A. spectral density(PSD) and time-frequency correlated features extraction techniques

A.   
Selection of Brain Regions

EEG signals are acquired from 21 electrodes, placing them
on the scalp of human subjects. The EEG signals are recorded on a separate
computer having 8 GB RAM with CPUclock of

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

 

Low

Medium

High

Activation level

Odor 1 (Perfume)

 

 

 

 

Odor 2 (Dettol)

 

 

 

 

Odor 3 (Acetic acid)

 

 

 

 

Odor 4 (Alchohol)

 

Fig.5. Section of brain region and
frequency bands from scalp maps of three concentration level of four stimuli

3.4 GHz.Fig. 5 shows the scalp
maps of one subject. Scalp maps of different concentration level of different
stimuli have been recorded during the experiment.
From the scalp maps, it can be observed that pre-frontal, Frontal, parietal,
and temporal lobes exhibit significant activations during the experiment. Here, the pre-frontal region is hardly activated inolfactory
sensing and processing. Besides these, frontal and mainly parietal, temporal
are found to take significantly active participation during the experiment. The
paper 6 reveals that temporal lobe and frontal lobe are highly associated
with human olfactory signal processing. Therefore, we select F3, F4,
F7, F8 and FZ (for frontal lobe) P3
and P4 (from parietal lobe), T3 and T4(from
temporal lobe) and FP1 and FP2 (for prefrontal lobe)for
extracting necessary information using signal processing techniques.Fig.6shows the electrode positions over the scalp in our
brain during experiment, where the red colored electrodes are used to collect
data. Electrodes are placed over the scalp using 10-20 electrode placement
system 26.

Fig.6. Electrode position of
our experiment

 

B.   
EEG Feature Extraction and Feature Selection

Featureextraction
is very much important for EEG classification problem for accurate decoding of
mental tasks. There is variety of EEG feature extraction schemes: time-domain
feature extraction techniques like Hjorth parameters 18, Autoregressive
parameters 19 and also frequency-domain feature extraction techniques like
power spectral density(PSD) and time-frequency correlated features extraction
techniques likediscrete wavelet transform (DWT) 20. We first plot the raw EEG
signal pattern recorded from the specific brain regions during the experiment.
Fig. 7(a)showsthe plot of raw EEG signalscaptured from the temporal lobe during
the smell stimuli presentation and Fig. 7(b)represents the filtered EEG signals
for four different odor stimuli.

Power
spectral density (PSD), is a well-known frequency-domain feature extraction
technique to extract EEG signal power distribution. PSD is applied on the
filtered EEG signal acquired from prefrontal, frontal, parietal and temporal
lobe regions. Besides PSD, It is important to mention here that filtering of
EEG signal is done by using a standard Chebyshev 21 band pass infinite
impulse response (IIR) filter of order 10, which has the pass band frequency of
0.5-13 Hz. The selection is made so because of the superior performance of
Chebyshev filter as compared to its standard counterparts including Butterworth
and Elliptic filter. Now, for each subject and each vowel sound, PSD extract
10×12×328 feature sets  (since, here,
experiment is repeated 10 times and number of selected electrodes 12). Fig.8(a)
and Fig.8(b)present the PSD features extractedfrom the above brain regions.A.   
Selection of Brain Regions

EEG signals are acquired from 21 electrodes, placing them
on the scalp of human subjects. The EEG signals are recorded on a separate
computer having 8 GB RAM with CPUclock of

 

Low

Medium

High

Activation level

Odor 1 (Perfume)

 

 

 

 

Odor 2 (Dettol)

 

 

 

 

Odor 3 (Acetic acid)

 

 

 

 

Odor 4 (Alchohol)

 

Fig.5. Section of brain region and
frequency bands from scalp maps of three concentration level of four stimuli

3.4 GHz.Fig. 5 shows the scalp
maps of one subject. Scalp maps of different concentration level of different
stimuli have been recorded during the experiment.
From the scalp maps, it can be observed that pre-frontal, Frontal, parietal,
and temporal lobes exhibit significant activations during the experiment. Here, the pre-frontal region is hardly activated inolfactory
sensing and processing. Besides these, frontal and mainly parietal, temporal
are found to take significantly active participation during the experiment. The
paper 6 reveals that temporal lobe and frontal lobe are highly associated
with human olfactory signal processing. Therefore, we select F3, F4,
F7, F8 and FZ (for frontal lobe) P3
and P4 (from parietal lobe), T3 and T4(from
temporal lobe) and FP1 and FP2 (for prefrontal lobe)for
extracting necessary information using signal processing techniques.Fig.6shows the electrode positions over the scalp in our
brain during experiment, where the red colored electrodes are used to collect
data. Electrodes are placed over the scalp using 10-20 electrode placement
system 26.

