Our reference: NSM 6644
P-authorquery-v9
AUTHOR QUERY FORM
Journal: NSM
Please e-mail or fax your responses and any corrections to:
E-mail: corrections.esch@elsevier.thomsondigital.com
Article Number: 6644
Fax: +353 6170 9272
Dear Author,
Please check your proof carefully and mark all corrections at the appropriate place in the proof (e.g., by using on-screen
annotation in the PDF file) or compile them in a separate list. Note: if you opt to annotate the file with software other than
Adobe Reader then please also highlight the appropriate place in the PDF file. To ensure fast publication of your paper please
return your corrections within 48 hours.
For correction or revision of any artwork, please consult http://www.elsevier.com/artworkinstructions.
Any queries or remarks that have arisen during the processing of your manuscript are listed below and highlighted by flags in
the proof. Click on the ‘Q’ link to go to the location in the proof.
Location in
article
Q1
Q2
Query / Remark: click on the Q link to go
Please insert your reply or correction at the corresponding line in the proof
Please confirm that given names and surnames have been identified correctly.
The country name has been inserted for all the affiliation. Please check, and correct if necessary.
Please check this box or indicate your approval if
you have no corrections to make to the PDF file
Thank you for your assistance.
ARTICLE IN PRESS
G Model
NSM 6644 1–3
Journal of Neuroscience Methods xxx (2013) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Journal of Neuroscience Methods
journal homepage: www.elsevier.com/locate/jneumeth
Basic Neuroscience
Short communication
1
2
Roboneuron: A simple and robust real-time analog spike simulator
and calibrator
3
4
5
6
7
8
Q1
Tyler Dickerhoff a , Abidin Yildirim b , Timothy J. Gawne c,∗
a
Department of Neuroscience, Vision Science Research Center, University of Alabama at Birmingham (UAB), Birmingham, AL 35294, United States
Vision Science Research Center, University of Alabama at Birmingham (UAB), Birmingham, AL 35294, United States
Q2 c Department of Vision Sciences, University of Alabama at Birmingham (UAB), Birmingham, AL 35294, United States
b
9
10
11
12
13
14
15
h i g h l i g h t s
•
•
•
•
A simple and robust real-time analog spike simulator is presented.
Files for ordering commercial grade circuit boards are included.
The system generates two asynchronous overlapping spikes.
Spikes can fire at a set rate or under external voltage control.
16
17
a r t i c l e
i n f o
a b s t r a c t
18
19
20
21
22
Article history:
Received 3 April 2013
Received in revised form 14 May 2013
Accepted 15 May 2013
23
29
Keywords:
Spiking model
Circuit
Neuron
Spike sorter
Calibrator
30
1. Introduction
24
25
26
27
28
31
32
33
34
35
Background: Modern computerized spike recording systems are increasingly powerful and sophisticated.
However, this increases the importance of performing validation by recording signals from a system with
a known input–output relationship.
New method: We present here a simple and robust analog circuit that uses a minimum number of commonly available components to simulate two independently spiking neurons. The two neurons generate
asynchronous overlapping spikes. These can be independently set to spike at either a constant rate, or at
a rate set by an external control voltage.
Results: The circuit is simple enough to easily assemble by hand, however, standard files for ordering
commercial printed circuit boards are also supplied. Several units were built by different people, using
both hand-assembly and commercially manufactured printed circuit boards: all worked well. The circuit
is robust with respect to supply voltages and component values.
Comparison with existing methods: Existing analog circuits tend to be complex, hard to assemble, and use
hard-to-find components. Digital simulators typically require specific development systems that have
steep learning curves and are likely to change radically or become unavailable very quickly. This system
has been optimized to be robust, simple, and use only commonly available components.
Conclusions: When validating a system there could be an advantage to using a calibrator that is robust,
whose input–output relationship is simple, and whose design is stable over time.
© 2013 Published by Elsevier B.V.
“I would have written a shorter letter, but I did not have the
time.” – Blaise Pascal.
As computerized data acquisition systems become more complex and sophisticated, the need to run validation trials on systems
with known response properties becomes ever more critical.
∗ Corresponding author at: University of Alabama at Birmingham, 924 South 18th
Street, Birmingham, AL 35294, United States. Tel.: +1 205 934 5495;
fax: +1 205 934 5725.
E-mail address: tgawne@gmail.com (T.J. Gawne).
Several options for doing this exist. One can take a primary sensor
– a photocell, or a microphone, or a strain gauge, etc., depending
upon the experiment – and record the resultant signal. This is a
very useful technique, but it cannot in general test the detection
and sorting of action potentials.
