group.  Figure 4 shows serial changes in individual total scores of line length judgement task before (pre), immediately after (post), and 10 min (post 10) after tSMS/sham stimulation for a period of 20 min (tSMS over P4 and C6 and sham). In line with the preliminary experiment, we found two subgroups showing either a leftward (9 of 16, 6 males and 3 females) or rightward bias (7 of 16, 4 males and 3 females). The total score and the number of errors in total and in each score (−1, −2, +1, and +2) are summarized in Table 2 for the left-biased group and in Table 3 for the right-biased group.

Effects of tSMS on the Scores of Line Length Judgement Task
Brain Sci. 2020, 10, x FOR PEER REVIEW 6 of 14   Since including the task scores from two subgroups in one analysis can result in the mean total score close to zero, we analyzed the effects of stimulation site and time separately for two subgroups. The two-way repeated-measures ANOVA revealed a significant interaction between stimulation site and time for the left-biased group (F(4,32) = 2.898, p = 0.037, η 2 = 0.266). For the tSMS over C6, a post hoc analysis showed a significant difference between before and immediately after 20 min of tSMS (post) (p < 0.001) ( Figure 5). In addition, at immediately after 20 min of stimulation, a significant difference was revealed between two tSMS stimulation conditions (C6 and P4). No effect of stimulation site or time was observed for the right-biased group. Analysis without an outlier from the left-biased group can be found in Appendix A Figure A1.  Table 2. The total score and the number of errors in the line length judgement task for the left-biased group (mean ± standard error of the mean (SEM)). n = 9. * p < 0.001 vs. pre, † p < 0.05 vs. tSMS over P4.  Since including the task scores from two subgroups in one analysis can result in the mean total score close to zero, we analyzed the effects of stimulation site and time separately for two subgroups. The two-way repeated-measures ANOVA revealed a significant interaction between stimulation site and time for the left-biased group (F (4,32) = 2.898, p = 0.037, η 2 = 0.266). For the tSMS over C6, a post hoc analysis showed a significant difference between before and immediately after 20 min of tSMS (post) (p < 0.001) ( Figure 5). In addition, at immediately after 20 min of stimulation, a significant Brain Sci. 2020, 10, 1006 7 of 14 difference was revealed between two tSMS stimulation conditions (C6 and P4). No effect of stimulation site or time was observed for the right-biased group. Analysis without an outlier from the left-biased group can be found in Appendix A Figure A1. In the left-biased group, the magnitude of bias was significantly reduced immediately after tSMS over C6, and the bias was significantly smaller for C6 as compared to P4 at post. * p < 0.001 vs. pre, † p < 0.05 vs. tSMS over P4.

Discussion
In this study, 9 of 16 subjects consistently underestimated the length of the right-side segment in line length judgement task, and the remaining 7 subjects consistently underestimated the length of the left-side segment. In these subjects showing the leftward bias, there was an improvement in the task score after an application of tSMS over C6 but not over P4 or after sham stimulation. In the other subjects showing the rightward bias, no significant intervention effect was observed with any stimulation.

Leftward and Rightward Biases in Healthy Individuals
Neurologically healthy individuals tend to show a leftward bias [39]. This phenomenon is known as pseudoneglect and can be observed in various visuo-spatial tasks [40,41]. In contrast to this previous observation, 7 of 16 subjects showed a rightward bias in the present study. Although at a first glance our finding seems contradictory, enormous individual differences have been observed in a relatively large number of previous studies on pseudoneglect [42][43][44][45]. In particular, Manning and colleagues found a rightward bias in 10 of 22 subjects with a large between-subject variability and proposed that this large variability could be the cause of frequent unsuccessful replication of leftward bias in healthy individuals [46]. Also, Jewell and McCourt pointed out in their review that, in addition to the problem of individual differences, scanning, sex, age, and hand dominance could influence the amplitude of pseudoneglect, even though their meta-analysis demonstrated the existence of leftward bias [41]. Thus, what we observed in the present study is not necessarily unique. Supportively, in our preliminary study, all the subjects consistently showed either a leftward or rightward bias over all sessions regardless of the stimulus presentation time. Furthermore, intersession reproducibility of the bias was quite high. From these data, there are assumed to be a certain number of individuals who show a rightward bias. Moreover, the high consistency and reproducibility suggest that the change in the task score by an application of tSMS over C6 in the left-biased group was not due to changes in attention, habituation, or fatigue resulting from repeated sessions.

Putative Mechanisms Underlying Change in the Score of Line Length Judgement Task by tSMS Over C6
SMFs have constant intensity and direction over time with a frequency of 0 Hz, and are different from electromagnetic fields that vary over time. Although growing evidence indicates that SMFs influence the central nervous system, it has not been fully understood exactly how this occurs. In a recent review on the effectiveness of SMFs at cellular level, SMFs were ensured to have an impact on cellular systems [47]. In particular, radial pair recombination and biomolecules reorientation by In the left-biased group, the magnitude of bias was significantly reduced immediately after tSMS over C6, and the bias was significantly smaller for C6 as compared to P4 at post. * p < 0.001 vs. pre, † p < 0.05 vs. tSMS over P4.

