Complexity of neural outputs elicited by transcranial magnetic stimulation

Complex neural activity of the motor cortex is posited to serve as the foundation for a large repertoire of activation patterns crucial for executing movements. As transcranial magnetic stimulation (TMS) predominantly activates monosynaptic fast-conducting corticospinal projections, which are involved in dexterous movement control, the complexity of neural outputs elicited by TMS may reflect an underlying repertoire of activation patterns crucial for executing dexterous movements. We proposed to quantify dimensionality of multimuscle motor-evoked potentials (MEPs) through dimensionality reduction as an integrated measure to reflect the complexity of neural outputs elicited by TMS. For its validation, we conducted two experiments focusing on different stimulus parameters: stimulus intensity (experiment 1: n = 40) and size of motor mapping (experiment 2: n = 35). The present findings demonstrated that lower intensities resulted in higher complexity and vice versa but no effects of different sizes of motor mapping on complexity. Analyses incorporating disparities in MEP amplitude across different muscles supported that complexity was effectively captured through dimensionality reduction. Notably, complexity was minimally influenced by the selection of muscles and was associated with individual differences in motor cortex anatomy. We performed two fingers alternating tapping as an index of dexterous movements, and the results demonstrated its association with the complexity of neural outputs elicited by TMS: higher complexity corresponded to better dexterous performance. The proposed complexity measure might reflect neural processes aligned with the principle of motor abundance. The framework complements well-established MEP analyses and offers a novel perspective for investigating the motor system.NEW & NOTEWORTHY The study introduces a framework to assess the complexity of neural outputs elicited by transcranial magnetic stimulation (TMS). We demonstrated that complexity, obtained from multimuscle motor-evoked potentials (MEPs) through dimensionality reduction, varied depending on stimulus intensities, was minimally influenced by the selection of muscles, and was associated with individual differences in motor cortex anatomy. Importantly, individuals with higher complexity exhibited better dexterous performance. The framework offers a novel perspective for investigating the motor system.