Fichier PDF

Partagez, hébergez et archivez facilement vos documents au format PDF

Partager un fichier Mes fichiers Boite à outils PDF Recherche Aide Contact



charvet ieee2008 256 .pdf



Nom original: charvet_ieee2008_256.pdf
Titre: BioMEA: A 256-Channel MEA System with Integrated Electronics
Auteur: Guillaume Charvet, Olivier Billoint, Lionel Rousseau, Blaise Yvert

Ce document au format PDF 1.4 a été généré par The Engineering in Medicine and Biology Conference Management System / PDFlib+PDI 6.0.2 (Perl 5.8.0/Linux), et a été envoyé sur fichier-pdf.fr le 20/07/2018 à 14:43, depuis l'adresse IP 132.168.x.x. La présente page de téléchargement du fichier a été vue 86 fois.
Taille du document: 1.7 Mo (4 pages).
Confidentialité: fichier public




Télécharger le fichier (PDF)









Aperçu du document


Proceedings of the 29th Annual International
Conference of the IEEE EMBS
Cité Internationale, Lyon, France
August 23-26, 2007.

ThB08.2

BioMEATM : A 256-channel MEA system with integrated electronics
G. Charvet, O. Billoint, L. Rousseau and B. Yvert

Abstract— In order to understand the dynamics of large
neural networks, where information is widely distributed
over thousands of cells, one of today’s challenges is to
successfully record the simultaneous activities of as many
neurons as possible. This is made possible by using
microelectrodes arrays (MEAs) positioned in contact
with the neural tissue. Thanks to microelectronics’
microfabrication technologies, it now becomes possible to
build high density MEAs containing several hundreds of
microelectrodes. However, increasing the number of
electrodes using conventional electronics is difficult to
achieve. Moreover, high density devices addressing all
channels independently for simultaneous recording and
stimulation are not readily available. Here, we present a
256-channel in vitro MEA system with integrated
electronics allowing simultaneous recording and
stimulation of neural networks. Both actions are
performed independently on all channels.

networks. These 256-electrode arrays are interconnected to
four 64-channel ASICs dedicated to the amplification and
the multiplexing of the signals and stimulation. Each ASIC
includes one amplification stage and one current generator
per channel, and has the capability to rapidly switch between
recording and stimulation. The realization of high density
MEA systems with integrated electronics offer new
possibilities for both in vitro and in vivo studies of large
neural networks.
II. ORGANIZATION OF THE BIOMEA SYSTEM
The BioMEA™ system comprises a high density
electrode array (256 3D micro-electrodes) interconnected to
four dedicated 64-channel ASICs (amplification, analog
multiplexing, current stimulation) running in parallel,
specific acquisition boards, and a user-friendly software.

I. INTRODUCTION

M

ultielectrode arrays (MEAs) provide an elegant way to
probe the neural code distributed over large
populations of neurons either in vitro [1][2][3] or in
vivo[4][5]. MEAs also offer the possibility to deliver
electrical stimulation to neural networks[6], making them
promising technologies to build neural prosthesis[7][8][9].
The increasing number of MEA studies show that the
understanding of neural network dynamics can be best
achieved when the highest number of neurons can be
recorded simultaneously. However, increasing the number of
electrodes using conventional electronics is difficult to
implement into compact device. Moreover, high density
devices addressing all channels independently for
simultaneous recording and stimulation are not yet available.
To overcome these limitations, we designed high density
multielectrode systems with integrated electronics allowing
simultaneous recording and stimulation of neural networks.
This paper presents a new 256-channel system named
BioMEA™. Either planar or 3D electrodes arrays were
realized by deep reactive ion etching techniques of a Silicon
substrate and tested on whole embryonic spinal cord
Manuscript received April 2, 2007.
G. Charvet is with the CEA – LETI - MINATEC Department DTBS,
Grenoble, 38054 France (e-mail: guillaume.charvet@cea.fr).
O. Billoint is with the CEA – LETI - MINATEC Department DCIS,
Grenoble, 38054 France (e-mail: olivier.billoint@cea.fr).
B. Yvert is with the CNIC UMR5228 – CNRS and Université
Bordeaux1&2, Talence, France (e-mail: b.yvert@cnic.u-bordeaux1.fr)
L. Rousseau is with the Groupe ESIEE, Noisy-le-Grand, France (e-mail:
l.rousseau@esiee.fr)

