Resorbable PLGA Microneedles to Insert Ultra-fine Electrode Arrays
in Neural Tissue for Chronic Recording
Frederik Ceyssens
1
, Marta Bovet Carmona
2
, Dries Kil
1
, Marjolijn Deprez
3
, Bart Nuttin
3
,
Aya Takeoka
4
, Detlef Balschun
2
and Robert Puers
1
1
ESAT-MICAS, KULeuven, Kasteelpark Arenberg 10, Leuven, Belgium
²Dept. of Psychology, KULeuven, Leuven, Belgium
3
Experimental Neurosurgery and Neuroanatomy, KULeuven, Leuven, Belgium
4
NERF, Leuven, Belgium
1 OBJECTIVES
It has been shown that the mechanical rigidity of
neural implants is a key factor that causes the
formation of scar tissue around the implant. Clearly,
this reduces performance and lifetime. Therefore,
recent work has focused on very compliant,
polymer-based implants (Weltman, 2016). However,
such implants need a temporary reinforcement to aid
their insertion in neural tissue (Lecomte 2018).
Recent work has even indicated that it is possible
to keep polymer thin-film neural electrode arrays in
close contact with neural tissue over chronic
timescales, without the formation of scar tissue
(Zhou 2017). To achieve this, injection of an
electrode array suspended in liquid was used, which
is hard to upscale to higher electrode counts and
relatively cumbersome.
In this work, we are investigating the use of
microneedles fabricated out of short-chain, fast
resorbing polylactic-co-glycolic acid (PLGA,
Purasorb PLDG 5002A) as a temporary
reinforcement.
This technique improves the existing arrays,
injected using a capillary, in terms of controllability
and upscalability to larger electrode counts.
Therefore, the objective was to device a
fabrication process, that allows to micromachine
needle-shaped PLGA structures and to embed ultra
fine electrode arrays in those needles. A second
objective was the long-term in vivo testing of the
electrode arrays in rats.
2 METHODS
The electrode arrays are fabricated by a lithography
based processing technology published earlier
(Ceyssens, 2015). The process was adapted to
reduce the implant thickness to only 1 µm, yielding
an ultra-flexible implant backbone.
The resulting arrays were designed to contain a
linear array of 16 iridium oxide electrodes, aimed at
single neuron recording. The electrodes are 15
micrometer in diameter. Polyimide
(HDMicrosystems PI2611) is used as insulation
material. After fabrication, an Omnetics Nano
connector is attached to connect an external
amplifier during testing. The micromachined wires
between the connector and the electrodes are only 10
µm wide.
Separate microneedles for support are fabricated
out of a short chain PLGA, that resorbs over a period
of 2-3 weeks after implantation. Molding or
picosecond UV laser machining is used. The needles
have a cross-section of 0.35 x 0.25 mm². In a final
step, the needles are bonded to the electrode array
using thermocompression.
For in vitro testing, the needle arrays were
implanted transdurally in the right sensorimotor
cortex of four rats, 4 mm right from Bregma. At
intervals, at least one week apart, the rats were
anesthetized using usoflurane. The spontaneous
spiking activity was recorded. In a second test,
biphasic electric pulses (800 µA amplitude, 0.1 ms
per phase) were applied subcutaneously to the left
forepaw to evoke a muscle contraction. Meanwhile,
the presence and strength of the evoked potential in
the brain was monitored.
6
Ceyssens, F., Carmona, M., Kil, D., Deprez, M., Nuttin, B., Takeoka, A., Balschun, D. and Puers, R.
Resorbable PLGA Microneedles to Insert Ultra-fine Electrode Arrays in Neural Tissue for Chronic Recording.
In Extended Abstracts (NEUROTECHNIX 2018), pages 6-9
Copyright © 2018 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
To quantify the electric recordings, the results
were filtered (FIR HP filter, 200 Hz stopband, 250
Hz passband) and a spike counting algorithm based
on a moving window was adopted. Spike threshold
was set to 4 standard deviations below the window
average. To quantify the evoked potential, a FIR HP
filter with 10 Hz stopband was used and post-
stimulation peak-to-peak (PTP) values were
recorded. Also, the RMS ratio after/before
stimulation was determined.
After approximately four months, the rats were
sacrificed and a histological evaluation of the
implantation sites was performed (GFAP and NeuN
staining).
3 RESULTS
3.1 Fabrication and in Vitro Test
The fabrication process yielded about 50% devices
without defects (Figure 1: example from early
fabrication run). The arrays are 1 µm thick in total,
and the metal tracks are 4 µm wide, with 3.5 µm of
insulating PI on every side. An example needle is
shown in Figure 1. The measured electrode
impedance was around 200 kOhm at 1 kHz.
As an insertion test in Agar gel proved
insufficient adhesion between the implant backbone,
a second version was fabricated in which the
backbone was squeezed between two PLGA needles
of half the thickness. This proved to have sufficient
adhesion.
Figure 1: PLGA needle with electrodes.