A commercially available system exists, the “Spike Simulator”
(FHC Inc., Bowdoin ME), it is an extremely useful device but does
not have external control inputs. There are published analog
circuits (French and Stein, 1970; Schweitzer-Tong, 1983), but they
require hand-circuit breadboarding construction, or are sensitive
to component values or use archaic parts. VLSI or other digitalbased systems have also been designed (Delbruck and Liu, 2004;
Li et al., 2010; Peterson and Ohzawa, 1999; Saito et al., 2008), but
0165-0270/$ – see front matter © 2013 Published by Elsevier B.V.
http://dx.doi.org/10.1016/j.jneumeth.2013.05.010
Please cite this article in press as: Dickerhoff T, et al. Roboneuron: A simple and robust real-time analog spike simulator and calibrator. J Neurosci
Methods (2013), http://dx.doi.org/10.1016/j.jneumeth.2013.05.010
36
37
38
39
40
41
42
43
44
45
46
47
48
G Model
NSM 6644 1–3
ARTICLE IN PRESS
T. Dickerhoff et al. / Journal of Neuroscience Methods xxx (2013) xxx–xxx
2
Fig. 1. Circuit diagram for the roboneuron real-time spike simulator and calibrator.
69
they cannot be constructed without either specialist equipment, or
specific software libraries that need to be downloaded, installed,
patched, understood, updated, etc. and which in any event are
likely to either go away or be radically changed in the near future.
Digital circuits can also produce large amounts of broadband noise,
which in some applications may prove problematic. Also, without
great care digital circuits are likely to produce simulated spikes
at multiples of an internal clock, which could potentially be a
confound in some settings.
Here we present a simple carefully debugged real-time analog
spike simulator and calibrator circuit. It simulates two spikes with
different waveforms, creates asynchronous overlapping spikes, and
the firing rate of either one or both spikes can be controlled by
an analog input voltage. It uses common parts, and is thus likely
to be able to be constructed without modification for decades to
come. It has a simplified design with a minimum of complexity,
thus making it tractable for hand-circuit construction, but we also
make the design files available so that printed circuit boards can
be ordered from commercial fabricators to further aide construction. The circuit has also proven useful as a demo in an educational
setting.
70
2. Methods
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
71
72
73
The circuit diagram is shown in Fig. 1. A parts list, the design
files for a printed circuit board, and photographs of the completed
design, are given in the supplementary material.
The circuit uses only a single 9-V battery, and is robust with
respect to power supply voltage. There are only two kinds of integrated circuits in the design: an LM324 quad operational amplifier,
which is extremely common, and an LM231 voltage-to-frequency
converter chip. The LM231 is not quite as common as the LM324,
but it has been around for over a decade, is well stocked at all major
electronics supply houses, and provides pulses of the appropriate
shapes and frequency ranges stably and with a minimum of components. We attempted several alternative designs using both the
ubiquitous 555 chip and some other voltage-controlled oscillator
chips, and found it surprisingly difficult to get these circuits to operate reliably for the purposes of this design without adding in a lot
of extra level-shifting or other support circuitry.
Both simulated spikes are rectangular pulses with maximum
amplitudes of 2.5 V. This level was chosen to allow headroom for
generating superimposed spikes. Simulated spike A has a width of
0.8 ms, and a maximum frequency of approximately 300 Hz. Simulated spike B has a width of 0.4 ms, and a maximum frequency
of approximately 600 Hz. As the design was optimized for use as
a calibrator, no attempt was made to simulate the details of neuronal dynamics: the relationship between the control voltages and
the firing rates is linear over the entire range. However, the simulator will tend to spike synchronously with the onset of a large step
increase in control voltage.
Even though most single-unit labs have audio monitors, we
included a provision for an AC-coupled audio output: this is especially useful for classroom demonstrations. There is also a simple
Please cite this article in press as: Dickerhoff T, et al. Roboneuron: A simple and robust real-time analog spike simulator and calibrator. J Neurosci
Methods (2013), http://dx.doi.org/10.1016/j.jneumeth.2013.05.010
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
G Model
NSM 6644 1–3
ARTICLE IN PRESS
T. Dickerhoff et al. / Journal of Neuroscience Methods xxx (2013) xxx–xxx
3
103
voltage divider to provide a low-level output: no attempt was made
to model a high-impedance source, this is solely so that the output
can be fed into a unit preamplifier without saturating it.
rectangular pulses make it completely clear that the recorded data
is from a simulator, and the rectangular pulses are also a good way
to check how much distortion your analog filters are adding to the
recorded signals.
