Amir M. Sodagar (York University)
Brain-implantable microsystems for intra-cortical interfacing to the brain are nowadays of increasing interest. When aiming at high-density neural recording and live streaming of the recorded information through a wireless link, the limited bandwidth of the wireless link becomes a bottleneck. To overcome the challenge, reduction and compression of the data being telemetered off the implant appears to be an effective solution.
This tutorial starts with an overview of brain-implantable microsystems. Then, design requirements and challenges of such devices is briefly reviewed. The tutorial then focuses on the data reduction and compression techniques that enable a brain implant to concurrently telemeter multiple intra-cortically sensed neural signals to the outside world in the real time. In addition to their signal processing side, efficient hardware implementation of such techniques is discussed, which is of crucial importance when it comes to realize them on compact and low-power brain implants.