This article has been accepted for publication in a future issue of this journal, but has not been

fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JERM.2019.2904024, IEEE Journal
of Electromagnetics, RF and Microwaves in Medicine and Biology
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 1

Passive Impedance Matching For Implanted

Brain-Electrode Interfaces
Wei-Chuan Chen, Student Member, IEEE, Katrina Guido, and Asimina Kiourti, Member, IEEE

Batteryless and wireless brain implants matched to high-impedance clinical electrodes.

• Radio-Frequency backscattering is employed to enable batteryless and wireless brain implants that are: a)
matched to high-impedance clinical electrodes, and b) tolerant to DC voltage.
• As compared to previous wireless and batteryless brain implants, the proposed approach offers a remarkable
improvement in sensitivity by 25 times.
• Unobtrusive monitoring of deep brain signals may significantly improve the individual’s physical and mental
well-being (e.g., for patients with epilepsy, Alzheimer’s, Parkinson’s, and more).
• Batteryless brain implants matched to high-impedance electrodes can readily be employed to clinical
applications.
• Improvements on the interrogator side help suppress the phase noise and improve the demodulated signal
integrity.

2469-7249 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JERM.2019.2904024, IEEE Journal
of Electromagnetics, RF and Microwaves in Medicine and Biology
IEEE JOURNAL OF ELECTROMAGNETICS, RF AND MICROWAVES IN MEDICINE AND BIOLOGY 1

Passive Impedance Matching For Implanted

Brain-Electrode Interfaces
Wei-Chuan Chen, Student Member, IEEE, Katrina Guido, and Asimina Kiourti, Member, IEEE

Abstract—We propose a new technique for matching the high impedance of sub-cranial electrodes to wireless brain implants that
is: a) passive, b) highly tolerant to the DC offset voltage caused by the electrochemical reaction in the recording electrode, and c)
complemented by an improved external interrogator design that exhibits reduced phase noise. As compared to previous wireless and
batteryless brain implants, the proposed approach offers a remarkable improvement in sensitivity by 25 times. The proposed system
consists of an external interrogator and a neuro-recorder implanted under the scalp. For operation, the interrogator sends a 2.4
GHz carrier signal to “turn on” the implant. This carrier self-biases a PNP Bipolar Junction Transistor (BJT) that enables matching
to the recording electrode at frequency fneuro in a batteryless manner. Concurrently, the recorded neuropotentials (at frequency
fneuro ) pass through a Schottky diode that allows them to mix with the carrier and generate a 4.8 GHz ± fneuro modulated signal.
The latter is then transmitted back to the interrogator for demodulation. To verify the implant’s operation, in-vitro measurements
are presented. Measurement results demonstrate that emulated neuropotentials as low as 200 µVpp can be detected at a 33 kΩ
electrode impedance. As such, the proposed system presents a game-changing capability for a wide range of applications.
Keywords—Biomedical telemetry, brain implant, DC offset, electrode, impedance matching, neurosensing, passive, wireless.
I. I NTRODUCTION TABLE I
VOLTAGE A ND F REQUENCY R ANGE O F S IGNALS G ENERATED B Y T HE

R ADIO-FREQUENCY (RF) backscattering is widely used H UMAN B RAIN [10]

