CNS*2026 Halifax

Introduction
Transcranial magnetic stimulation (TMS) is a promising technique to alleviate symptoms related to cerebellar pathologies. However, the outcomes of TMS are variable and it is not understood how exactly neuronal activity is affected by it. Many models, at different scales, have been developed to try to address the open questions on the functioning of TMS. The state of the art are multi-scale models, which integrate MRI generated head models of the electric field induced by TMS and biologically realistic neuron models. A limitation of multi-scale models lies in their validation [1]. In this work we present a framework to model and validate TMS targeting the cerebellum.


Methods

We calculated the mean electric field induced by TMS in the cerebellar vermis using the SimNIBS simulator. We developed a biologically realistic model of the cerebellar cortex [2] and coupled the TMS induced electric field to its neurons using NEURON’s extracellular mechanism.
We generated the input-output curve of the cerebellar network, in baseline and in stimulated conditions, by increasing the input mossy fibre frequency from 0 to 150 Hz in 3 Hz steps. 
We normalised the input-output curves in the two conditions and utilised them in a vertical oculomotor system model [3] to predict eye movements. We compared predicted eye movements in the baseline and stimulated conditions with experimental data to validate the TMS model.



Results

The maximum mean electric field induced by TMS in the cerebellum vermis is 30.4 V/m. The stimulation does not alter the mean firing rate of the neurons, but it alters the spike timing and the instantaneous firing rate. 
We reproduced the eye movements of a healthy subject with the network in baseline conditions. Eye movements in the stimulated condition show a drift during gaze holding. Saccades are not affected by the stimulation.
Experimentally it was found that the application of TMS generated an increase in the amplitude of ipsilateral reflexive saccades and in the acceleration of ipsilateral smooth pursuit movements.



Discussion

The vertical oculomotor system model produces observable differences in predicted eye movements following stimulation. This suggests that the model can serve as a validation framework for transcranial magnetic stimulation models targeting cerebellar circuits involved in saccades and gaze holding.
Predicted eye movements do not accurately reproduce the experimental findings. This may suggest that the employed model is not detailed enough to reproduce experimental results. It may also suggest eye movement changes seen experimentally  are not primarily driven by neuronal activity within the targeted cerebellar region, but may also arise from activation of more superficial neural structures or neighbouring regions in the cerebral cortex.



References

1. Shahid, S. S., Bikson, M., Salman, H., Wen, P., & Ahfock, T. (2014). The value and cost of complexity in predictive modelling: Role of tissue anisotropic conductivity and fibre tracts in neuromodulation. Journal of Neural Engineering, 11(3), 036002. https://doi.org/10.1088/1741-2560/11/3/036002 
   
2. Bernasconi, E., Yousif, N., & Steuber, V. (2024). 32nd Annual Computational Neuroscience Meeting: CNS*2023 - P194 Development of a biologically realistic cerebellar model to study the effects of non-invasive stimulation. Journal of Computational Neuroscience, 52(1), 3–166. https://doi.org/10.1007/s10827-024-00871-5 
   
3. Glasauer, S., & Rössert, C. (2008). Modelling drug modulation of nystagmus. Progress in Brain Research, 171, 527–534. https://doi.org/10.1016/S0079-6123(08)00675-4
   

Acknowledgement
-