Roboneuron was not intended to duplicate the firing statistics of
real neurons, but to generate spikes in a deterministic manner for
testing and calibration. In principle one could generate spikes with
an approximate Poisson distribution by feeding in a random signal
to the voltage control inputs: the timing capacitors will integrate
the random signal over time and this will add jitter to the interspike interval timing. The problem is that it is surprisingly hard to
generate robust random noise signals using only analog components, the approach tends to need tweaking, be sensitive to noise,
etc. (see Horowitz and Hill, 1989). If you really need to simulate the
actual dynamics of real neurons it would be better to just bite the
bullet and use a microprocessor-based system.
For visual studies we have found that one of the most useful
accessories is a silicon solar cell of about 1 cm2 , of the kind commonly found in science fair kits. Such a cell can drive the external
control inputs of this simulator with reasonable illumination levels,
and it does not require a power supply or any additional support
circuitry.
An advantage of this circuit is that, most likely, it will be possible
to construct copies of it without change for decades to come. When
validating a system there could be an advantage to using a calibrator
whose design is stable over time.
104
3. Results
Acknowledgments
This work was supported by NSF grant IOS 0622318 and NEI
grant P30 EY003039 (Core), and a grant from the Oak Ridge Associated Universities (ORAU). The authors acknowledge the technical
assistance of Jerry Millican and Alexander Zotov.
117
A sample output of the simulator is shown in Fig. 2. Several
boards have been assembled using either hand-construction techniques, or the commercially manufactured printed circuit board:
all worked well.
There is only one limitation of this circuit of which we are aware:
at some levels of spike amplitude and frequency the two simulated
spikes will occasionally phase-lock. This does not happen commonly, but in any analog design there will always be some degree of
coupling between circuit elements and thus in an analog circuit this
cannot be completely eliminated. People building their own versions from hand should try and keep the two voltage-to-frequency
subsections of the circuit as far away from each other as possible,
especially the timing capacitors.
118
4. Discussion
Fig. 2. Sample waveform output of roboneuron. Here both simulated neurons are set
to fire at a relatively high rate in order to demonstrate the asynchronous overlapping
spikes.
101
102
105
106
107
108
109
110
111
112
113
114
115
116
119
120
121
122
123
124
125
126
127
At first we had thought to mimic the shape of an extracellularly recorded action potential, but realized that rectangular pulses
work as well if not better for our purposes. First, trying to create
a relatively realistic spike shape using an underdamped bandpass
filter would have required either an active filter with a split power
supply, or a passive filter with an inductor. This seemed pointless because rectangular pulses of different widths and amplitude
should span the range of any spike sorter. Rectangular pulses make
it obvious if a spike sorter can handle at least simple cases of
overlapping spikes correctly, and any failure is obvious. Finally,
Appendix A. Supplementary data
Supplementary material related to this article can
be found, in the online version, at http://dx.doi.org/
10.1016/j.jneumeth.2013.05.010.
References
Delbruck T, Liu SC. A silicon early visual system as a model animal. Vision Res
2004;44:2083–9.
French AS, Stein RB. A flexible neural analog using integrated circuits. IEEE Trans
Bio-Med Eng 1970;Bm17:248–53.
Horowitz P, Hill W. The art of electronics. 2nd ed. Cambridge, UK: Cambridge University Press; 1989. p. 658.
Li G, Talebi V, Yoonessi A, Baker CL. A FPGA real-time model of single and multiple
visual cortical neurons. J Neurosci Meth 2010;193:62–6.
Peterson M, Ohzawa I. VNS (Visual neuron simulator). http://neurovision.
berkeley.edu/Demonstrations/VSOC/vsoc/vsoc main.html#VNS [08.05.99].
Saito Y, Suginohara H, Ohzawa I. Real-time visual neuron simulator.
http://visiome.neuroinf.jp/modules/xoonips/detail.php?item id=3308
[18.01.08].
Schweitzer-Tong DE. The photoneuromime – an artificial visual neuron for dynamic
testing of computer-controlled experiments. Behav Res Methods Instrum
1983;15:9–12.
Please cite this article in press as: Dickerhoff T, et al. Roboneuron: A simple and robust real-time analog spike simulator and calibrator. J Neurosci
Methods (2013), http://dx.doi.org/10.1016/j.jneumeth.2013.05.010
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179