today for applications as diverse as passive radio fre-
Neural Signals Voltage Range Frequency Range
quency identication (RFID), the Internet of Things (IoT), and
so on. One emerging application of RF backscattering relates Local Field Potential 20∼ 2000 µVpp 1∼500 Hz
to wireless and fully-passive monitoring of deep brain activity
Action Potential 20∼ 2000 µVpp 250 Hz∼10 kHz
[1]–[6]. Possible clinical applications include: 1) detection and
interruption of early epileptic seizures, 2) behavioral studies Electroencephalogram 2∼ 100 µVpp 0.5 Hz∼50 Hz
of consciousness, 3) understanding and improving the brain’s
functionality for patients with Alzheimer’s and Parkinson’s
disease [7], [8]. In brief, an external (wearable) interrogator 50 Ω function generator that emulated neuropotential activity.
sends a carrier signal toward an underlying batteryless implant. But unfortunately, the impedance of clinical electrodes is in
As soon as the implant turns on, it mixes the carrier with the the order of tens of kΩ, implying a signicant deterioration
sensed neuropotentials, and further backscatters the resulting in sensitivity under in-vivo conditions. As an example, the
mixing products back to the interrogator. Contrary to state- sensitivity of the brain implant reported in [3] degrades from
of-the-art integrated circuit (IC) brain implants, the aforemen- 20 µVpp to 5 mVpp when attached to a 33 kΩ resistor. Addi-
tioned approach does not require implantable pre-ampliers, tionally, the DC offset voltage caused by the electrochemical
digital controllers, Analog-to-Digital Converters (ADC), or reaction in the recording electrode is unaccounted for in [3].
other power-hungry components [9]–[11], thus eliminating Assuming an in-vivo scenario, this DC offset greatly changes
batteries and temperature increase in the surrounding tissue the bias of the implanted diode used for mixing, further
[12]. decreasing the system sensitivity.
Our latest research has demonstrated a miniature (8.7 mm Conventional integrated chips (ICs) used for neural record-
x 10 mm) wireless and fully-passive brain implant with a ing utilize a rst-stage neural amplier that serves to increase
remarkable 20 µVpp sensitivity under in-vitro conditions [3]. the input impedance and cancel the DC offset voltage [13].
Such high sensitivity theoretically covers all signals generated However, these neural ampliers are typically associated with
by the human brain, including neural spikes and local eld high power consumption and require very stable DC voltage
potentials (LFPs) [10]. Nevertheless, the aforementioned brain supplies. Expectedly, integration of ampliers is not a viable
implant was optimized with in-vitro considerations in mind. solution for our class of passive brain implants.
That is, the implant was designed to match the impedance of a As an alternative, we herewith report a novel method for
passively matching the high impedance of clinical electrodes
This paper is expanded from a presentation at the 2018 IEEE International to batteryless brain implants. To do so, a Bipolar Junction
Symposium on Antennas and Propagation and USNC-URSI Radio Science
Meeting held in Boston, MA, USA.
Transistor (BJT) is integrated into the implant and the overall
W. Chen, K. Guido, and A. Kiourti are with the ElectroScience Lab- circuit is redesigned accordingly. In brief, the BJT is self-
oratory, Department of Electrical and Computer Engineering, The Ohio biased by the external carrier and serves as an impedance
State University, Columbus, OH 43212 USA (e-mail: [email protected]; buffer between the electrode and the circuitry. Concurrently,
[email protected]; [email protected];)
This work was supported by the National Science Foundation under Grant the DC offset voltage issue is eliminated. Modications on
1763350. the interrogator side of the circuit are also proposed to reduce

2469-7249 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JERM.2019.2904024, IEEE Journal
of Electromagnetics, RF and Microwaves in Medicine and Biology
IEEE JOURNAL OF ELECTROMAGNETICS, RF AND MICROWAVES IN MEDICINE AND BIOLOGY 2

