Assessing the effects of TMS coil size on quantitative mapping of human motor cortex

51 

Abstract Submission 

Evgenii Kim^1,2, Mohammad Daneshzand^1,2, Sergey Makarov³, Konstantin Weise^4,5, Ole Numssen⁴, Thomas Knösche⁴, Dylan Edwards^6,7, Tommy Raij^1,2, Aapo Nummenmaa^1,2

¹Massachusetts General Hospital, Boston, MA, ²Harvard Medical School, Boston, MA, ³Worcester Polytechnic Institute, Boston, MA, ⁴Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Saxony, ⁵Leipzig University of Applied Sciences, Leipzig, Germany, ⁶Thomas Jefferson University, Philadelphia, PA, ⁷Moss Rehabilitation Research Institute, Philadelphia, PA

Evgenii Kim  
Massachusetts General Hospital|Harvard Medical School
Boston, MA|Boston, MA

Mohammad Daneshzand  
Massachusetts General Hospital|Harvard Medical School
Boston, MA|Boston, MA

Sergey Makarov  
Worcester Polytechnic Institute
Boston, MA

Konstantin Weise  
Max Planck Institute for Human Cognitive and Brain Sciences|Leipzig University of Applied Sciences
Leipzig, Saxony|Leipzig, Germany

Ole Numssen  
Max Planck Institute for Human Cognitive and Brain Sciences
Leipzig, Saxony

Thomas Knösche  
Max Planck Institute for Human Cognitive and Brain Sciences
Leipzig, Saxony

Dylan Edwards  
Thomas Jefferson University|Moss Rehabilitation Research Institute
Philadelphia, PA|Philadelphia, PA

Tommy Raij  
Massachusetts General Hospital|Harvard Medical School
Boston, MA|Boston, MA

Aapo Nummenmaa  
Massachusetts General Hospital|Harvard Medical School
Boston, MA|Boston, MA

Transcranial Magnetic Stimulation (TMS) is a powerful non-invasive neurostimulation method that is established in preoperative functional mapping for neurosurgical interventions[1]. The geometry of the coil plays a crucial role in shaping the induced electric field (E-field), thus influencing the spatial specificity and efficacy of TMS pulses (larger coils provide enhanced depth penetration but with a trade-off of decreased focality)[2]. The spatial extent of the E-field determined by the coil type has been a crucial consideration for improved quantitative TMS protocols, where the goal is to isolate a cortical "hotspot" for a target muscle[3]. Yet, it is not well-characterized how invariant the mapping results are with respect to the size of the TMS coil, and/or if using different coils would provide complementary information.

Data were acquired from a single healthy, right-handed male subject who provided written informed consent. The workflow is illustrated in Fig. 1. Before TMS session, T1- and T2-weighted MR images were obtained using a Siemens 3 Tesla scanner. The head model was created from MRI using the SimNIBS headreco[4]. TMS pulses were delivered using a MagPro X100 stimulator (MagVenture, Denmark). Two figure-of-eight coils, B60 and B35 (MagVenture; with coil diameter 2x75 mm and 2x46 mm, respectively) were used in separate runs during the same session. Each run involved 250 TMS single biphasic pulses at 5-second intervals. The stimulation intensity for B60 was set at 150% of the resting motor threshold (rMT), which corresponds to 68% of the maximum stimulator output (MSO). Meanwhile, B35 was operated at 133% of the rMT (limited by 100% MSO). For each stimulation, each coil was randomly positioned around the motor cortex. Coil locations were recorded using a neuronavigation system (TMS Navigator, Localite, Germany). Motor evoked potentials (MEPs) were recorded from three finger muscles - first dorsal interosseous (FDI), abductor digiti minimi (ADM), and abductor pollicis brevis (APB). MEPs were sampled at 5 kHz and post-processed with a bandpass filter (100-1000 Hz). Coil positions and the head model were utilized to compute the E-fields using an LU-based BEM-FMM[5]. The identification of the "hotspot" for each muscle was determined by maximum R² scores derived from logarithmic sigmoid regression, establishing the correspondence between induced magnitude E-field distribution and MEPs[6].

A goodness-of-fit map illustrating the functional cortical representation of the fingers under two coils is presented in Fig. 2. The functional map was further derived by combining B60 and B35 datasets. To ensure an equivalent number of data points, 125 random trials were selected from the B60 coil and 125 from the B35 coil. The identified "hotspots" were consistently within a 3 mm range across three coil configurations (B60, B35, combined) for each muscle. As expected, the B35 coil with higher focality of E-field exhibited an improved localization in the goodness-of-fit map. Specifically, the B60 mapping showed a "spread" of the hot spot neighboring gyri (Fig. 2, red arrows) that is likely associated with the limitations of the TMS coil focality, rather than the actual extent of the cortical motor representation. The results from the combination of two coil types produced a functional map that captures the "best of both worlds" – the spatial spread to the neighboring gyri of the B60 coil is reduced, while the R₂ scores of the B35 are enhanced.

Our study showed that TMS mapping results across two coils of different sizes produced qualitatively similar outcomes, indicating the overall robustness of the approach. Combining both coil types produced a balanced mixture of both features in terms of sensitivity and specificity. In the future, by utilizing multichannel TMS coil arrays[7], we may generate a wider range of E-field patterns and further reduce the ambiguity in the localization maps.

Non-invasive Magnetic/TMS ¹

TMS ²

Cortex

Transcranial Magnetic Stimulation (TMS)

Other - Functional mapping