Wireless Communication Systems from the Perspective of Implantable Sensor Networks for Neural Signal Monitoring

Abstract

Recent advances in modern neurocomputing heading toward promising clinical applications of implantable neuronal sensing devices have shown the utmost necessity of wireless communication systems that allow real-time monitoring of neural signals. The design of a wireless transmission system for this particular application shall meet several requirements involving source compression of the high data rate neural recording, communication with a standard device as bridge between body area and remote server, and high fidelity of the received signal to ensure effective brain activity monitoring. A wireless transmission system over Bluetooth and 3G is analyzed for its application to the real-time transmission of neural signals captured by implanted micro-electrode array sensors. Average compression rate of 75% of the neural signal is achieved through detection using nonlinear energy operator preprocessing and automatic threshold adaptation. The wireless transmission of these signals integrates a Bluetooth transmission from the information source to a conventional mobile device and then over 3G to a remote server, without intermediate storage on the mobile phone. Reconstruction of the coded neural signal provides the input to high-performance spike classification algorithm allowing the tracking of individual neuron spiking patterns.