Discussion
In this study, 9 of 16 subjects consistently underestimated the length of the right-side segment in line length judgement task, and the remaining 7 subjects consistently underestimated the length of the left-side segment. In these subjects showing the leftward bias, there was an improvement in the task score after an application of tSMS over C6 but not over P4 or after sham stimulation. In the other subjects showing the rightward bias, no significant intervention effect was observed with any stimulation.

Leftward and Rightward Biases in Healthy Individuals
Neurologically healthy individuals tend to show a leftward bias [39]. This phenomenon is known as pseudoneglect and can be observed in various visuo-spatial tasks [40,41]. In contrast to this previous observation, 7 of 16 subjects showed a rightward bias in the present study. Although at a first glance our finding seems contradictory, enormous individual differences have been observed in a relatively large number of previous studies on pseudoneglect [42][43][44][45]. In particular, Manning and colleagues found a rightward bias in 10 of 22 subjects with a large between-subject variability and proposed that this large variability could be the cause of frequent unsuccessful replication of leftward bias in healthy individuals [46]. Also, Jewell and McCourt pointed out in their review that, in addition to the problem of individual differences, scanning, sex, age, and hand dominance could influence the amplitude of pseudoneglect, even though their meta-analysis demonstrated the existence of leftward bias [41]. Thus, what we observed in the present study is not necessarily unique. Supportively, in our preliminary study, all the subjects consistently showed either a leftward or rightward bias over all sessions regardless of the stimulus presentation time. Furthermore, intersession reproducibility of the bias was quite high. From these data, there are assumed to be a certain number of individuals who show a rightward bias. Moreover, the high consistency and reproducibility suggest that the change in the task score by an application of tSMS over C6 in the left-biased group was not due to changes in attention, habituation, or fatigue resulting from repeated sessions.

Putative Mechanisms Underlying Change in the Score of Line Length Judgement Task by tSMS Over C6
SMFs have constant intensity and direction over time with a frequency of 0 Hz, and are different from electromagnetic fields that vary over time. Although growing evidence indicates that SMFs influence the central nervous system, it has not been fully understood exactly how this occurs. In a recent review on the effectiveness of SMFs at cellular level, SMFs were ensured to have an impact on cellular systems [47]. In particular, radial pair recombination and biomolecules reorientation by diamagnetic anisotropy effects were proposed to be conclusive as such SMFs' effectiveness was consistently reported in all the previous studies: these changes subsequently result in susceptibility of biomolecules, intracellular structural modifications, and changes in the enzymatic reactions [47]. Rosen and colleagues also hypothesized that diamagnetic anisotropy by SMFs can cause a reorientation of phospholipids and hence an alteration of ion channels within them. Indeed, there is evidence that SMFs can influence membrane resting and action potentials by altering voltage-gated potassium [48], sodium [49], and calcium [50] channels. Another possibility is that the gradient of SMFs pushes calcium ions (positive ions) away from astrocytes, reducing the possibility of glutamate release [1]. Additionally, a recent study proposed a new hypothesis that magnetic pressure associated with gradients of the Zeeman influences a surface tension on the structure of the channel proteins, consequently altering the kinetics of voltage channel gating mechanics [51]. At current understanding, even though there are many hypotheses, there is no firm conclusion as to what mechanism is responsible for the change in the central nervous system by tSMS. Nevertheless, tSMS has been consistently demonstrated to have an inhibitory effect on the cortical excitability in various areas, including the motor [1,11], somatosensory [14,52], visual [53,54], and temporal [55] cortices, as well as SMA [16,17] and cerebellar region [56].
In the present study, tSMS applied over C6 (right STG) but not P4 (right PPC) altered the score of line length judgement task. Specifically, in the left-biased (right spatial neglect) group, the magnitude of bias or neglect was reduced immediately after tSMS over C6. Considering the inhibitory effect that tSMS has on the cortical excitability, it was expected that tSMS over the right PPC would reduce the leftward bias as in a previous study demonstrating improved leftward bias in healthy adults by inhibitory rTMS over the right PPC [34]. This contradictory result is hard to be explained, but one possibility is that tSMS was not strong enough to modulate the PPC. It is conceivable that a sufficient magnetic flux density did not reach into the PPC. Meanwhile, we find new evidence that inhibition of the right STG by tSMS can improve a leftward bias in healthy adults, which may be attributed to a relatively new hypothesis that USN is caused by lesions in the ventral attentional network, including the inferior parietal lobule, STG, and inferior frontal gyrus [57,58]. Following a finding that most post-stroke patients with USN had lesions in the ventral attentional network and not in dorsal attentional network consisting of the superior parietal lobule, frontal eye field, and intraparietal sulcus, Corbetta and colleagues hypothesized that lesions in the ventral attentional network bring about dysfunction of the dorsal parietal area, which consequently mediates a rightward bias in post-stroke patients [57,58]. Indeed, there is limited but supportive evidence of possible linkages between ventral and dorsal brain regions [59]. Therefore, we hypothesize that tSMS-induced reduction in the excitability of the STG, a part of the ventral attentional network, caused a change in the function of the dorsal parietal area, as a result inducing a rightward bias or, in other words, improving a leftward bias. Alternatively, tSMS over C6 might have influenced the superior longitudinal fasciculus that anatomically supports the ventral and dorsal attentional networks, as some research argues that lesions in the fasciculus is involved in the occurrence of USN [60]. However, in accordance with Coulomb's law, the magnetic field strength decreases in inverse proportion to the square of the distance, and it is confirmed that the magnetic field strength of about 500 mT at the surface decreases to a low level (under 50 mT) at 5 cm from the magnet surface in actual measurement [13,61] and computer simulation [62][63][64]. Hence, we doubt that tSMS reached into the superior longitudinal fasciculus, a deep white matter tract. Additionally, the intervention effect of tSMS over C6 was transient, and there was no prolonged after-stimulation effect in the present study. As indicated in previous tSMS studies, when tSMS is applied over the motor cortex for less than 20 min, its after-stimulation effect lasts only for a few minutes. On the other hand, when the application time is over 30 min, the after effect can be as long as 30 min [12]. It is necessary to investigate whether cortical plastic change occurs only with long-term stimulation and also whether longer tSMS stimulation of the temporal cortex is necessary for a long-lasting change in visuo-spatial cognition.

