A low-cost reliable online noise level estimation for accurate spike detection in extracellular recordings

On-line unsupervised spike detection is a fundamental building block for autonomous neural implants. To accurately monitor neurological activity, the spike detectors should be able to operate at a wide range of firing rates. This can only be achieved by a careful estimation of instantaneous noise level in the recording. In this paper, we propose a system for estimating noise level in extracellular recordings, which is reliable across a wide range of firing rates, and thus leads to accurate spike detection.