the phase noise and further improve the integrity of the A. Electrode Interface: Circuit Model and Impedance
retrieved neuropotentials. To validate the above, an in-vitro When a clinical sub-cranial electrode is immersed inside
measurement setup is considered with a series resistor used to a liquid-ionic conductor (electrolyte or buffer), the cations
account for the clinical electrode impedance. and anions react with the electrode and create an electrolyte-
electrode double layer (namely, the Helmholtz double layer)
II. M ETHODS AND P ROCEDURES [14]. The rst layer of the Helmholtz double layer is composed
The block diagram of the proposed neurosensing system of the ions absorbed on the surface by the charged electrodes.
with impedance-matching capabilities is shown in Fig. 1. The The second layer consists of the oppositely charged ions
system consists of two parts: 1) a brain implant placed under attracted by the Coulomb force. With these two oppositely
the scalp and attached to a recording electrode that penetrates charged layers, the metal electrode acts like a capacitor and
through the bone to the cortical cortex surface, and 2) an allows the neuropotentials to pass through to the associated
external interrogator placed outside the scalp to communicate neural recorder [15].
with the implanted sensor. Operation of the proposed system The equivalent model of a sub-cranial electrode is well
is summarized as follows. First, the external interrogator analyzed in the literature [16] and is illustrated in Fig. 2. As
transmits a 2.4 GHz carrier signal via the interrogator antenna seen, the circuit model is composed of the electrolyte solution
to activate the brain implant. This signal is rectied at the resistance (Rs ), the double layer interface resistance and
implanted diode and serves to self-bias the BJT. In turn, capacitance (Re and Ce ), and the metal electrode resistance
the high impedance of the BJT assists in matching to the (Rm ). Generally, the solution resistance (Rs ) and the metal
high impedance of the recording electrode. Concurrently, the electrode resistance (Rm ) are negligible as compared to the
implanted diode acts as a mixer that uses the 2.4 GHz carrier double layer interface resistance and capacitance (Re and Ce ).
to upconvert the brain signal (at frequency fneuro ) to 4.8 Because of the double layer capacitance, the impedance of the
GHz ± fneuro . This upconverted third-order product is then electrode is a complex number and changes with frequency.
backscattered by the implant’s antenna and eventually received Referring to Fig. 2, this electrode impedance degrades the
by the interrogator. This signal can then be directly observed signal amplitude at the input of the neuropotential monitoring
in the frequency domain using a spectrum analyzer and/or can system (Vin ) by means of a voltage divider:
be demodulated and observed in the time domain using an
Za
oscilloscope. Vin (ω) = Vsig (ω) × (3)
To boost system sensitivity, a power budget analysis is Za + Ze
hereafter presented. The ultimate aim is to lower the minimum where Vsig is the neuropotential amplitude generated within
detectable neuropotential level, expressed as: the brain, Ze is the complex electrode impedance, and Za
is the complex neuro-sensor impedance. The voltage divider
equation shows that the magnitude of Za would decrease Vin
M DSneuro [dBm]
(1) and cause phase distortion when Ze >> Za [16].
= Receiver Sensitivity[dBm] + Lsys [dBm]. To better understand the effect of electrodes on the neu-
where Lsys is the overall system loss, and rosensing system performance, we proceed to characterize via
Receiver Sensitivity is the minimum detectable signal electrode impedance spectroscopy the impedance of clinical
level of the receiver (viz. the interrogator). Referring to Fig. macro-electrodes currently used for Deep Brain Stimulation
1, the overall system loss is expressed as: (DBS) surgery at Ohio State’s Wexner Medical Center (FHC
microTargeting mTD differential electrode) [17]. A potentio-
Lsys [dB] = Lprop [dB] + Lconv [dB] stat with a three-electrode setup is employed per Fig. 3. Here,
(2) the macroelectrode contact of the DBS electrode acts as the
+Lcircuit [dB] + Lelectrode [dB] working electrode, while an Ag/AgCl and a Pt wire electrode
where Lprop is the propagation loss between the implanted are used as the reference and counter electrodes, respectively.
and interrogator antenna at 4.8 GHz ± fneuro , Lconv is All three electrodes are immersed inside a phosphate buffered
the conversion loss at the implanted mixer, Lcircuit is the saline (PBS) solution of pH 7.4 which mimics the pH, osmotic
impedance mismatch loss between the antenna and the mixer, concentration, and ion concentration of the human body.
and Lelectrode is the impedance mismatch loss between the Measurement results of electrode impedance magnitude and
recording electrode and the implant. phase as a function of frequency are shown in Fig. 4. As
As is expected, to improve system sensitivity, Lsys must be seen, the impedance magnitude reduces with frequency. At the
minimized. Accordingly, Lprop , Lconv , Lcircuit and Lelectrode smallest frequency where neuropotentials may be identied,
must be minimized to the greatest extent. Techniques for viz. at 0.5 Hz, the electrode impedance is as high as 33 kΩ.
reducing Lprop , Lconv and Lcircuit were explored in our Added to the above, the electrochemical reaction that takes
previous research [3]. Basically, low Lprop could be achieved place at the electrode interface will give rise to different
by optimizing the implanted and interrogator antenna pair. DC voltage levels across different recording electrodes [18]–
Lconv and Lcircuit , on the other hand, could be reduced by [20]. This voltage difference, which may be as high as 50
employing a matching network between the implanted antenna mV , known as the DC offset voltage, can have detrimental
and the Schottky diode. In this work, we instead focus on consequences. In conventional battery-enabled ICs, this offset
reducing Lelectrode by employing a new circuit design. is known to saturate the rst-stage neural amplier, while in

2469-7249 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JERM.2019.2904024, IEEE Journal
of Electromagnetics, RF and Microwaves in Medicine and Biology
IEEE JOURNAL OF ELECTROMAGNETICS, RF AND MICROWAVES IN MEDICINE AND BIOLOGY 3

