WIRELESS ELECTROENCEPHALOGRAPHY (EEG) IN MICE MODEL

1
 Section 1: Introduction and Literature review
1.1 Introduction:
The human brain is one of the most complex systems in the universe. There are various
technologies to record brain activities in the form of waves and electroencephalography
(EEG) is one of them. EEG is one of the brain signal processing techniques that allows
gaining the understanding of the complex inner mechanisms of the brain and abnormal brain
waves that have shown to be associated with brain disorders.(Gurumurthy et al., 2013)

Electroencephalographyand related methodologies help in predicting the likelihood that novel

therapies and compounds will exhibit clinical efficacy early in preclinical development. These
analysis, including quantitative EEG (e.g., brain mapping) and evoked/event-related potentials
(EP/ERP), can provide a physiological endpoint that may be used to facilitate drug discovery,
optimize lead or candidate compound selection, as well as afford patient stratification and
Go/No-Go decisions in clinical trials(Leiser et al., 2011). EEG recordings are conducted in
mice to understand neural activity and how it relates to specific behaviours or cognitive tasks
(e.g., memory, learning, and decision-making) that are relevant to understandbrain disorders
such as epilepsy, schizophrenia and Alzheimer’s disease.

The EEGLAB (a toolbox of MATLAB)environment for electrophysiological signal

processing is the centre of a growing framework of tools developed and released by
researchers at the Swartz Centre for Computational Neuroscience (SCCN) at UCSD. These
include software for synchronized multimodal recording (SNAP, LSL, XDF), MoBILAB (an
open source toolbox for analysis and visualization of mobile brain/body imaging data), the
HeadIT data, and a growing set of toolboxes that operate as EEGLAB plug-ins like BCILAB.
MEEG is a new plug-in developed by the authors for analysis of MEG and MEEG
(synchronized MEG plus EEG) data.

EEG signals contain more relevant information about brain disorders and different type of
signals generated by the activation of particular regions of the brain. Signals in the form of

2
 dataset are already loaded to the tool so we will be using that signals to plot the data and
visualization of the time-frequency domain plots which can be displayed all together.

1.2 Wireless EEG Recording System:

For an EEG-based Web Based Sensory Network, the EEG sensors are electrodes placed on a
person’s head, usuallyfollowing an international convention (e.g., the international 10–20
system). An EEG sensor is alsoreferred to as the EEG channel. The number of sensors
depends on the application: some systemsrequire few electrodes, while others require a few
hundred. Every sensor is wired to a single centralmicroprocessor unit that usually has three
main components: a buffer (to store the EEG data streamcoming from the different EEG
channels; this buffer acts as memory), the microprocessor itself (tocarry out computations
needed before transmission) and a low-power radio (to wirelessly transmit thedata). The
combination of the EEG sensors and the microprocessor unit is referred to as the sensornode.
This sensor node is battery powered. The sensor node transmits the EEG data to the server
nodewirelessly. The server node is comprised of two main blocks: a low power radio receiver
(to receivethe transmitted EEG data) and a computing resource (to carry out any post-
transmission computations,storage and any other desired operations). We assume that there is
no constraint on the energy supply orthe computational power at this server node.(Fauvel &
Ward, 2014)

Figure 1: Block diagram of wireless EEG recording system

1.3 EEG recording system in Mice:

A wireless portable EEG recording system for animals is designed,
manufactured and then tested in animal. The system consists of four
modules: 1) EEG collecting module with the wireless transmitter and
receiver; 2) Preamplifier, band pass filter and 50 Hz trapper, 3) power
management module and 4) display interface for showing EEG signals. The
3
 EEG data are modulated first and then emitted by the wireless transmitter
after being amplified and filtered. The receiver demodulates and displays
the signals in voltage through serial port. The system is designed as SMD
with small size and lightweight and is fabricated of electronic components
that are commercially available. The measuring electrodes is implanted on
the skull and the acquisition transmitter was strapped or placed
underneath the skin on the back of the mouse.(Dian Zhang et al., 2011)

