IEEE Solid-States Circuits Magazine - Fall 2023 - 42

Closed-loop neuromodulation systems and
BCIs combine knowledge from neuroscience,
engineering, and computer science to facilitate
real-time, bidirectional interaction between the
brain and implantable, wearable, or assistive
devices.
past decade, the neuromodulation and
BCI fields have witnessed significant
growth in both academic and industrial
sectors [1], [2], [3], [4]. Nevertheless,
their clinical impact has remained
relatively limited to date. This is partially
due to the intricate nature of
the human brain, challenges in device
translation and validation, concerns
on long-term reliability, and the lack of
technologies with high efficacy needed
for widespread use. Overcoming these
challenges requires technological advancements,
interdisciplinary collaborations,
and extensive clinical studies
to unlock the transformative potential
of neural interfaces for widespread
clinical applications.
Invasive and Noninvasive
Neurotechnologies
Table 1 summarizes the neural recording
and stimulation modalities
used in both invasive and noninvasive
neurotechnologies, highlighting
their advantages and disadvantages.
Currently, invasive neuromodulation
strategies based on electrical stimulation,
such as deep brain [4] and cortical
[5] stimulation, have emerged
as promising therapeutic options
for a wide range of brain disorders,
including medication-resistant epilepsy,
dystonia, essential tremor, and
obsessive-compulsive disorder (OCD),
among others. While optogenetics
has shown great promise in animal
studies, its translation to human and
clinical applications is still in its early
stages due to concerns related to genetic
modification of neurons and
associated ethical considerations. In
invasive neural interfaces, the sensing
modality for closed-loop feedback
may include high-resolution action
potentials, or lower-frequency signals,
such as electrocorticography (ECoG),
local field potential (LFP), and stereoelectroencephalography
(sEEG).
Conversely, there is an increasing
interest in developing noninvasive or
minimally invasive technologies for
individuals with milder neurological
or psychiatric conditions. These
approaches utilize wearable scalp or
subscalp EEG, functional near-infrared
spectroscopy [6], peripheral or behavioral
measures for sensing, and
noninvasive methods like focused
ultrasound [7], transcranial magnetic
stimulation [8],
transcranial direct
current stimulation, and temporal interference
[9] for brain intervention.
The aim is to deliver effective stimulation
with reduced risks compared
to invasive methods and enhance patient
convenience. However, noninvasive
approaches face challenges, such
as low efficacy and limited spatial resolution.
Additionally, these methods
are often bulky, lack portability, and
involve complex setup and positioning.
Given the predominant use of
invasive technologies in both commercial
and research-based closedloop
systems, our focus in this article
will be limited to invasive interfaces.
Current and Emerging
Neural Devices
The neuromodulation and BCI device
markets are expected to grow significantly
by 2030, driven by technological
advancements, the increasing
incidence of neurological and mental
disorders around the world, and the
aging population. Key industry players
like Medtronic, Boston Scientific,
and Abbott-as well as pioneering
neurotech companies and startups,
such as NeuroPace, Blackrock Neurotech,
Neuralink, Cortec, Synchron,
and Paradromics-are actively involved
in the development of BCI and
neurostimulation devices, with a focus
on invasive systems (Figure 1).
The FDA-approved closed-loop
responsive neurostimulation (RNS)
system [5] analyzes ECoG signals from
up to four channels to detect and disrupt
seizures, by comparing a specific
epileptic " feature " against a predetermined
threshold. The AspireSR vagus
nerve stimulator by LivaNova is capable
of dynamically adapting stimulation
based on changes in ictal heart
TABLE 1. COMPARISON OF INVASIVE AND NONINVASIVE NEUROTECHNOLOGIES FOR NEURAL RECORDING
AND NEUROSTIMULATION.
Invasive
Noninvasive
SENSING MODALITY
AP, ECoG, LFP, stereo-EEG
EEG, fNIRS, subscalp EEG
STIMULATION TYPE
DBS, cortical, epicranial,
optogenetics
tDCS, focused US, TMS, TI
PROS
High resolution, High
efficacy, invisible
No major surgery, safe
and accessible
CONS
Surgical intervention,
higher risk and cost
Low resolution, low
efficacy, obtrusive
AP: action potential; DBS: deep-brain stimulation; ECoG: electrocorticography; EEG: electroencephalography; fNIRS: functional near-infrared
spectroscopy; LFP: local field potential; tDCS: transcranial direct current stimulation; TI: temporal interference; TMS: transcranial magnetic stimulation;
US: ultrasound.
42
FALL 2023
IEEE SOLID-STATE CIRCUITS MAGAZINE
IEEE Solid-States Circuits Magazine - Fall 2023

Table of Contents for the Digital Edition of IEEE Solid-States Circuits Magazine - Fall 2023
