Compressed Sensing Based Neural Signal Processing and Performance Analysis

Abstract:

Measurement of neural signal provides important value for study of brain function and the pathogenesis of neurological. With emerging extensive research of electrical activity, more and more neural signal need to be collected, transmitted and stored, making the compression processing of neural signal become important part of digital signal processing. In recent years, ASIC-based wireless neural signal acquisition system has been developed rapidly, encountered strict restrictions on power consumption which is dominant determined by the data rate and complexity of algorithm. In order to reduce power consumption, lower data rate and algorithm with lower complexity needed to be selected when design a neural acquisition system. This paper focus on neural signal compression method based on compressed sensing and its performance and compare it with conventional compression algorithm. We compare complexity of various algorithms in the view of circuit complement, show that the complexity of neural signal compression can be dramatically reduced by using specially designed compressed sensing matrix, thereby reducing the system power consumption.

• Periodical: Applied Mechanics and Materials (Volumes 513-517)
• Pages: 1595-1599
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.513-517.1595
• Online since: February 2014
• Authors: Yu Xi Zhang*, Wen Gui Fan, Jin Ping Sun
• Keywords: Algorithm (GA), Compression, Neural Signal

References

[1] Ding Fei. Neurobiology. Beijing: Beijing Science Press, 2007. 386-465. (in Chinese).

[2] M. S. Chae, Zhi Yang, M. R. Yuce, et. al. A 128-Channel 6 mW Wireless Neural Recording IC With Spike Feature Extraction and UWB Transmitter [J]. Neural System and Rehabilitation Engineering, IEEE Trans, vol. 17, no. 4, 2009, p.312 – 321.

DOI: 10.1109/tnsre.2009.2021607

[3] Christoph Buulach, Ulrich Bihr. Evaluation Study of Compressed Sensing for Neural Spike Recording[C]. 34th Annual International Conference of the IEEE EMBS, 2012, p. pp.3507-3511.

[4] Z. Charbiwala, V. Karkare, S. Gibson, et. Compressive Sensing of Neural Action Potentials Using a Learned Union of Supports [C], Body Sensor Networks, 2011, p. p.53–58.

DOI: 10.1109/bsn.2011.28

[5] Aharon M, Elad M, Bruckstein A. M. The K_SVD: An algorithm for designing of overcomplete dictionaries for sparse representation [J]. IEEE Transaction on Signal Processing, 2006, V54(11): 4311-4322.

DOI: 10.1109/tsp.2006.881199

[6] R.Q. Quiroga,Z. Nadasdy, and Y. Ben-Shaul. Unsupervised spike detection and sorting with wavelets and superparamagnetic clustering[J]. Neural Comput, 2004, vol. 16. pp.1661-1687.

DOI: 10.1162/089976604774201631

[7] BaraniukR.G. Alecture on compressive sampling [J]. IEEE Signal Processing Magazine, 2007, V24(4): 118-121.

[8] Donoho D.L. Compressed sensing [J]. IEEE Transactions on Information Theory, 2006, V52(4): 1289 -1306.

[9] Mallat S., Zhang Z. Matching pursuits with time-frequency dictionary[J]. IEEE Transactions on signal Processing, 1993, V41(12) : 3397-3415.

DOI: 10.1109/78.258082