Figure 2: Wireless EEG recording system for Mice

1.4 EEG Signal Analysis

EEGLAB is a toolbox and GUI, running under the cross-platform MATLAB environment
(The MathWorks, Inc.) for processing collections of single-trial and averaged EEG data of
any number of channels. Available functions include EEG data, channel and event
information importing, data visualization, pre-processing that includeartefact (eye movement,
skin etc) rejection, filtering, epoch selection, and averaging, ICA and time/frequency
decompositions including channel and component cross-coherence supported by bootstrap
statistical methods based on data resampling.
The first thing we need is some raw EEG data to process. These often include the application
of filters, such as a high-pass filter to remove the DC components of the signals and the drifts
(usually a frequency cut-off of 1 Hz is enough). A low pass filter can also be applied to
remove the high frequency components. In EEG, we currently rarely study frequencies above
90 Hz, which correspond to the Gamma range. Once our signals are clean, i.e., pre-processed,
it is quite common to cut them in epochs of a few seconds and then extract features out of
each one of these. This allows us to have many features from a single EEG recording, which
is always good when performing statistics or when applying classifiers.

1.5 Signals and its Characteristics

4
EEG waveforms are classified according to their frequency, amplitude, shape and position of
the electrodes on the scalp. Frequency (Hz) is the basic unit used to determine normal or
abnormal rhythms. The classification of waveforms based on signal frequency: alpha, beta,
theta, delta and gamma. Shape, head distribution and symmetry property of waves are also
the basis of recognition. Certain pattern of waveform is normal at specific age, state of
alertness and sleep. Frequency bands categorize the continuous rhythms of the brain or brain
waves. Brain wave frequency differs corresponds to different behavior and mental states of
the brain.

Delta (0.1 - 4 Hz)

Delta wave is highest in amplitude and slowest wave. It represents the grey matter of the
brain. This brain wave activity is found in all stages of sleep. It is normal, rarely observed by
experienced mediators and is dominant rhythms in infants. Delta waves are abnormal for
adults who are awake and it induces GH.

Figure 3: Delta Waveform

Theta (4 – 8 Hz)
Theta waves are related to subconscious activity. They are observed in deep relaxation and
meditation. Psychotics can gather awareness information by using this level. It is abnormal in
adults but itis normal for children under 13 years. It encourages the production of GH,
serotonin hormone that increase relaxation and get relief from pain, cortical hormone that
helps for memory and learning.

Figure 4: Theta Waveform

Alpha (8 - 13 Hz)
Alpha waves represents white matter of the brain. Seen in all age groups commonly found in
adults who is awake but relaxed with closed eyes. It occurs on both sides of the head but

5
slightly higher amplitude in non-dominant side and recorded from the occipital and parietal
regions of the brain. It acts as a bridge between conscious and subconscious mind. It induces
the production of serotonin, a chemical reactor that increases relaxation and relief from pain.
An abnormal case is alpha coma, caused by hypoxic ischemic encephalopathy of destructive
process in the pons.

Figure 5: Alpha Waveform

Beta (13 - 30 Hz)

Beta waves are concerned with behavior and actions. It is related to the sense of what we see,
touch, hear, smell and taste. It is usually seen in both sides of frontal and parietal lobes. Beta
waves induces the production of Cortisol, which is the major age accelerating hormone within
the brain and interface with learning and memory. They occur in conscious state like talking,
problem solving, judgment, and decision-making.

