Implantable medical devices and brain attractors: Network modulation and design practice

Implantable medical devices are increasingly employed for neurorehabilitation to restore lost motor function through brain computer interfacing. Proper circuit design for both sensing and stimulation is key to technology success, including the optimization of device capabilities and the exploitation of neural coding properties, while enhancing patient safety. To date peripheral nerve and brain implant devices adopt similar design strategies for neural interfacing that rely on Hodgkin Huxley formulations of spiking properties. Dynamical systems theory is revealing, however, that brain operation fundamentally differs from that of peripheral nerves, and is characterized by continuous operation, oscillatory networking properties, and the optimization of set points that enable both stability and maximum flexibility of response. Dynamical elements like attractors are nonlinearly and, in cases of sustained activity, functionally independent of spiking activity relative to output, adopting numerous configurations described by widely varying mathematical relations between spiking activity and performance. These features introduce new physiological considerations for brain implant design strategies.