Clinical Application
The incident of USN after a stroke in the right hemisphere ranges from 30% to 81% [65][66][67], and this disorder can be chronic (more than 1 year after stroke) in one third of the patients [68]. To alleviate the symptoms of neglect, various therapeutic interventions, such as visuospatial training, prism adaptation, and pharmacologic treatments, have been examined in research studies and also attempted in clinical practice [69]. However, no treatment for neglect has been established thus far. As reported in three recent reviews about the effectiveness of NIBS on USN symptoms [29,31,70], inhibitory rTMS is one of the promising tools for future treatment, but TMS is expensive and requires a quite high-level of technical skill for its use. On the other hand, the NdFeB magnet is an inexpensive industrial product that is easily available, and application of the magnet on the scalp does not require a high operational skill. Therefore, if tSMS applied over right hemisphere (either over C5 or P3) is found to attenuate neglect symptoms in post-stroke patients in future studies, then this new NIBS technique may become a clinically useful tool. Furthermore, it may be used as an in-home intervention along with other rehabilitation exercise programs. Meanwhile, as shown in this and previous tSMS studies, a short after-effect may be a disadvantage compared with a traditional NIBS tools. Since tSMS over the motor cortex for 30 min has been demonstrated to result in an after-effect of 30 min [10], it should be investigated whether a longer duration of tSMS over the temporal cortex similarly induces long-lasting after-effects on visuo-spatial cognitive function. Given the short after-effect, another potential clinical use could be to repeat tSMS in combination with rehabilitation for 4-8 weeks, like interventions using rTMS and tDCS [71,72]. Clinical investigation is clearly warranted to confirm the potential benefit of using tSMS.

Conclusions
The present study demonstrated that application of tSMS over the right STG reduces a leftward bias and thus modulates visuo-spatial cognition in healthy adults. Further study is needed to clarify the neurophysiological mechanism underlying this modulation by tSMS over the temporal but not parietal cortex. Funding: This work was partially supported by grants from the Japan Society for the Promotion of Science (19H01091, 19H03977, and 20K19708).

Conflicts of Interest:
The authors declare no conflict of interest. Figure A1. These figures show results without an outlier from the left-biased group. (A) Serial changes in the individual total scores of line length judgement task before (pre), immediately after (post), and 10 min (post 10) after tSMS for 20 min. Scatter plots present the individual values for each stimulation condition: (a1) P4, (a2) C6 and (a3) sham stimulation. Nine subjects showed a rightward bias (reddish lines), and seven subjects showed a leftward bias (blue lines). (B) Serial changes in the total scores of line length judgement task before (pre), immediately after (post), and 10 min (post 10) after tSMS for 20 min, separately presented for the left-biased group (b1) and right-biased group (b2). In the left-biased group, the magnitude of bias was significantly reduced immediately after tSMS over C6, and the bias was significantly smaller for C6 as compared to P4 at post. * p < 0.001 vs. pre, † p < 0.05 vs. tSMS over P4.