Figure 6: Beta Waveform

Gamma (30-100 Hz)

Gamma waves are associated with perception and consciousness. This wave is analysed after
the development of digital EEG as analog is restricted to record and measure rhythms lesser
than frequency of 25 Hz. These waves are occurred during hyper alertness and integration of
sensory inputs. Gamma properly combines the senses and memory together for an ultimate
experience.(Kumar & Bhuvaneswari, 2012)

Figure 7: Gamma Waveform

6
 Table 1 : Brain waves and their characteristic

Types Frequency State Hormone Location

Delta 0-4 Deep sleep GH, Melatonin Frontally in adults, posteriorly
in children
Theta 4-8 Deeply relaxed Serotonin, Acetylcholine, Anti- Thalamic Region
cortisol, Endorphins, GH
Alpha 8-13 Day dreaming, Serotonin, Acetylcholine, Posterior regions
calm Endorphins,
Beta 13-30 Active thinking, Adrenaline, Cortisol, Frontal and Parietal
alert Norepinephrine, Dopamine
Gamma 30-100 Concentrated, Serotonin, Endorphins Somatosensory cortex
focused

7
 Section 2: Methodology
2.1 EEG Data Recording in Mice:
The ketamine-induced changes in the cortical biopotentials were recorded using a wireless
EEG system. For EEG transmitter implantation, mice were deeply anesthetized using the
cocktail of ketamine (90 mg/kg) and xylazine (10 mg/kg) administered via intraperitoneal
route. Mice were kept on a heating pad (37°C) during the whole microsurgical procedure. A
midline scalp incision was made from the eyes to the neck to expose the skull. Two tiny burr
holes, 1 at each side, were trephined manually on the skull over the parietal lobe of both the
hemisphere about 1.5 mm posterior to bregma and 2.0 mm lateral to the sagittal suture. Two
screws were placed in the holes while touching the surface of the dura and connected to the
transmitter electrodes. Finally, the assembly was secured to the parietal bone using
dental acrylic cement and topical tissue adhesive (n-butyl-2-cyanoacrylate).
The transmitter was placed subcutaneously at the back of the mice. After the surgery, mice
were allowed to recover for 7 days before the acquisition of EEG data. EEG data acquisition
was made through a wireless stellar telemetry system. EEG data were recorded continuously
in packets of 10 seconds at the time interval of 5 minutes for 30 minutes at the rate of 200
samples per second before (baseline) and after INK administration. (Goswami et al., 2021)

2.2 EEG Signal Processing:

2.2.1 Start EEGLAB
Type eeglab at the MATLAB command line and hit enter. EEGLAB will automatically
add itself to the MATLAB path.

8
 Figure 8: EEGLAB Start Page

2.2.2 Load or import file

Import EDF file of the dataset by selecting “From EDF/EDF+/GDF files (BIOSIG
toolbox)”

Figure 9: Import file

2.2.3 Save the Required data from the EDF file

The first EEG channel was taken as animal1, and the second EEG channel was taken as
animal2 and setting the sample frequency to 200 Hz.

2.2.4 Plot the EEG signal

The EEG signal of animal1 and animal2 was plotted with respect to time.

9
2.2.5 PSD of Raw data
Create a PSD for the raw animal1 and animal2 data.

2.2.6 Open filter designer app and design a filter

Type filterDesigner in the Command Window and design a band pass filter for all the
brain [Link] create a filter for as follows:
a. Delta (0.1 - 4 Hz)
b. Theta (4 – 8 Hz)
c. Alpha (8 - 13 Hz)
d. Beta (13 - 30 Hz)
e. Gamma (30-100 Hz)

10
 Figure 100: Filter Designing App

2.2.7 Export the filter as object to the workspace.

Figure 11: Export designed filter to workspace

2.2.8 Create a filtered data

Using the following, create a filtered data for each filter of animal1 and animal2.

11
2.2.9 Create the epoch of the data
Using the following code, create the epoch for the filtered data. Similarly do the same for
the rest of the brain waves (filters).

2.2.10 PSD of filtered data

Using the following code, create PSD of the filtered data. Similarly do the same for the
rest of the brain waves (filters).

12
 Section 3: Result and Discussions
The visualization of EEG data was done. The first two data was selected as animal1 and
animal2 and further preprocessing of data was done on that. The waveforms below the
second wave is the calculated values from the from the recorded EEG signal.

Figure 12: Channel activity from EEGLAB

The amplitude vs time was plotted of both the animal1 and animal2. This is done to
Visualize the EEG data with respect to time.