Fig. 1. Block diagram of the proposed impedance-matching neurosensing system.

our previous fully-passive neurosensing system [3], the offset

is anticipated to change the bias point of the employed antipar-
allel diode pair (APDP) mixer and deteriorate the sensitivity.
These real-world concerns for clinical electrodes are here-
after taken into account. Given that neural signals may be as
low as 0.5 Hz in frequency, a capability to match to at least 33
kΩ of electrode impedance is necessary for the neurosensing
system of Fig. 1. Concurrently, the ability to overcome the DC
offset voltage is a key requirement for the design.
Fig. 3. Three electrode setup used to measure the electrode impedance using
a potentiostat.
B. Passive Brain Implant with High Input Impedance
The operating principle of the proposed technique used to
passively match the electrode-implant interface and eliminate
the DC offset is summarized in Fig. 5. As seen, the implant
consists of: a) an implantable antenna used for wireless
backscattering, b) a Schottky diode that acts as a rectier
in DC mode and as a mixer in RF mode, c) a matching
network used to mitigate Lcircuit between the antenna and
the Schottky diode (composed of two microstrip lines with
open- and short-ended microstrip lines), d) the high-impedance
clinical electrodes, and e) a PNP BJT added between the
Schottky diode and the electrodes to serve as an impedance
buffer.
Circuit operation is composed of two modes, viz. the DC Fig. 4. Clinical brain electrode impedance measurement.
mode (Fig. 5(a)) and the RF mode (Fig. 5(b)), as analyzed
below.
1) DC Mode. To activate the neural sensor, the interrogator implant, the Schottky diode acts as a rectier that serves
transmits a 2.4 GHz carrier signal. Once received by the to create DC current and self-bias the BJT per Fig. 5(a).
Following biasing of the BJT, brain signals coming from
its base should pass through the BJT and, eventually, get
upconverted by the Schottky diode. To do so, the BJT emitter
is connected right after the Schottky diode, while its collector
is connected to ground. Comparing the voltages at the base
(VB ), collector (VC ), and emitter (VE ) terminals, the BJT may
either operate in the forward-active region (VE > VB > VC )
or the saturation region (VE > VB < VC ), Table I. In both
cases, signal may ow from the base to the emitter, while
Fig. 2. Equivalent circuit model of the electrode. the DC voltage at the base may be neglected. This unique

2469-7249 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JERM.2019.2904024, IEEE Journal
of Electromagnetics, RF and Microwaves in Medicine and Biology
IEEE JOURNAL OF ELECTROMAGNETICS, RF AND MICROWAVES IN MEDICINE AND BIOLOGY 4

Fig. 6. Demodulated version of -115 dBm neural signal at 100 Hz. The plot
compares the currently reported system vs. the one previously reported in [3].

Fig. 5. Proposed neural implant design: (a) DC mode, and (b) RF mode.

TABLE II
N ODE VOLTAGE AND BJT O PERATION R EGIONS

Voltage B-E Junction B-E Junction Mode

VE < VB < VC Reverse Forward Reverse-active

VE < VB > VC Reverse Reverse Cut-off

VE > VB < VC Forward Forward Saturation

Fig. 7. Measurement set-up used to assess the neurosensing system perfor-
VE > VB > VC Forward Reverse Forward-active mance.

