You are on page 1of 14

SSVEP Channel Selection

Using Adaptive
Filtering

Presented By:-

Vijay Rohilla
Contents
• Introduction
• Steady State Visual Evoked Potential (SSVEP)
• Purposed Work
• ADAPTIVE FILTERS
• Channel Selection Algorithm
• RESULT
• CONCLUSIONS
• References
Introduction
 Brain Computer Interface (BCI) is a multidisciplinary research area. Brain computer interfaces
provides an additional channel to communicate with computer based on the measurement of
direct neural activity.

 Steady State Visual Evoked Potential based Brain Computer Interface has many advantages
over other method of Brain Computer Interface like high Signal to Noise Ratio and high
Information Transfer Rate .
Steady State Visual Evoked Potential (SSVEP)
SSVEP signals are the periodic Electroencephalogram (EEG) responses to the
visual stimulus with a pre-defined or periodic flickering pattern of specific
frequency. The user put his/her attention at a flickering pattern and the system
determines the frequency of this flickering pattern response over occipital region
of the brain.
If the pattern whose frequency matches with the frequency response over the
occipital region then this imply that the user wanted to select that particular
pattern and that pattern is selected. These flickering pattern may relates to any
command for the machine or computer.
Although the mechanism of SSVEP generation is not definite, and there is some
difference of its amplitude among subjects frequency is very critical information
in SSVEP data.
Purposed Work
In this paper we purposed the application of adaptive filtering for the detection of
frequency of SSVEP signal and extended this concept in recognizing the sensor
location which is best in the term of frequency content relative to the frequency of
the flickering pattern.
METHODOLOGY AND MATERIAL
 DATABASE DISCRIPTION
• We have used database of EEG system: 128-channel BIOSEMI (active electrodes),
sampling rate: 256 Hz.15 seconds of SSVEP available (excluding 5 seconds of
offset and onset) for 4 subjects.
• 3 SSVEP stimulation frequencies available (8 Hz, 14 Hz, 28 Hz). 5 trials available
for each SSVEP frequency. This database is provided at free of cost and available
on internet for anyone to download.
ADAPTIVE FILTERS
• SSVEP signal 𝑥𝑖 (𝑛) a random signal and can be treated as the zero mean
stationary signal, is obtained from the a
• channel/sensor placed on the scalp. The signal contains the EEG recorded over
the scalp. Here i is the number of the channel. 𝑤𝑖𝑛 is the weight of the variable
filter which changes its value in each iteration .

Block diagram of general adaptive filter


Channel Selection Algorithm
For a specific frequency, channels selected in a trials are then compared and
normalized with the reference to the channel which has the highest spectral
component, and relative value is assigned to each channel based on which the best
three channel is selected in a trial. This process is repeated for all trial and thus we
get 15 selected channels for a specific frequency

channel selection algorithm


• Channel selection for the subject-1 at three different frequencies 8Hz, 14Hz, and
28Hz for five trials are given below in table 1(a), 1(b), and 1(c) respectively. The
channel selected in the

Channel selection for subject 1 for 8Hz frequency Channel selected for subject 1 for 14Hz frequency
Channel selected for subject 1 for 28Hz frequency

Among channels in tables-1(a), 1(b), and 1(c) selection values


for channels B1, B2, and B3 are highest hence B1, B2 and B3 are
selected as the best channel locations for subject 1 in reaction to
8Hz, 14Hz, and 28Hz flickering frequency
RESULT
• The maximum value of the X(f) corresponds to the stimulus frequency, signal to
noise ratio is taken as the performance evaluation parameter ratio of X(f) over
the mean of adjacent 16 value is defined as SNR.