1-4244-0788-5/07/$20.00 ©2007 IEEE

Fig1 : Architecture of the system BioMEATM
III. 256-ELECTRODE ARRAYS
In order to be as close to the neurons as possible, the
microelectrodes of in vitro systems may be shaped as 3D
needles, at the tip of which the recording site is.
Pioneering techniques use classic isotropic etching,
(either plasma or wet etching). However, the technique’s
drawback is the limitation impose on the electrode’s pitch
by the electrode’s. The electrode’s pitch cannot be
smaller than twice the electrode’s height: for h = 80 µm,
p > 160 µm. These isotropic processes thus prevent the
fabrication of dense 3D electrode arrays with large aspect
ratios.

171

These limitations have been overcome using Deep Reactive
Ion Etching (DRIE) [10] [11]. A specific process was
developed in which silicon substrate is etched anisotropically
in order to obtain microneedles. Our MEAs have no
limitations neither in height nor diameter and pitches down
to 50 µm could be achieved. This process offers the
possibility to manufacture various shapes of electrodes on
silicon or glass substrate. Realized electrode arrays (Fig. 2)
are composed of 256 electrodes with a basis of 40µm, a
height of 80 µm arranged in an 8 x 32 matrix with a pitch of
250 µm.

Each channel of this 64-channel CMOS chip (Fig. 4) is
interfaced with neurons via a microelectrode array and
includes a low noise, variable gain measurement channel.
The preamplifier and the amplifier are based on the same
structure which provides a unity DC gain and an AC gain of
respectively 75 and 10 in the 1Hz – 3kHz bandwidth. Both
can be separately switched to follower configuration.
Stimulations are performed by using, for each channel, an 8to-1 analog multiplexer (fed by 8 external input signals for
the whole ASIC) and a voltage-to-current converter allowing
uniform current stimulation (+/- 400µA peak max.)
independently of the electrode impedance. The measured
transfer function of the voltage-to-current converters shows a
good linearity and a limited drift between channel 1 and
channel 64. A global control signal was added to cut the
residual DC current at the output of the 64 voltage-to-current
converters once the stimulation patterns have been applied.

Fig 2 : 256 Electrode array
IV. THE 256-CHANNEL SYSTEM
Interfacing neurons through MEAs using discrete electronics
rapidly limits the number of channels, creating the need for
highly integrated electronics to achieve sufficient spatial
resolution[12][13][14]. This will be especially the case for in
vivo studies in small animals (mouse).
A. Integrated Electronic
A dedicated ASIC named AGNES (Asic for General
Neurons Electrical Study) (fig 3) has been developed in
order to allow simultaneous recording and stimulation on 64
channels.

Fig 4 : Architecture of the ASIC AGNES
A Sample & Hold circuit allows snapshot style images of the
64 channels (with no time delay between channels as in a
sequential reading), and can stand a maximum sampling
frequency of 50kHz, which leads to a data output frequency
of 3.2MHz (time-multiplexed). To get rid of the random DC
offset potential existing at the electrode-electrolyte interface,
the ASIC can be supplied with floating VSS and VDD. The
Circuit’s size is 2.4mm x 11.2mm (0.35µm CMOS process)
(Table I).
TABLE I
PERFORMANCE OF THE ASIC AGNES
Parameter
Power consumption
Input-Referred Noise
Signal Bandwidth
Maximum Sampling Frequency
Maximum Multiplexing Frequency
Input Impedance
Variable Gain
Maximum Stimulation Current
Area (in 0.35µm CMOS)

Fig 3: ASIC AGNES

172

Result
125mW
4.3 µV rms
0.08Hz to 3kHz
50 kHz
3.2 Mhz
> 1012 ohms
1, 10, 75, 750
+/- 400µA
27mm²

B. Electronic System set-up
The BioMEA™ system interfaces on 256 Electrodes and
performs both measurements and stimulations. This system
uses four 64-Channels ASICs mounted on a mechanical
support allowing electrical interconnections between four
ASICs and one 256-channel array, recording and stimulation
electronics capabilities, and a PC running a specific firmware
and providing a user interface (fig 5).