feature implies tolerance to DC offset. Simulations indicate GHz phase noise before it enters into the circulator. The
that the input impedance of the self-biasing BJT circuit is superior performance of this interrogator is highlighted in Fig.
219 kΩ and remains almost constant across the entire neural 6. Here, a 100 Hz neural signal of -115 dBm is considered
frequency range (0.5 Hz to 1 kHz). Due to the high input at the implant side, and the demodulated signal is contrasted
impedance of the BJT, the circuit of Fig. 5 can readily match at the output of the newly proposed interrogator vs. the one
to the high-impedance electrodes. proposed in [3]. Clearly, the addition of the two bandpass
2) RF Mode. The Schottky diode now serves as a mixer, lters signicantly improves the retrieved signal integrity.
as shown in Fig. 5(b). That is, the diode utilizes the 2.4
GHz carrier signal to upconvert the brain signals (at frequency D. Measurement Setup
fneuro ) and give rise to the third-order harmonic component
(4.8 GHz ± fneuro ). This upconverted signal is backscattered The in-vitro measurement setup used to validate the neu-
toward the interrogator and is, eventually, demodulated to rosensing system of Fig. 1 is shown in Fig. 7. As depicted, a
recover the neuropotentials in the time domain. signal generator (Agilent SG386) feeds a 2.4 GHz carrier with
10 dBm signal level to the interrogator. An arbitrary func-
tion generator (Keysight 33500B) emulates neuropotentials
C. Interrogator with Improved Phase Noise Performance
as sinusoidal waveforms (at frequency fneuro ). To consider
Instability of the signal generator that is used to generate a worst-case scenario for electrode impedance in this study,
the 2.4 GHz carrier creates phase noise. In the time domain a 33 kΩ resistor is used to represent this impedance per
and frequency domain, this noise appears as uctuations and Fig. 4. The improved interrogator of Fig. 1 demodulates
as a skirt centered at the carrier signal, respectively. While out- the neuropotentials in the time domain. The demodulated
putting the 2.4 GHz carrier, the signal generator also produces neuropotentials are then visualized via an oscilloscope.
a 4.8 GHz harmonic which is, in turn, associated with its own
phase noise. Unfortunately, this latter phase noise interferes III. R ESULTS
with the demodulation process and has not been accounted for
to date. To improve the interrogator’s phase noise performance, A. Stand-Alone Circuit Performance
the design in Fig. 1 is proposed. Compared to the previously As a rst step, performance of the implanted circuit is tested
employed interrogator system [3], two extra bandpass lters in a stand-alone wired conguration. That is, the implanted
are added between the circulator and the splitter. Both lters antenna is not considered in the design, but rather the im-
are centered at 2.4 GHz and are used to suppress the 4.8 planted circuit is directly connected to a circulator, as shown

2469-7249 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JERM.2019.2904024, IEEE Journal
of Electromagnetics, RF and Microwaves in Medicine and Biology
IEEE JOURNAL OF ELECTROMAGNETICS, RF AND MICROWAVES IN MEDICINE AND BIOLOGY 5

Fig. 9. The demodulated time domain 100 Hz signal of (a) circuit alone Vin
= 100 µVpp , (b) with the antenna in air Vin = 100 µVpp , and (c) with the
antenna in pig skin Vin = 200 µVpp in series with a 33 kΩ resistor

impedance of 33 kΩ, the sensitivity of the neurosensing system

in [3] is 50 times lower.

B. Integrated System Performance

Performance of the complete wireless system is then vali-
dated in free space (Fig. 8(b)) and via a tissue-emulating model
(Fig. 8(c)). In both cases, the implant circuit of Fig. 8(a) is
attached to an antenna that serves as the wireless interface
between the neuro-sensor and the interrogator. Here, the patch
antenna design reported in [3] is considered, which exhibits
dual-band resonances at 2.4/4.8 GHz and a footprint of 40
mm x 40 mm. Again, sensor miniaturization falls outside the
scope of this work, yet can be readily performed using already
available techniques [3]. The interrogator antenna follows the
design in [4], while the overall system layout follows the
design of Fig. 1.
Referring to Fig. 8(b), the implanted and interrogator an-
tennas are placed in free-space with a distance of ∼ 0.1 mm
between the two. Results show that neural signals as low as
Fig. 8. Fabricated BJT Circuit (a) directly connected to circulator, (b)
100 µVpp can be retrieved for an electrode impedance of 33
connected to antenna in air, and (c) connected to antenna in pig skin kΩ. An example demodulated waveform at 100 Hz is shown in
Fig. 9 (dotted/black), at the minimum detectable level of 100
µVpp . Referring to Fig. 8(c), the implanted antenna is placed
in Fig. 8(a). Here, a proof-of-concept circuit, 56.28 mm × under a 2 mm-thick layer of pig skin. The permittivity and loss
28.49 mm in size, is considered. The circuit is fabricated on tangent of the pig skin are measured using the Agilent 85070E
Rogers RO4003C substrate (ǫr = 3.38, tanδ = 0.0021) of Dielectric Probe Kit and further compared vs. the theoretical
thickness 32 mils (0.813 mm). Miniaturization is outside the skin properties [21] shown in Fig. 10. Results in this case show
scope of this particular work, yet can be readily performed via that neural signals as low as 200 µVpp can be retrieved for
techniques explored in the past [3]. Referring to Fig. 8(a), the an electrode impedance of 33 kΩ. An example demodulated
2.4 GHz carrier is set as the input to port 1 of the circulator and waveform at 100 Hz is shown in Fig. 9 (solid/blue) at the
is routed directly to the implant through port 2. The implant minimum detectable level of 200 µVpp . This slight degradation
mixes the carrier with the emulated neural signals and outputs in performance is expected given the losses associated with
the 4.8 GHz ± fneuro product to port 2 of the circulator. The biological tissues. Indeed, Fig. 11 compares the transmission
latter signal is routed to port 3 of the circulator where it is, coefcient between the two antennas in free space and with pig
eventually, demodulated and plotted in the time domain. skin used as a separation medium. As seen, the transmission
Results show that neural signals as low as 100 µVpp coefcient degrades by ∼ 3dB at 2.4 GHz and by ∼ 8dB at 4.8
can be retrieved for an electrode impedance of 33 kΩ. An GHz. As mentioned in Section III.A, higher signal levels result
example demodulated waveform at 100 Hz is shown in Fig. 9 in less noisy waveforms, while smaller electrode impedances
(dashed/red), at the minimum detectable level of 100 µVpp . result in improved sensitivity. For comparison, and assuming
Expectedly, higher signal levels result in less noisy wave- the same electrode impedance of 33 kΩ, the sensitivity of
forms, while smaller electrode impedances result in improved the neurosensing system in [3] is 25 times lower. Specic
sensitivity. For comparison, and assuming the same electrode Absorption Rate (SAR) simulations are also performed for the

