24-Channel Dual-Band Wireless Neural Recorder With Activity-Dependent Power Consumption

Srinjoy Mitra, Jan Putzeys, Francesco Battaglia, Carolina Mora Lopez, Marleen Welkenhuysen, Cyriel Pennartz, Chris van Hoof, Refet Firat Yazicioglu

Research output: Chapter in Book/Report/Conference proceedingConference contribution


Recording neural activity from freely behaving animals has become extremely important for basic research in neuroscience. One of the major limitations in this domain arises from the need of a low-power, lightweight, wireless recording unit that can be used on small animals for at least a couple of hours. To conserve power, it is essential to reduce the large data volume from tens or even hundreds of recording sites. Since it is widely considered that the primary information in neural recording is limited to the infrequent action potentials (AP or spike), the data can be considerably reduced by using a system based on AP activity [1]. This paper presents a complete activity-dependent wireless system that utilizes an ASIC for recording simultaneous AP and local field potential (LFP) signals, improving the state of the art in terms of performance-to-weight ratio. In spite of large data compression, we demonstrate a method to accurately preserve the AP shape, essential for further processing and spike sorting. Finally, this is the first system (see Fig. 16.4.6 comparison table) particularly designed to take advantage of tetrode (four closely placed electrodes) based neural recording units (Fig. 16.4.1).
Original languageEnglish
Title of host publication2013 IEEE International Solid-State Circuits Conference Digest of Technical Papers (ISSCC)
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Number of pages2
ISBN (Electronic)978-1-4673-4516-3
ISBN (Print)978-1-4673-4515-6
Publication statusPublished - 28 Mar 2013


Dive into the research topics of '24-Channel Dual-Band Wireless Neural Recorder With Activity-Dependent Power Consumption'. Together they form a unique fingerprint.

Cite this