16 × |𝑋(𝑓)|2
𝑆𝑁𝑅 = 8
σ𝑘=1[|𝑋( 𝑓 = 0.25 × 𝑘 |2 + |𝑋 𝑓 − 0.25 × 𝑘 |2 ]

. SNR for selected channel of sub1 Trial 1


CONCLUSIONS
• SSVEP based BCI depends on the neurological activities of the brain which differs
significantly from person to person hence to locate the optimal sensor location
for a person is very essential since lesser the number of sensors required for the
detection of brain signal results in the unreliable signal due to poor spatial
resolution of these signals which make it difficult for practical application

• The optimal channel locations as suggested in this paper is an attempt in the


same direction of reducing the number of sensor but with the less possible loss of
information.
• Channels selected in this paper shows the sensor location which gives the best
spectral component of the flickering frequency for respective users/subjects.
These results may be used further in designing a better BCI system with less
number of the sensors.
References
• [1] Feldbauer C., Pernkopf F., and Rank Erhard. “Adaptive Filters”.
• [2] Y. Zhang, G. Zhou, J. Jin, X. Wang, A. Cichocki “Frequency recognition in SSVEP-based
BCI using multiset canonical correlation analysis”.
• [3] Zheng-Hua Wu and De-Zhong Yao “A Study on SSVEP Based BCI”. Journal of
electronic science and technology of china, vol. 7, no. 1, March 2009
• [4] Lijing Meng, Jing Jin, Xingyu Wang. “A Comparison of Three Electrode Channels
Selection Methods Applied to SSVEP BCI”. 4th International Conference on Biomedical
Engineering and Informatics (BMEI). 2011.
• [5] Yijun Wang, Zhiguang Zhang, Xiaorong Gao, Shangkai Gao. “Lead selection for
SSVEP-based brain-computer interface”. Proceedings of the 26th Annual International
Conference of the IEEE EMBS San Francisco, CA, USA • September 1-5, 2004.
• [6] Jiahui Pan, Yuanqing Li, Rui Zhang, Zhenghui Gu, and Feng Li. “Discrimination
Between Control and Idle States in Asynchronous SSVEP-Based Brain Switches: A
Pseudo-Key- Based Approach”. IEEE transactions on neural systems and rehabilitation
engineering, vol. 21, no. 3, May 2013.
• [7] Henrik Gollee, Ivan Volosyak, Angus J. McLachlan, Kenneth J. Hunt, and Axel Graser. “An SSVEP-Based
Brain– Computer Interface for theControl of Functional Electrical Stimulation”. IEEE Transactions on
biomedical engineering, vol. 57, no. 8, August 2010.
• [8] Brendan Z. Allison, Moore Jackson, Dennis J. McFarland, Gerwin Schalk, and Jonathan R. Wolpaw.
“Towards an Independent Brain - Computer Interface Using Steady State Visual Evoked Potentials”. Clin
Neurophysiol. 2008 February ; 119(2): 399–408.
• [9] Gernot R. Müller-Putz, and Gert Pfurtscheller. “Control of an Electrical Prosthesis With an SSVEP-
Based BCI”. IEEE Transactions on biomedical engineering, vol. 55, no. 1, January 2008.
• [10] Zhenghua Wu and Dezhong Yao. “Frequency detection with stability coefficient for steady-state
visual evoked potential (SSVEP)-based BCIs”. J. Neural Eng. 5 (2008) 36–43.
• [11] Ola Friman*, Ivan Volosyak, and Axel Gräser. “Multiple Channel Detection of Steady-State Visual
Evoked Potentials for Brain-Computer Interfaces”. IEEE Transactions on biomedical engineering, vol. 54,
no. 4, April 2007.
• [12] Hsiang-Chih Chang, Po-Lei Lee, Men-Tzung Lo, I.-Hui Lee, Ting-Kuang Yeh, and Chun-Yen Chang.
“Independence of Amplitude-Frequency and PhaseCalibrations in an SSVEP Based BCI UsingStepping
Delay Flickering Sequences”.Neural systems and rehabilitation engineering, vol. 20, no. 3,May 2012.
• [13] Akitoshi Itai and Arao Funase. “Spectrum Based Feature Extraction Using Spectrum Intensity Ratio
for SSVEP Detection”. 34th Annual International Conference of the IEEE EMBS San Diego, California USA,
28 August - 1 September,2012.
• [14] Toshihisa Tanaka, Cheng Zhang, and Hiroshi Higashi. “SSVEP frequency detection methods
considering background EEG”. SCIS-ISIS 2012, Kobe, Japan, November 20-24, 2012.
• [15] H. Cecotti,.“Classification of Steady-State Visual Evoked Potentials based on the Visual Stimuli Duty
Cycle”. 978-1- 4244-8132-3/10/$26.00 ©2010 IEEE
Thank You

You might also like