Fig.6. Electrode position of
our experiment

 

B.   
EEG Feature Extraction and Feature Selection

Featureextraction
is very much important for EEG classification problem for accurate decoding of
mental tasks. There is variety of EEG feature extraction schemes: time-domain
feature extraction techniques like Hjorth parameters 18, Autoregressive
parameters 19 and also frequency-domain feature extraction techniques like
power spectral density(PSD) and time-frequency correlated features extraction
techniques likediscrete wavelet transform (DWT) 20. We first plot the raw EEG
signal pattern recorded from the specific brain regions during the experiment.
Fig. 7(a)showsthe plot of raw EEG signalscaptured from the temporal lobe during
the smell stimuli presentation and Fig. 7(b)represents the filtered EEG signals
for four different odor stimuli.

Power
spectral density (PSD), is a well-known frequency-domain feature extraction
technique to extract EEG signal power distribution. PSD is applied on the
filtered EEG signal acquired from prefrontal, frontal, parietal and temporal
lobe regions. Besides PSD, It is important to mention here that filtering of
EEG signal is done by using a standard Chebyshev 21 band pass infinite
impulse response (IIR) filter of order 10, which has the pass band frequency of
0.5-13 Hz. The selection is made so because of the superior performance of
Chebyshev filter as compared to its standard counterparts including Butterworth
and Elliptic filter. Now, for each subject and each vowel sound, PSD extract
10×12×328 feature sets  (since, here,
experiment is repeated 10 times and number of selected electrodes 12). Fig.8(a)
and Fig.8(b)present the PSD features extractedfrom the above brain regions.A.   
Selection of Brain Regions

EEG signals are acquired from 21 electrodes, placing them
on the scalp of human subjects. The EEG signals are recorded on a separate
computer having 8 GB RAM with CPUclock of

 

Low

Medium

High

Activation level

Odor 1 (Perfume)

 

 

 

 

Odor 2 (Dettol)

 

 

 

 

Odor 3 (Acetic acid)

 

 

 

 

Odor 4 (Alchohol)

 

Fig.5. Section of brain region and
frequency bands from scalp maps of three concentration level of four stimuli

3.4 GHz.Fig. 5 shows the scalp
maps of one subject. Scalp maps of different concentration level of different
stimuli have been recorded during the experiment.
From the scalp maps, it can be observed that pre-frontal, Frontal, parietal,
and temporal lobes exhibit significant activations during the experiment. Here, the pre-frontal region is hardly activated inolfactory
sensing and processing. Besides these, frontal and mainly parietal, temporal
are found to take significantly active participation during the experiment. The
paper 6 reveals that temporal lobe and frontal lobe are highly associated
with human olfactory signal processing. Therefore, we select F3, F4,
F7, F8 and FZ (for frontal lobe) P3
and P4 (from parietal lobe), T3 and T4(from
temporal lobe) and FP1 and FP2 (for prefrontal lobe)for
extracting necessary information using signal processing techniques.Fig.6shows the electrode positions over the scalp in our
brain during experiment, where the red colored electrodes are used to collect
data. Electrodes are placed over the scalp using 10-20 electrode placement
system 26.

Fig.6. Electrode position of
our experiment

 

B.   
EEG Feature Extraction and Feature Selection

Featureextraction
is very much important for EEG classification problem for accurate decoding of
mental tasks. There is variety of EEG feature extraction schemes: time-domain
feature extraction techniques like Hjorth parameters 18, Autoregressive
parameters 19 and also frequency-domain feature extraction techniques like
power spectral density(PSD) and time-frequency correlated features extraction
techniques likediscrete wavelet transform (DWT) 20. We first plot the raw EEG
signal pattern recorded from the specific brain regions during the experiment.
Fig. 7(a)showsthe plot of raw EEG signalscaptured from the temporal lobe during
the smell stimuli presentation and Fig. 7(b)represents the filtered EEG signals
for four different odor stimuli.