3.2 In Vivo Testing
The implantations (Figure 2) went as planned. No
wound inflammation or abnormal behaviour of the
animals was observed.
Figure 2: Implantation, showing transparent PLGA
microneedle in burr hole (lower left), connecting cable and
Omnetics connector (upper right).
About 50 days after implantation, the headstage
of rat 3 disconnected from the skull, after which no
measurements were possible on the animal. In all
other animals, it was possible to observe signals
throughout the duration of the experiment.
In general, action potentials and evoked
potentials could be clearly observed over the entire
course of the 4 month time span. There were no
channels dropping out. A typical measurement of
spontaneous action potentials (4 months after the
start of the experiment) is shown in Figure 3.
Figure 3: Sample of observed spontaneous spikes (rat 4, 4
months after implantation). Red crosses show peaks
detected by the algorithm used.
Figure 4 shows an example of an evoked
potential recording, also after 4 months.
The number of observed spontaneous spikes per
second, and the size of the evoked potential are
shown in Figure 5. The average PTP value (over all
channels) is around 2 mV, and about 70 spikes per
second can be seen per channel. Especially the latter
measurement shows a relatively large variation,
though.
Resorbable PLGA Microneedles to Insert Ultra-fine Electrode Arrays in Neural Tissue for Chronic Recording
7
Figure 4: Example of evoked potential recording (rat 4,
channel 11, 4 months after implantation). The largest peak
is the stimulation artifact, which is followed by a train of
relatively low-frequency responses.
3.3 Histology
Due to the shallow insertion depth, it was not
possible to remove the rat brains for histology
without pulling out the implant, though it was still
possible to inspect the glial scar.
Histology revealed that there was still limited
glial scarring present. The remaining scar is a factor
4 6 times smaller in cross sectional area than the
original size of the PLGA needle. An example is
given in figure 6. In later work, a full analysis of all
histological data will be included.
4 DISCUSSION
Briefly, though no full integration (i.e. the electrode
array floating in between neurons without any scar
in between) with the neural tissue was achieved, the
observed astrocytic scar was minimal. Viable
neurons were seen to be present inside the volume
that used to be taken up by the resorbable structure.
Qualitatively, the measurements were stable and
a similar response was seen over the four months of
the experiment.
Quantitatively, the number of spikes per second
and the peak-to-peak values of the evoked potential
was seen to vary about a factor of 2 in between
experiments. As there is no clear upwards or
downwards trend and this variation is even present
between measurements just a few days apart on the
same rat, we can assume this is likely due to inherent
variability in the experimental setup.
This includes the depth of the anesthesia, the
exact positioning of the electrode used for
stimulation and natural variations.
Figure 5: Top: measured PTP value (per rat, averaged over
all 16 channels) of the evoked potential over time. Bottom:
Number of spontaneous spikes per second.
Figure 6: Scar after end of experiment. Red: GFAP stain.
Green: NeuN stain, revealing the presence of viable
neurons up to 100 µm close the center axis of
implantation. The dashed rectangle indicates the
approximate boundary of the original PLGA needle.
5 CONCLUSION
We conclude that this method of inserting ultra-fine
electrodes arrays is practical and yields only minor
tissue damage. Combined with the polyimide-based
neural electrode array presented, we were able to
NEUROTECHNIX 2018 - 6th International Congress on Neurotechnology, Electronics and Informatics
8
record evoked potentials and action potentials for at
least four months.
ACKNOWLEDGEMENTS
We would like to thank Ester Tooten (UCL,
Belgium) and Karin Jonckers (VIB, Belgium) for
their help in obtaining these results.
Furthermore, our gratitude goes to our funding
sources: FWO-Flanders for Frederik Ceyssens’
research fellowship, KULeuven IDO fund and the
Hercules Foundation for heavy equipment (AKUL
034 and ZW1115). The research leading to these
results has received funding from the European
Research Council under the European Union’s
Seventh Framework Programme (FP7/2007-
2013)/ERC grant agreement n° 340931.
REFERENCES
Weltman, A., Yoo, J., Meng, E. 2016. Flexible,
penetrating brain probes enabled by advances in
polymer microfabrication. Micromachines, 7(10), 180
Lecomte, A., Descamps, E., Bergaud, C. 2018. A review
on mechanical considerations for chronically-
implanted neural probes. Journal of neural
engineering, 15(3), 031001..
Zhou, T., Hong, G., Fu, T. M., Yang, X., Schuhmann, T.
G., Viveros, R. D., Lieber, C. M. 2017. Syringe-
injectable mesh electronics integrate seamlessly with
minimal chronic immune response in the brain.
Proceedings of the National Academy of Sciences,
114(23), 5894-5899.
Ceyssens, F., Puers, R. 2015. Insulation lifetime
improvement of polyimide thin film neural implants."
Journal of neural engineering 12.5 054001.
Resorbable PLGA Microneedles to Insert Ultra-fine Electrode Arrays in Neural Tissue for Chronic Recording
9