eight 64kB RAM blocks. A clock signal synchronizes the
digital to analog conversion at a maximum sampling rate of
10MHz. The system also provides an adjustable power
supply for all ASICs. Indeed, the ASIC can be supplied with
floating [VSS,VDD]: from [0v,5v] to [-2.5v,2.5v]. The large
data amount and the low latency require high performance
communications and signal processing capabilities. To
handle the large amount of output data, an FPGA running at
48 MHz and a USB 2.0 interface chip were used.
2) Acquisition software and user interface
The system is supervised by a PC for data acquisition and it
provides a user interface. The software is an essential
component of the system. It manages the USB data handling,
the data visualization, and the stimulus generation. Low
latency and high data throughput are key requirements.
The software allows real time acquisition of 256 channels
and data storage on all 256 channels at a time. The output
data files are CED spike2 data files (SON32 data format) so
that the acquired data can be read back offline using Spike2
software. Channels are displayed during acquisition and it is
possible to navigate through data during recording. A user
friendly interface allows the configuration of the acquisition
and of the stimulation parameters.
V. BIOLOGICAL TESTS
Developing neural networks generate spontaneous activity
[15][16] that is important for the maturation of a functional
circuit. We have been using multielectrode arrays to record
from a whole embryonic mouse hindbrain-spinal cord
preparations isolated in vitro at embryonic days E13-E16.5.
We found spontaneous activity in the medulla characterized

Fig 5 : BioMEA™ system setup
1) Recording and stimulation electronics controls
The recording and stimulation electronics controls includes
analog signal adaptation, analog-to-digital converters (ADC),
digital-to-analog converters (DAC), and also a digital
interface (microcontroller and FPGA) for ASIC protocol and
data transmission. This hardware electronic system was
designed to control up to 4 AGNES ASICs. There are four
14-bit ADCs for simultaneous conversions on the four
ASICs’ analog sampled outputs. The maximum sampling
frequency for each channel is 50 kHz/Channel. So, each
ADC are running at 3,2 MHz (64 channels per one analog
sampled signal). For Stimulation, eight 14-bit DACs for
simultaneous voltage-controlled patterns generation are
implemented. Eight voltage-controlled patterns are stored in

173

Fig 6 : Recording with BioMEATM of spontaneous bulbocervical Local Field Potentials (LFP)

principally by local field potentials (LFP) recurring every 13 minutes (Fig. 6). This activity, which resembles sharp
waves found in the cortex, could be suppressed by a
pharmacological blockade of AMPA/Kainate glutamatergic
receptors using CNQX (10 µM).
A more detailed spatiotemporal mapping of these
spontaneous LFP could be obtained using a transparent 256electrode array (Fig. 7). Mapping of this 256-electrode array
is adapted to the shape of the spinal cord (8mm x 2mm).
Maps were built using surface spline interpolation [17].

(CEA-LETI), A. Bourgerette (CEA-LETI), A. Defontaine
(CEA-LETI), R. Escola (CEA-LETI), S. Lagarde
(CEA-LETI), C. Moulin (CEA-LETI), F. Goy (Bio-Logic),
B. Mercier (Groupe ESIEE), P. Meyrand (CNRS-CNIC), R.
Guillemaud (CEA-LETI) for their expertise and participation
in the project.
REFERENCES
[1]

G. T. Kovacs. Electronic sensors with living cellular components.

[2]

Proceeding of the IEEE, vol. 91, pp. 915-929, 2003
W. L. Rutten. Selective electrical interfaces with the nervous system.
Annual Review of Biomedical Engineering, vol. 4, pp. 407-452,

[3]

2002.
Beggs JM, Plenz D (2003) Neuronal Avalanches in Neocortical
Circuits. J Neurosci 23:11167-11177.

[4]
[5]

G. Buzsaki. Large-scale recording of neuronal ensembles. Nature
Neuroscience, vol. 7, pp. 446-451, 2004.
Nicolelis MA, Ghazanfar AA, Faggin BM, Votaw S, Oliveira LM

[6]

Reconstructing the engram: simultaneous, multisite, many single
neuron recordings. Neuron 18:529-537, 1997.
Eytan D, Brenner N, Marom S Selective Adaptation in Networks of

[7]

Cortical Neurons. J Neurosci 23:9349-9356, 2003.
E. Margalit et al.. Retinal prosthesis for the blind. Survey of
ophthalmology. vol. 47, pp. 335-356, 2002.

