Compressive multichannel cortical signal recording

Abstract

This paper presents a novel approach to acquire multichannel wireless intracranial neural data based on a compressive sensing scheme. The designed circuits are extremely compact and low-power which confirms the relevance of the proposed approach for multichannel high-density neural interfaces. The proposed compression model enables the acquisition system to record from a large number of channels by reducing the transmission power per channel. Our main contributions are the twofold. First, a CMOS compressive sensing system to realize multichannel intracranial neural recording is described. Second, we explain a joint sparse decoding algorithm to recover the multichannel neural data. The idea has been implemented at system as well as circuit levels. The simulation results reveal that the multichannel intracranial neural data can be acquired by compression ratios as high as four.

A Kronecker Compressed Sensing formulation for energy efficient EEG sensing

“…At the receiver, the signal is assumed to be sparse in some domain W (usually Gabor [1,2], wavelet [3] or DCT [4]); the sparsifying transform is chosen such that it is orthogonal. Using orthogonality, it is possible to express (13) as follows:…”

“…The analysis-synthesis pair only holds for orthogonal transforms and tight-frames. In general the Gabor [1,2] or spline [3] basis is not orthogonal or tight-frame. Thus they do not follow the nice analysissynthesis form.…”

“…Previous CS based techniques [1][2][3][4] have been able to reduce the energy consumption required for communication. It samples the full signal (thus does not reduce the sensing cost); projects the sampled signal onto a lower dimension thereby compressing it; and finally transmits the compressed signal -thereby reducing the communication cost.…”

“…The previous CS based methods [1][2][3][4] formulated the problem in such a way that it was not theoretically possible to reduce the sensing energy -it required sampling the full signal. In this work, we propose to reduce the sensing energy, simply by randomly under-sampling the EEG signal.…”

“…In such a sensing scenario, it is not possible to use traditional sparsifying transforms like wavelet [3] or DCT [4] owing to the incoherency requirement of CS [5]. The aforesaid transforms are not incoherent with the Dirac sampling basis (under-sampling).…”