Does TMS of the precentral motor hand knob primarily stimulate the dorsal premotor cortex or the primary motor hand area?

Back in 1985, Barker and colleagues introduced transcranial magnetic stimulation (TMS) as a non-invasive technique to stimulate the human motor cortex [1]. Targeting the cortical motor hand representation, Barker et al. [1] showed that TMS can evoke amotor response in contralateral hand muscles. Their seminal report kickstarted thewidespread clinical and scientific use of TMSwhich continues to increase. However, many questions regarding the biophysical and neurophysiological underpinnings of TMS still remain open. Electrophysiological recordings at the level of the upper spinal cord showed that TMS evokes a series of descending volleys in the fast-conducting corticospinal axons of pyramidal cells originating in the primary motor hand area (M1-HAND) [2]. The latencies of the descending corticospinal volleys indicted that these pyramidal cells were activated indirectly by TMS via a “trans-synaptic”mechanism [2,3]. Yet, it is still unclear which neuronal structures in the precentral gyrus are primarily excited by the TMS pulse and cause trans-synaptic excitation of corticospinal output neurons in M1-HAND. This lack of knowledge spans from the macroanatomical to the cellular level. It remains to be clarified which part of the precentral gyrus, which neuronal cell types and which axonal structures (i.e., axon hillock, bends or terminals) are most readily stimulated by the TMS pulse. To tackle these questions, Aberra and colleagues [4] developed a novel biophysically-based multi-scale modelling framework to simulate the responsiveness of a range of cortical cell types to single-pulse TMS at the single-neuron level. Their multi-scale modelling and simulation frameworkmerged state-of-the-art computations of the TMS-induced E-field withmorphologically realistic models of various types of cortical neurons. The E-field calculations used finite element models of the human head based on magnetic resonance imaging (MRI) data of a healthy individual, considering the geometry of the various tissue compartments [5]. The celltype specific neuronal models incorporated existing knowledge about neural membrane properties and axonal morphology, leveraging published models from the Blue Brain project [6]. The latter enabled significant extensions to previous modelling studies