13
Figure 113: Amplitude vs Time graph for animal1 and animal2

Figure 14: Amplitude vs time graph for Delta wave

Figure 125: Amplitude vs time graph for Theta wave

Figure 1613: Amplitude vs time graph for Alpha wave

14
 Figure 1714: Amplitude vs time graph for Beta wave

Figure 18: Amplitude vs time graph for Gamma wave

Power spectral density specifies the power levels of the frequency components present in a signal.
It is denoted as PSD inshort. The PSD specifies the power of various frequencies present in the
signal and we can determine the range of power over which the signal frequencies are operating
at. Essentially, the PSD profile is a plot of the power over frequency. Using these graphs the
frequency at which maximum power of the specific brain waves occur and the range of PSD can
be depicted.

Figure 15: Mean Power vs frequency graph of raw data

15
Figure 20: Mean Power vs frequency graph for Delta wave

Figure 21: Mean Power vs frequency graph for Theta wave

Figure 22: Mean Power vs frequency graph for Alpha wave

Figure 23: Mean Power vs frequency graph for Beta wave

16
 Figure 2416: Mean Power vs frequency graph for Gamma wave

Section 4: Conclusion
In the present study, EEG waves (both filtered and unfiltered) data were analysed using using
filter designer app from MATLAB. The cortical biopotentials were recorded in mice using
wireless telemetric system. Data was analysed for amplitude and maximum mean power at a
specific frequency range belonging to various filtered waveforms. Epoch has been created
using the filters exported to the workspace and PSD graph was plotted. The PSD graph
depicted the frequency at which the maximum power of the wave was seen. We have
observed mean power vs frequency is a better representation of filtered biopotentials.
Although, analysis of changes in amplitude can be of great significance while recording
evoked/event-related potentials (EP/ERP).
Our telemetric EEG recording system offers several advantages over other tethered recording
systems. The light weighted transmitter is fully implantable due to its small size and cause
minimal behavioral impairment or discomfort. In addition, its ‘radio wave telemetery’
permits the range of the signal to longer distance (up to 5 m) between emitter and receiver,
so data was recorded without any loss of signal and disturbance to the mice. This allows full
flexibility of experimental designs. A higher acquisition and transmission rate makes this
setup even more powerful. The data collected from individual mice is reliable. Thus, we
17
conclude that telemetry system data analysis using MATLAB will facilitate future work
directed at understanding the changes in cortical biopotentials for various neurological
disorder and other tasks related research involving changes in neuronal activities.

Section 5: Reference and Bibliographies

1. Dian Zhang, Xin Gao, & Dong Chuan Yu. (2011). A wireless electroencephalogram
telemetry recording system for roaming animals. 2011 International Conference on
Applied Superconductivity and Electromagnetic Devices, 192–195.
[Link]

2. Fauvel, S., & Ward, R. (2014). An Energy Efficient Compressed Sensing Framework for
the Compression of Electroencephalogram Signals. Sensors, 14(1), 1474–1496.
[Link]

3. Goswami, N., Aleem, M., & Manda, K. (2021). Intranasal Ketamine for Acute Pain:
Behavioral and Neurophysiological Safety Analysis in Mice. Current Therapeutic
Research, 94, 100627. [Link]

4. Gurumurthy, S., Mahit, V., & Ghosh, R. (2013). Analysis and Simulation of Brain Signal
Data by EEG Signal Processing Technique using MATLAB. International Journal of
Engineering and Technology, 5, 2771–2776.

5. Kumar, J. S., & Bhuvaneswari, P. (2012). Analysis of Electroencephalography (EEG)

Signals and Its Categorization–A Study. Procedia Engineering, 38, 2525–2536.
[Link]

6. Leiser, S. C., Dunlop, J., Bowlby, M. R., & Devilbiss, D. M. (2011). Aligning strategies
for using EEG as a surrogate biomarker: A review of preclinical and clinical research.
18
Biochemical Pharmacology, 81(12), 1408–1421.
[Link]

19