[8]
[9]

5 ms between 2 ma

Fig 7 : Map of Mouse spinal cord activities (LFP)
with a 256-electrode array and the BioMEATM System

VI. CONCLUSION
Microelectrode arrays have been realized with a DRIE
protocol allowing any aspect ratios of 3D microelectrodes. A
64-channel ASIC with recording and stimulation capabilities
for each channel has been designed and produced. A
complete 256-channel in vitro setup has been built. This
system includes 4 ASICs, the electronics controlling the
acquisition and the stimulation phases, and a user-friendly
software interface. The output data files are CED spike2 data
files, so that acquired data can be read back offline using
Spike2 software. First validations of the full (64- and 256channel) systems have been performed on acute spinal cord
preparations.

ACKNOWLEDGMENT
This work is supported by The French Ministry of
technology (NEUROCOM RMNT Project and ACI), the
Région Aquitaine, and grants from the Fyssen and FRM
foundations.
The authors wish to thanks J-P. Rostaing (CEA-LETI), S.
Gharbi (CEA-LETI), M. Trevisiol (CEA-LETI), V. Perrais
(Groupe ESIEE), S. Joucla (CNRS-CNIC), M. Antonakios

Nicolelis MA Actions from thoughts. Nature 409:403-407, 2001.
Jackson A, Mavoori J, Fetz EE Long-term motor cortex plasticity
induced by an electronic neural implant. Nature 444:56-60, 2006.
[10] JOURNAL OF MICROELECTROMECHANICAL SYSTEMS, VOL.
13, NO. 2, APRIL 2004 “The MEMSNAS Process: Microloading
Effect for Micromachining 3-D Structures of Nearly All Shapes”
Tarik Bourouina, Member, IEEE, Takahisa Masuzawa, and Hiroyuki
Fujita, Member, IEEE
[11] “Advanced silicon etching techniques based on deep reactive ion
etching (DRIE) for silicon harms and 3D micro- and nano-structures”
F.Marty, L.Rousseau, B.Saadany, B.Mercier, O.Français, Y.Mita,
T.Bourouina
[12] R. A. Blum, J. D. Ross, C. M. Simon, et al., "A Custom
Multielectrode Array with Integrated Low-Noise Preamplifiers,"
presented at Annual International Conference of the IEEE
Engineering in Medicine and Biology - Proceedings, 2003, pp. 33963399.
[13] W. Dabrowski, P. Grybos, P. Hottowy, et al., "Development of
integrated circuits for readout of microelectrode arrays to image
neuronal activity in live retinal tissue" presented at IEEE Nuclear
Science Symposium Conference Record, 2003, pp. 956-960.
[14] L. Berdondini, P. D. Van Der Wal, O. Guenat, et al., "High-density
electrode array for imaging in vitro electrophysiological activity",
Biosensors and bioelectronics, 2005, vol. 21, pp. 167-174.
[15] Penn AA, Riquelme PA, Feller MB, Shatz CJ (1998) Competition in
retinogeniculate patterning driven by spontaneous activity. Science
279:2108-2112.
[16] Yvert B, et al.. Multiple Spontaneous Rhythmic Activity Patterns
Generated by the Embryonic Mouse Spinal Cord Occur Within a
Specific Developmental Time Window. J. Neurophysiol., vol. 91, pp.
2101-2109, 2004.
[17] Perrin F, Pernier J, Bertrand O, Giard MH, Echallier JF. Mapping of
scalp
potentials
by
surface
spline
interpolation.
Electroencephalography and Clinical Neurophysiology 66:75-81,
1987.

174


charvet_ieee2008_256.pdf - page 1/4
charvet_ieee2008_256.pdf - page 2/4
charvet_ieee2008_256.pdf - page 3/4
charvet_ieee2008_256.pdf - page 4/4

Documents similaires


Fichier PDF charvetieee2008256
Fichier PDF charvetieee200864
Fichier PDF 0054
Fichier PDF 1
Fichier PDF charvet embc 2013
Fichier PDF diagramme de classe


Sur le même sujet..