Power
spectral density (PSD), is a well-known frequency-domain feature extraction
technique to extract EEG signal power distribution. PSD is applied on the
filtered EEG signal acquired from prefrontal, frontal, parietal and temporal
lobe regions. Besides PSD, It is important to mention here that filtering of
EEG signal is done by using a standard Chebyshev 21 band pass infinite
impulse response (IIR) filter of order 10, which has the pass band frequency of
0.5-13 Hz. The selection is made so because of the superior performance of
Chebyshev filter as compared to its standard counterparts including Butterworth
and Elliptic filter. Now, for each subject and each vowel sound, PSD extract
10×12×328 feature sets  (since, here,
experiment is repeated 10 times and number of selected electrodes 12). Fig.8(a)
and Fig.8(b)present the PSD features extractedfrom the above brain regions.A.   
Selection of Brain Regions

EEG signals are acquired from 21 electrodes, placing them
on the scalp of human subjects. The EEG signals are recorded on a separate
computer having 8 GB RAM with CPUclock of

 

Low

Medium

High

Activation level

Odor 1 (Perfume)

 

 

 

 

Odor 2 (Dettol)

 

 

 

 

Odor 3 (Acetic acid)

 

 

 

 

Odor 4 (Alchohol)

 

Fig.5. Section of brain region and
frequency bands from scalp maps of three concentration level of four stimuli

3.4 GHz.Fig. 5 shows the scalp
maps of one subject. Scalp maps of different concentration level of different
stimuli have been recorded during the experiment.
From the scalp maps, it can be observed that pre-frontal, Frontal, parietal,
and temporal lobes exhibit significant activations during the experiment. Here, the pre-frontal region is hardly activated inolfactory
sensing and processing. Besides these, frontal and mainly parietal, temporal
are found to take significantly active participation during the experiment. The
paper 6 reveals that temporal lobe and frontal lobe are highly associated
with human olfactory signal processing. Therefore, we select F3, F4,
F7, F8 and FZ (for frontal lobe) P3
and P4 (from parietal lobe), T3 and T4(from
temporal lobe) and FP1 and FP2 (for prefrontal lobe)for
extracting necessary information using signal processing techniques.Fig.6shows the electrode positions over the scalp in our
brain during experiment, where the red colored electrodes are used to collect
data. Electrodes are placed over the scalp using 10-20 electrode placement
system 26.

Fig.6. Electrode position of
our experiment

 

B.   
EEG Feature Extraction and Feature Selection

Featureextraction
is very much important for EEG classification problem for accurate decoding of
mental tasks. There is variety of EEG feature extraction schemes: time-domain
feature extraction techniques like Hjorth parameters 18, Autoregressive
parameters 19 and also frequency-domain feature extraction techniques like
power spectral density(PSD) and time-frequency correlated features extraction
techniques likediscrete wavelet transform (DWT) 20. We first plot the raw EEG
signal pattern recorded from the specific brain regions during the experiment.
Fig. 7(a)showsthe plot of raw EEG signalscaptured from the temporal lobe during
the smell stimuli presentation and Fig. 7(b)represents the filtered EEG signals
for four different odor stimuli.

Power
spectral density (PSD), is a well-known frequency-domain feature extraction
technique to extract EEG signal power distribution. PSD is applied on the
filtered EEG signal acquired from prefrontal, frontal, parietal and temporal
lobe regions. Besides PSD, It is important to mention here that filtering of
EEG signal is done by using a standard Chebyshev 21 band pass infinite
impulse response (IIR) filter of order 10, which has the pass band frequency of
0.5-13 Hz. The selection is made so because of the superior performance of
Chebyshev filter as compared to its standard counterparts including Butterworth
and Elliptic filter. Now, for each subject and each vowel sound, PSD extract
10×12×328 feature sets  (since, here,
experiment is repeated 10 times and number of selected electrodes 12). Fig.8(a)
and Fig.8(b)present the PSD features extractedfrom the above brain regions.