2469-7249 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JERM.2019.2904024, IEEE Journal
of Electromagnetics, RF and Microwaves in Medicine and Biology
IEEE JOURNAL OF ELECTROMAGNETICS, RF AND MICROWAVES IN MEDICINE AND BIOLOGY 6

Fig. 12. Demodulated waveform of a 100 µVpp and 100 Hz signal subject
to: (a) 0 V offset, (b) +50 mV offset, and (c) -50 mV offset in series with a
33 kΩ resistor

function generator this time provides the emulated neuropo-

tentials as well as an unwanted DC offset voltage. Referring
to Fig. 12, demodulated waveforms are presented for a 100
µVpp and 100 Hz neural signal subject to : 0 V (solid/red),
+50 mV (dashed/black) and -50 mV (dotted/blue) DC offset.
As seen, the proposed implant can tolerate even the most
extreme ± 50 mV DC offset at the minimum detectable level
of 100 µVpp .

IV. C ONCLUSION
Fig. 10. a) Measured permittivity, and (b) measured conductivity of pig skin
versus the reference skin properties reported in [21]. In this work, we proposed a wireless and batteryless brain
implant that is matched to the high impedance of recording
electrodes as well as resistant to DC offset. Experimental
results show that neuropotential detection sensitivity is im-
proved by 25 times vs. the state-of-the-art at an example
electrode impedance of 33 kΩ. Notably, impedance measure-
ments of macroelectrodes using 3-lead electrode impedance
spectroscopy indicate that this resistance covers the range
anticipated in clinical settings. To date, signals as small as
100 µVpp (in free space) and 200 µVpp (in-vitro, using pig
skin) can be captured at a worst-case scenario of 33 kΩ
electrode impedance. This implies that the system can monitor
all neural spikes and most of the local eld potentials (LFPs)
in real-world settings, viz. at high impedances. By contrast,
previous wireless and batteryless implants neglected this high
impedance and rather matched the implants to a 50 Ω function
generator that emulated brain activity. As such, the proposed
Fig. 11. Measured transmission coefcient (S21 ) of the implanted and
approach is transformational for fully-passive and wireless
interrogator antenna system (a) through air, and (b) through pig skin. neuropotential acquisition in clinical settings. Future work will
focus on: a) implant miniaturization using high-permittivity
substrates and stacking approaches (antenna stacked upon the
10-cm-radius spherical head model of [3]. Results indicate that circuit), and b) in-vivo testing in animals.
SAR averaged over 1g of tissue equals 0.862 W/kg (at 10 dBm
power). This value conforms to the strictest FCC requirements
of SAR1g < 1.6 W/kg for uncontrolled environment exposure ACKNOWLEDGMENT
[22].
The authors would like to thank Prof. Liang Guo in
The Ohio State University ECE department for electrode
C. DC Offset Tolerance impedance modeling and Prof. Vibhor Krishna in The Ohio
To verify the DC offset tolerance of the implant, the State University Wexner Center for providing the clinical
measurement setup of Fig. 8(a) is adopted. In this case, the electrode.

