Encoding of Movement in Local Field Potentials from the Wall of

Motor Cortical Lesions in Rats

Ioana Nica

, Marjolijn Deprez

, Frederik Ceyssens

, Kris van Kuyck

, Robert Puers

, Bart Nuttin

2,4

and Jean-Marie Aerts

Division Animal and Human Health Engineering, Department of Biosystems, KU Leuven, Leuven, Belgium

Research Group Experimental Neurosurgery and Neuroanatomy, KU Leuven, Leuven, Belgium

Division Microelectronics and Sensors (MICAS), Department of Electrical Engineering, KU Leuven, Leuven, Belgium

Department of Neurosurgery, University Hospitals Leuven, Leuven, Belgium

1 OBJECTIVES

The objective of the current study was to investigate

if electrical oscillatory activity from within the

cavity wall of a motor cortical lesion can be used as

a biomarker in decoding movement. We show

results from 3 rats with unilateral lesion in the

forelimb area of the motor cortex, for which local

field potential (LFP) spectra present significant

modulation within the frequency bands of 6-10 Hz

and 45-90 Hz, corresponding to movement episodes.

2 METHODS

Rats were anaesthetized with a mixture of ketamine

(Nimatek®) and medetomidine hydrochloride

(Narcostart®). A craniotomy over the forelimb area

of the primary motor cortex was made (coordinates:

1.5 mm posterior to 5 mm anterior to bregma, and

0.5 mm to 4.5 mm lateral to bregma), after which the

exposed brain tissue was aspirated to a depth of 1.5

mm. Three weeks later, a polyimide-based thin film

electrode array (Ceyssens et al., 2013) containing 16

platinum electrode contacts, each with a diameter of

350 µm (Fig. 1.a.) was implanted against the cavity

wall. The implant and connector (Omnetics) were

secured in place with stainless steel screws in the

skull and dental cement (Fig. 1.b.).

Open field tests were performed 1 month after

electrode implantation. The rats were placed in

metal cages (dimensions 36x38x35 cm), being free

to move. They were monitored for ~30 min with a

SONY HDR-AS15 camera (sample rate 30 Hz)

placed laterally. Brain electrical activity was

wirelessly recorded at the same time, using the W16

headstage model from Multichannel Systems (MCS

GmbH, Reutlingen, Germany), (see Fig.1.c.). The

LFPs were sampled at 10 kHz and preamplified in

the range of 1Hz to 5 kHz within the headstage.

a). b). c).

Figure 1: a). Electrode array; b). Top view of the

implantation site; c). View of the headstage.

Analyses of both video and LFP recordings were

performed offline in MATLAB (version 2014b; The

MathWorks). An automatic algorithm based on pixel

intensity was implemented to discriminate between

activity and resting intervals. The threshold was set

empirically so as to capture activity such as walking,

self-grooming, standing on hind limbs. Intervals

when no movement of the limbs could be detected

were labelled as part of the ‘resting state’. The

corresponding LFP intervals were then extracted and

divided in 2-sec epochs. On average, 225 (± 55)

epochs per rat were extracted for the active state and

335 (± 80) epochs, respectively, for the resting state.

The LFPs were low-pass filtered and down-

sampled using an equiripple FIR decimator (300 Hz

cutoff frequency, decimation factor of 10) so that

final sampling rate was 1 kHz. We computed the

power spectrum using the ‘periodogram’ Matlab

function (frequency resolution of 0.5 Hz) and we

multiplied power at each frequency bin with squared

frequency to account for the 1/f

decay specific to

brain signals (Miller et al., 2009; Buszaki et al.,

2012). A peak in power ~8 Hz and ~50 Hz was

observed on datasets from both behavioural states.

Nica, I., Deprez, M., Ceyssens, F., Kuyck, K., Puers, R., Nuttin, B. and Aerts, J..

Encoding of Movement in Local Field Potentials from the Wall of Motor Cortical Lesions in Rats.

Figure 2: Mean normalized power between significant

channels (± 95% confidence limits) for each animal.

3 RESULTS

We performed two-sample t-tests with unequal

variances to test significant differences between two

sets of features (mean power in theta (6-10 Hz) and

high-gamma (45-90 Hz) band for all 16 electrodes),

to investigate how well each feature discriminates

activity from rest. For each rat, a subset of electrodes

showed significant increases (p<0.05) in theta power

(14 electrodes for rat 1, 9 for rat 2 and 6 for rat 3).

High gamma power was significantly higher

(p<<0.01) in active state on all electrodes.

4 DISCUSSION

We report prominent theta and gamma activity in the

forelimb region of the rat motor cortex, on LFPs

recorded from the wall of a lesion. We show that

these oscillations are strongly linked to motor

behaviour state, in an open-field experimental setup

that required neither learning nor reward. Our results

corroborate previous studies reported in literature on

cortical LFPs from healthy rats, during treadmill

running (von Nicolai et al., 2014), or during reward-

motivated forelimb movement (Igarashi et al. 2013).

Since the results revealed that specific subsets of

electrodes are relevant for each subject, the

robustness of these features could be investigated in

a longitudinal test, while optimizing a decoder able

to detect activity state on a single-trial basis.

Motor cortical lesions can induce various

impairments, therefore it would be of interest to

search for a correlation between theta-gamma

activity and the type of deficit the subject exhibits.

In conclusion, our results suggest that

informative signal features can be extracted from

electrical activity generated in the wall of a motor

cortical lesion. These features should be further

investigated to test the hypothesis that LFPs can help

parametrize state of impairment, on a subject-by-

subject basis.

REFERENCES

Buzsaki, G., Anastassiou, C.A., Koch, C., 2012. The

origin of extracellular fields and currents - EEG,

ECoG, LFP and spikes. Nat. Rev. Neurosci.

13(6):407-420. doi: 10.1038/nrn3241.

Ceyssens, F., van Kuyck, K., Velde, G. V.,

Welkenhuysen, M., Stappers, L., Nuttin, B., Puers, R.,.

2013. Resorbable scaffold based chronic neural

electrode arrays. Biomedical microdevices, 15(3): 481-

493.

Igarashi J., et al., 2013. A θ-γ oscillation code for neuronal

coordination during motor behaviour. J. Neurosci.

33:18515-18530.

Miller, K. J., Sorensen, L. B., Ojemann, J. G. & den Nijs,

M. 2009. Power-law scaling in the brain surface

electric potential. PLoS Comput. Biol. 5, e1000609.

Von Nicolai, C., et al., 2014. Corticostriatal Coordination

through Coherent Phase-Amplitude Coupling. J.

Neurosci. 34(17):5938-5948.

a) Rat 3

b) Rat 2

c) Rat 1