Studies of electrocortical recordings from a newly conceived chronically­implanted wireless device in monkeys

Introduction: Artificial Brain­Machine Interfaces (BMI) represent a prospective step forward the vicarial support or replacement of faulty brain functions. The width and intricacy of replacement of a "brain function" is significantly related to the extent and complexity that the function spans in the neural context. Another issue is the "ecological niche" the interface may best occupy in the surviving neural context and its coherence with the residue functions. Namely, the device must neither interfere nor disorganize the already existing information background in the plan of potential clinical requirement as rehabilitation, adjutancy or functional replacement. A prerequisite from a chronically implantable device is also represented by the ease of reciprocal conveyance of fine­grain information from and to the brain. Not least, a BMI must meet the requisite of long term compliance within a delicate context such as the nervous tissue, without provoking rejection responses. Methods and Results: We present here the electrophysiological results obtained from a non­human primate (Macaca fascicularis) chronically implanted with a novel implantable BMI platform, called Cyberbrain, a 16 channel totally wireless grid, rechargeable by induction [AB Medica (Milan, Italy), designed by one of the authors, PR]. The grid provides both the wireless transmission of epicortical recordings and, equally, the delivery of finely driven stimulations. The grid was implanted (by PR) over the sensorimotor cortex (13 electrodes over the primary motor cortex, 3 on the primary somatosensory cortex) in the deeply anaesthetized animal. Cortical sensory and motor recordings and stimulations have been performed during 6 months. In details, by motor cortex epicortical single spot stimulations (1 to 8V, 1 to 10 Hz, 500us, biphasic waves) we analyzed the motor topographic precision, evidenced by tunable finger movements of the anesthetized animal. The responses to light mechanical peripheral sensory stimuli (50 stimuli, 1ms, variable delays 1.5 to 4 s) were analyzed, both in ongoing spontaneous activity and after activations of each epicortical sensory lead. In the first, we investigated, by estimating the mutual information between the single lead activities, the detection dynamics of the responses to peripheral stimuli within a sensory cortical circuitry, in the second we evaluated the grid electrical interference with somatotopic natural stimuli sensory detection programs. Conclusions: These features provide important technical suggestion for long­term implanted BMI and help for future therapeutic applications in sensorimotor and neurodegenerative diseases.