2469-7249 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JERM.2019.2904024, IEEE Journal
of Electromagnetics, RF and Microwaves in Medicine and Biology
IEEE JOURNAL OF ELECTROMAGNETICS, RF AND MICROWAVES IN MEDICINE AND BIOLOGY 7

R EFERENCES Wei-Chuan Chen (S’17) received the B.S. degree

in electrical engineering from the National Taiwan
[1] C. W. Lee, A. Kiourti, J. Chae, and J. L. Volakis, “A high-sensitivity University, Taipei, Taiwan in 2014. He is currently
fully passive neurosensing system for wireless brain signal monitoring,” working toward the Ph.D. degree at The Ohio State
IEEE Transactions on Microwave Theory and Techniques, vol. 63, no. 6, University, Columbus OH, USA from 2015.
pp. 2060–2068, 2015. His research interests include antenna design,
[2] A. Kiourti, C. W. Lee, J. Chae, and J. L. Volakis, “A wireless fully pas- computational electromagnetic, RF/microwave cir-
sive neural recording device for unobtrusive neuropotential monitoring.” cuit and its applications in medical.
IEEE Trans. Biomed. Engineering, vol. 63, no. 1, pp. 131–137, 2016.
[3] C. W. Lee, A. Kiourti, and J. L. Volakis, “Miniaturized fully passive
brain implant for wireless neuropotential acquisition,” IEEE Antennas
and Wireless Propagation Letters, vol. 16, pp. 645–648, 2017.
[4] W. Chen, C. W. L. Lee, A. Kiourti, and J. L. Volakis, “A multi-
channel passive brain implant for wireless neuropotential monitoring,”
IEEE Journal of Electromagnetics, RF and Microwaves in Medicine and
Biology, vol. 2, no. 4, pp. 262–269, Dec 2018.
[5] H. N. Schwerdt, W. Xu, S. Shekhar, A. Abbaspour-Tamijani, B. C. Towe,
F. A. Miranda, and J. Chae, “A fully passive wireless microsystem for
recording of neuropotentials using rf backscattering methods,” Journal
of Microelectromechanical Systems, vol. 20, no. 5, pp. 1119–1130, 2011.
[6] H. N. Schwerdt, F. A. Miranda, and J. Chae, “Wireless fully pas-
sive multichannel recording of neuropotentials using photo-activated rf
backscattering methods,” IEEE Transactions on Microwave Theory and Katrina Guido received a B.S. degree in physics
Techniques, vol. 63, no. 9, pp. 2965–2970, 2015. from Stevens Institute of Technology, Hoboken, NJ
[7] G. Schalk and E. C. Leuthardt, “Brain-computer interfaces using elec- in 2018. She is currently working toward a doctorate
trocorticographic signals,” IEEE Reviews in Biomedical Engineering, in electrical engineering at The Ohio State Univer-
vol. 4, pp. 140–154, 2011. sity, Columbus, OH.
[8] K. D. Wise, A. M. Sodagar, Y. Yao, M. N. Gulari, G. E. Perlin, and Her research interests include medical sensing,
K. Naja, “Microelectrodes, microelectronics, and implantable neural antennas for medical applications, and bioelectro-
microsystems,” Proceedings of the IEEE, vol. 96, no. 7, pp. 1184–1202, magnetics.
July 2008.
[9] R. R. Harrison, “The design of integrated circuits to observe brain
activity,” Proceedings of the IEEE, vol. 96, no. 7, pp. 1203–1216, 2008.
[10] R. Muller, H.-P. Le, W. Li, P. Ledochowitsch, S. Gambini, T. Bjorninen,
A. Koralek, J. M. Carmena, M. M. Maharbiz, E. Alon et al., “A
minimally invasive 64-channel wireless µecog implant,” IEEE Journal
of Solid-State Circuits, vol. 50, no. 1, pp. 344–359, 2015.
[11] H. Ando, K. Takizawa, T. Yoshida, K. Matsushita, M. Hirata, and
T. Suzuki, “Wireless multichannel neural recording with a 128-mbps
uwb transmitter for an implantable brain-machine interfaces,” IEEE
transactions on biomedical circuits and systems, vol. 10, no. 6, pp.
1068–1078, 2016.
[12] S. Kim, P. Tathireddy, R. A. Normann, and F. Solzbacher, “Thermal
