Dual ‐ site TMS as a tool to probe effective interactions within the motor network: a review

Dual-site transcranial magnetic stimulation (ds-TMS) is well suited to investigate the causal effect of distant brain regions on the primary motor cortex, both at rest and during motor performance and learning. However, given the broad set of stimulation parameters, clarity about which parameters are most effective for identifying particular interactions is lacking. Here, evidence describing inter- and intra-hemispheric interactions during rest and in the context of motor tasks is reviewed. Our aims are threefold: (1) provide a detailed overview of ds-TMS literature regarding inter- and intra-hemispheric connectivity; (2) describe the applicability and contributions of these interactions to motor control, and; (3) discuss the practical implications and future directions. Of the 3659 studies screened, 109 were included and discussed. Overall, there is remarkable variability in the experimental context for assessing ds-TMS interactions, as well as in the use and reporting of stimulation parameters, hindering a quantitative comparison of results across studies. Further studies examining ds-TMS interactions in a systematic manner, and in which all critical parameters are carefully reported, are needed.

CS coil was placed over PMd with the handle pointing anteriorly, inducing an AP directed current flow (Bäumer et al., 2009;Groppa et al., 2012a;Groppa et al., 2012b;Vesia et al., 2018). This coil position caused a slight overlap of both coils in some cases (Vesia et al., 2018). Although the specific coil orientations were not described, both included figures and the description of the induced flow direction suggest that Civardi et al. (2001) and Byblow et al. (2007) used a similar coil orientation.
Alternatively, the TS coil was oriented perpendicular to the mid-sagittal line with the handle pointing either medially (Parmigiani et al., 2015) or laterally (Parmigiani et al., 2018), while the exact orientation of the CS coil was not defined but varied according to the mechanical interaction between the two coils (Parmigiani et al., 2015). Finally, Van Hoornweder et al. (2021) examined an alternative way of positioning both coils which allowed a small inter-coil distance of 35mm without inducing (over-)heating of the coils, using the edge of a butterfly coil and a small cooled figure-of-eight coil.
Here, the small coil was placed over left M1 with the short axis roughly parallel to the central sulcus, similar to, e.g., Groppa et al. (2012b), while the outer surface of one wing of a butterfly figure-of-eight coil was placed over the left PMd (Van Hoornweder et al., 2021).

Interhemispheric interactions
The CS coil, situated over PMv, was positioned tangentially to the skull with the handle pointing anteriorly (handle parallel to the mid-sagittal line) (Buch et al., 2010;Fiori et al., 2017), represented in Fehler! Verweisquelle konnte nicht gefunden werden.. In contrast, the exact coil orientation was not described in the study of Mochizuki et al. (2004b). Smaller coils (50 mm) were used, either to apply both CS and TS (Fiori et al., 2017) or only for the CS (Buch et al., 2010), to allow placement of both coils on the optimal target position.

Intrahemispheric interactions
The TS coil was always positioned with the handle pointing posteriorly and laterally at a 45° angle to the midsagittal line (Bäumer et al., 2009;de Beukelaar et al., 2016;Koch et al., 2010b;Lago et al.,  induce an AP-directed current in the brain but did not specify its exact position.

DLPFC Interhemispheric interactions
Both the CS and the TS coil were oriented perpendicular to the mid-sagittal line (Uehara et al., 2013).
Alternatively, either only the CS coil was oriented perpendicular while the handle of the TS coil was

Intrahemispheric interactions
Although none of the studies discussed in this section described the precise tilt and twist of the conditioning coil, they appear to differ from one another as based on the described coil orientations.
More specifically, the CS coil was either placed anterior and slightly lateral to the TS coil placed over

SMA Interhemispheric interactions
The CS coil was held with the handle pointing laterally, inducing a medially directed current in the underlying brain tissue (Picazio et al., 2014), or with the handle pointing forward, inducing an APdirected current (Fiori et al., 2017).

Intrahemispheric interactions
Regarding the coil orientation, most authors rotated the TS coil 45° to the midsagittal line, with its handle pointing posterior and lateral inducing a PA current in the brain (Arai et al., 2012;Mars et al., 2009;Shirota et al., 2012). Alternatively, the TS coil was rotated 180°, inducing an AP-directed current in the brain (Arai et al., 2012) or oriented with its handle pointing anterior, 45° relative to the midsagittal line, using a circular coil (Oliveri et al., 2003).

Cerebellum
As for the optimal orientation of the coil, while the TS coil was directed posterolateral at 45° away from the mid-sagittal line (inducing a PA current in the brain), most authors have oriented the CS coil horizontally, inducing an upward current in the cerebellum [e.g., (Daskalakis et al., 2004;Fernandez et al., 2018b;Schlerf et al., 2012;Spampinato et al., 2020b;Tanaka et al., 2018)], except for Ugawa et al. (1995), Fisher et al. (2009) and Werhahn et al. (1996 placing a circular coil (TS coil) over the vertex.
Additionally, Werhahn et al. (1996) used a vertically oriented CS coil, inducing a mediolateral current. Furthermore, Fisher et al. (2009) rotated the CS coil 180 degrees in some participants reversing the CS current direction (i.e., upward to downward). Not all studies described the specific coil orientation of either the CS coil (Spampinato et al., 2017) or both the TS and CS coil (Hardwick et al., 2014;Spampinato and Celnik, 2017).
Overview illustrating how different brain regions influence M1 in the context of different motor tasks. A green cross indicates a facilitatory influence, while a red cross indicates an inhibitory influence. Lack of modulation of an interaction during a certain task is indicated with a black cross. Lastly, a grey cross indicates that the interactions are modulated but the direction (i.e., facilitation and/or inhibition) is not yet clear (i.e., less than 50% of the studies demonstrated consistent results). It should be kept in mind that it is very difficult to reflect differences in timing or other nuances in this type of table. The table should therefore be viewed as a general overview only. For a more detailed explanation we refer the reader to the respective "task-related" sections in the results section.