IEEE Circuits and Systems Magazine - Q4 2020 - 29

I. Introduction
ptogenetics is a neural stimulation technique that
allows using light to control and monitor individual neurons in living tissue. This groundbreaking
approach has changed the way neuroscientists study the
brain and its associated diseases by enabling new promising research paradigms involving freely-behaving rodents, especially mice, which serve as common disease
models. As a result, the development of advanced optogenetic hardware combining multiple technologies, such as
bioMEMS, optics, microelectronics, and wireless technology has become a source of significant interest.
A paradigm holding promise to accelerate the development of new therapeutics against brain diseases
draws from a parallel utilization of optogenetics and
electrophysiology. These complementary a
- pproaches
can be used to close the loop and regulate the neural
circuits in real time through the utilization of smart implantable sensors and actuators [1]-[3]. Closed-loop (CL)
optogenetics consists of activating or silencing specific
neural circuits to reach or restore a normal activity level in a biological neural network. Among several future
revolutionary applications in medicine, this approach
is envisioned to allow artificial neural connections to
bypass a deficient neural circuit, to reinforce a specific
network, or to induce plasticity, i.e. repair neural connections artificially [3]-[7].
This paper covers the design of advanced systems
for performing CL optogenetics with live neural circuits.
First, we present an overview of the important paradigm
of CL optogenetics and its associated implementation
challenges. We show the critical importance of neural
electrophysiology, optogenetics, digital neural signal
processing and wireless systems design for the application of this paradigm to small laboratory mice. After reviewing the existing solutions, we present the design of a
complete wireless system for performing CL optogenetics, and we cover all modules, including a neural bioamplifier and a low-power programmable Delta-Sigma ^TR h
analog-to-digital converter, as well as energy-efficient
photostimulator circuits for delivering enough light inside the brain tissues to trigger the light-sensitized proteins inside the neurons. Then, we present the design
of advanced digital building blocks for implementing a
real time CL controller with very short latency, including a neural activity detector, a neural data compression
core, and a spike sorting core. We present a robust and
low-power strategy to detect, compress and classify
the neural waveforms in real time, over many recording
channels, in order to reliably close the loop between the

recording module and the photo-simulator. Finally, we
present experimental results obtained in vivo with an
electro-optic prototype utilized in CL with a freely moving mouse virally expressing ChR2-mCherry in inhibitory
neurons of the prelimbic cortex.
The paper is organized as follows: Section II presents
an overview of CL optogenetics, the design challenges
to address, and the state-of-the-art solutions; Section
III presents a complete electro-optic system for CL optogenetics; Section IV covers the design of all building
blocks: a mixed signal integrated circuit (IC) for neural
recording and optical stimulation; Section V covers the
design of the digital neural signal processing cores; Section VI presents a CL strategy and the implementation of
a CL controller; Section VII shows the system integration
and encapsulation; followed by in vivo results in Section
VIII and conclusion in Section IX.
II. Hardware for Closed-Loop Optogenetics
Performing optogenetic stimulation in parallel with electrophysiology recording to close the loop requires five
essential components (Fig. 1(a)): an optrode i.e. a device
that can sense biopotentials and deliver light [8] (e.g.
electrodes and optical fibers), a neural recording interface, a neural signal processor, a CL controller and an
optical stimulator. First, an optrode senses the low-voltage electrical activity of the brain on several channels
in parallel, whose voltages are further conditioned and
digitized by a neural recording interface. Then, a closed
loop can be established between the neural recording
interface and the photo-stimulator through a digital signal processing module and a dedicated controller. The
digitized neural signals are passed to an embedded processor, or an off-line computer to assess the activity patterns, and issue proper stimulation. The neural signal
processor and the controller can be used to trigger optical stimulation sequences when specific neural patterns
occurs. When activated, the optical stimulator delivers
proper electrical currents to enable the light sources
(laser, LED, ÂµLED, etc.), and deliver light through the
optrode. Then, the stimulation light activates a specific
neural circuit, changing in turn the recorded neural signal, thus implementing a feedback loop. It is worth noting that to predict the neural circuit behavior, the growing field of Computational optogenetics can provide
an accurate model of light-sensitive protein in realistic
tissue/organ settings [9], [10], which can greatly help in
the design of a precise CL system by predicting the behavior of the cells in feedback to stimulation. Moreover,
although this paper focuses on the electrical activity as

Gabriel Gagnon-Turcotte, Member, IEEE, Guillaume Bilodeau, Olivier Tsiakaka and Benoit Gosselin, Member, IEEE.

FOURTH QUARTER 2020

IEEE CIRCUITS AND SYSTEMS MAGAZINE

IEEE Circuits and Systems Magazine - Q4 2020

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