impact of an active 3-d microelectrode array implanted in the brain,”
IEEE Transactions on Neural Systems and Rehabilitation Engineering,
Asimina Kiourti (S’10–M’14) received the
vol. 15, no. 4, pp. 493–501, Dec 2007.
Diploma degree in electrical and computer
[13] E. Bharucha, H. Sepehrian, and B. Gosselin, “A survey of neural front
engineering from the University of Patras, Patras,
end ampliers and their requirements toward practical neural interfaces,”
Greece, in 2008, the M.Sc. degree in technologies
2014.
for broadband communications from University
[14] L. A. Geddes, “Historical evolution of circuit models for the electrode-
College London, London, U.K., in 2009, and the
electrolyte interface,” Annals of Biomedical Engineering, vol. 25, no. 1,
Ph.D. degree in electrical and computer engineering
p. 1, Jan 1997. [Online]. Available: https://doi.org/10.1007/BF02738534
from the National Technical University of Athens,
[15] W. Franks, I. Schenker, P. Schmutz, and A. Hierlemann, “Impedance
Athens, Greece, in 2013.
characterization and modeling of electrodes for biomedical applications,”
Dr. Kiourti is currently an Assistant Professor of
IEEE Transactions on Biomedical Engineering, vol. 52, no. 7, pp. 1295–
Electrical and Computer Engineering at The Ohio
1302, July 2005.
State University and the ElectroScience Laboratory, Columbus, OH, USA.
[16] M. J. Nelson, P. Pouget, E. A. Nilsen, C. D. Patten, and J. D. Schall,
From 2013 to 2016 she served as a Post-Doctoral Researcher and then a
“Review of signal distortion through metal microelectrode recording
Senior Research Associate at the ElectroScience Laboratory. During her
circuits and lters.” Journal of neuroscience methods, vol. 169 1, pp.
career, she has (co-)authored over 35 journal papers, 80 conference papers,
141–57, 2008.
and 7 book chapters. Her research interests include wearable and implantable
[17] FHC electrode description. [Online]. Available: https://www.fh-co.com/
sensors, antennas and electromagnetics for body area applications, and
[18] A. Bagheri, M. T. Salam, J. L. P. Velazquez, and R. Genov, “Low-
exible textile-based electronics.
frequency noise and offset rejection in dc-coupled neural ampliers:
Dr. Kiourti has received several awards and scholarships, including
A review and digitally-assisted design tutorial,” IEEE Transactions on
the IEEE Engineering inMedicine and Biology Society (EMB-S) Young
Biomedical Circuits and Systems, vol. 11, no. 1, pp. 161–176, Feb 2017.
Investigator Award for 2014, the IEEE Microwave Theory and Techniques
[19] R. R. Harrison, “A low-power, low-noise cmos amplier for neural
Society (MTTS) Graduate Fellowship for Medical Applications for 2012,
recording applications,” in 2002 IEEE International Symposium on
and the IEEE Antennas and Propagation Society (AP-S) Doctoral Research
Circuits and Systems. Proceedings (Cat. No.02CH37353), vol. 5, May
Award for 2011. She is currently serving as an Associate Editor for the IEEE
2002, pp. V–V.
Transactions on Antennas and Propagation.
[20] J. Ruiz-Amaya, A. Rodriguez-Perez, and M. Delgado-Restituto, “A low
noise amplier for neural spike recording interfaces,” Sensors, vol. 15,
pp. 25 313–25 335, 09 2015.
[21] S. Gabriel, R. Lau, and C. Gabriel, “The dielectric properties of
biological tissues: Ii. measurements in the frequency range 10 hz to
20 ghz,” Physics in medicine & biology, vol. 41, no. 11, p. 2251, 1996.
[22] U. F. C. Commission et al., “Report and order fcc 96–326,” 1996.

2469-7249 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.