Real-time Computation of E-field in Transcranial Magnetic Stimulation for Neuronavigation and Optimization

Transcranial magnetic stimulation (TMS) is a brain stimulation technique used to modulate the brain activity facilitating the treatment of many psychological and morphological diseases. During TMS, it is important that we optimize the E-field dosage at the targeted brain region for efficient stimulation. A real-time TMS solver is particularly useful in this regard as it allows coil placement over the scalp on-the-fly. We introduce a method of predicting the TMS-induced E-field in under 4 milliseconds for arbitrary coil placement over the scalp. First we compute a set of orthonormal mode functions for the TMS-induced E-field using random realizations of gaussian white noise magnetic currents defined on a fictitious surface separating the head and the coil. Next, we use electromagnetic reciprocity and Huygens’s principle to compute the mode expansion coefficients. A study on 16 MRI-derived real head models reveals that the method requires only 400 modes to achieve an error of less than 2% irrespective of coil model. This enables the method suitable for real-time neuronavigation system.