IEEE Circuits and Systems Magazine - Q4 2020 - 31

input to the CL systems, other sensing means that are
not covered in this paper exist to assess brain activity,
such as optical fluorescence signal recording [11].
The concept of a wireless system for CL optogenetics
is presented in Fig. 1(b). The embedded neural signal processor can detect, compress and sort the APs in real-time,
while the CL controller issues stimulation triggers dependent on the firing rate pattern recognition. The processed
neural data is transmitted wirelessly through telemetry
to allow for live monitoring and/or data storage and management in the cloud. The design of a complete system is
covered in detail as a use case in Section III.
A. Challenges & Requirements
The design of a wireless electro-optic platform intended
for performing CL optogenetics with live neurons is
facing several challenges and must address important
requirements.
1) Neural Electrophysiology
Extracellular APs can have very low amplitude ranging
from ten ÂµV up to 500 ÂµV, and a frequency content between 300 Hz and 5 kHz [16]. Recording this signal requires designing a suitable low-noise (IRN < 10 ÂµVrms [17],
[18]) and low-power bioamplifier, and a precise analogto-digital converter (ADC) that can be duplicated across
many parallel channels without consuming too many
resources. Sampling rates of 20 kHz and above are commonly used for better resolution [19]-[21].

several challenges, mainly regarding the material for
light delivery. In the most advanced designs, ÂµLEDs
are utilized in place to generate photons in situ, hence
avoiding the utilization of the waveguides [38]. In this
approach, the shank implementation (Fig. 2(c)) is widely used with these micro light sources on a (semi-) rigid
thin needle shaped substrate for an easier insertion in
brain tissues [31], [39].
4) Digital Neural Signal Processing & CL Controller
Analyzing the dense neural activity patterns in real time
to quickly issue proper stimulation feedback is very
challenging in terms of hardware resources and speed.
PC-based and hardwired systems are not practical for

Electrodes

Light Guide

(a)
Electro-Optic Array. (ZnO)

2) Optogenetic Stimulation
Providing optogenetic stimulation requires delivering
enough optical power to the genetically-modified neurons in order to properly activate its light sensitive proteins. For commonly used proteins, such as ChR2, the
minimum optical power per unit of area that is required
for activation is of around 0.1 mW/mm2 [22]. Once the
output light reaches that threshold, the exposed cells
emit APs whose firing rates depend on the light stimulation intensity and duration [23].
3) Optrode Design and Manufacturing
An optrode establishes an electro-optic interface between the hardware modules and the neural tissue.
Recent micro-optrodes range from a combination of
optical and electrical probes to engineered multimodal
fibers with dual capabilities (Fig. 2(a)), either in single
[17], [35], comb [36] or array implementations [37]. Although holding great promise, this approach has been
shown prone to photo-electric artifacts, mainly due to
fiber and electrode material. In this regard, the use of
light waveguides integrated into silicium overcomes
this limitation (Fig. 2(b)). However, these advances face
FOURTH QUARTER 2020 		

(b)
Microfabricated
Electrodes

ÂµLED

Microfabricated
Optrode Shank

(c)
Figure 2. Different types of optrodes: (a) packaged microelectrodes along with optical fibers, (b) optrode with integrated light waveguides and (c) shank implementation using
micro light sources on a (semi-) rigid thin needle.

IEEE CIRCUITS AND SYSTEMS MAGAZINE	

31
IEEE Circuits and Systems Magazine - Q4 2020

Table of Contents for the Digital Edition of IEEE Circuits and Systems Magazine - Q4 2020
