Hardware powered ultra-low latency (HarPULL) brain-body state dependent TMS technology for investigation of the motor system

Abstract

Сortico-spinal neuron excitability can be tracked using rhythmic oscillations whose phase plays a pivotal role in assessing the state of the pertinent neural network. Brain’s oscillatory activity can be non-invasively registered with electroencephalography (EEG) and magnetoencephalography (MEG). Using transcranial magnetic stimulation (TMS) contingent upon the parameters of the ongoing brain activity was shown to have numerous applications in both research and clinics. Despite the potential, the described closed-loop studies often deliver inconclusive results due to real-time phase tracking errors and unaccounted for movement related contextual factors.We present a novel truly real-time hardware-software complex for the low latency brain and body state dependent transcranial magnetic stimulation (TMS). The real-time part of the software is implemented on-board of a digital EEG-recording device controlled by a real-time operating system. High fidelity phase estimation is achieved by Kalman filter-based state-space modeling of brain rhythm using the parameters estimated from the pre-recorded segment of data. To get a better grip on the state of the specific muscles and to add the associated contingency to the TMS pulse triggering process we employ electromyographic (EMG) data channels and track their envelope. The trigger moment is contingent upon a set of user specified conditions accounting for brain rhythm phase, amplitude and EMG signal strength. The circular standard deviation of the phase tracking error is below 6 degrees. The delay from the user specified combination to the arrival of the magnetic field pulse at the recording electrode is below 5 milliseconds.Using this technology, for the first time, we performed adaptive motor mapping of a healthy subject in the agonist-antagonist ERD/ERS motor task with the possibility of triggering stimulation at the certain phase of movement according to the EMG activity ratio of several hand muscles, which opens broad opportunity for studying lateral inhibition in the motor cortex.