Detection of Target Frequency from SSVEP Signal Using Empirical Mode Decomposition for SSVEP Based BCI Inference System
- PDF / 1,064,309 Bytes
- 13 Pages / 439.37 x 666.142 pts Page_size
- 27 Downloads / 157 Views
Detection of Target Frequency from SSVEP Signal Using Empirical Mode Decomposition for SSVEP Based BCI Inference System Mukesh Kumar Ojha1 · Manoj Kumar Mukul1
© Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract This paper describes the effectiveness of feature obtained by power spectrum analysis (PSA) as well as the combined method of empirical mode decomposition (EMD) and PSA for the development of brain–computer interface (BCI) system using steady-state visual evoked potential (SSVEP). Accurate detection of SSVEP response from the recorded EEG signal is a difficult task for a new development of the BCI inference system. The EMD technique is a non-linear method of signal decomposition, which generates several intrinsic mode functions (IMFs) of different flickering frequencies. Prominent IMF signal of SSVEP plays a vital role in the accurate detection of frequency. The proposed method achieves the average detection accuracy of 81.45% over four subjects; in contrast, the conventional method of PSA achieves average detection accuracy of 80.43%. The achieved result indicates that the proposed method out performs state of the art by more than 1.02% over four subjects. Keywords Steady-state visual evoked potential (SSVEP) · Brain–computer interface (BCI) · Empirical mode decomposition (EMD) · Power spectrum analysis (PSA) · Intrinsic mode function (IMF)
1 Introduction A brain–computer interface (BCI) system enables a subject to communicate with a computer directly through the incoming brain signals from the subject’s brain [1–3]. BCIs are widely used in the field of research, augmentation, mapping, assisting, and human body function repairing. The most used brain signals are event-related potential (ERP), P300, and steady-state visual evoked potential (SSVEP) [4, 5] to design a BCI inference system. Nowadays, most of the BCI researchers are concentrating on the SSVEP-based BCI inference system due to the high-information transfer rate (ITR), minimal training time, and high signal-to-noise ratio (SNR) [6–8].
* Mukesh Kumar Ojha [email protected] 1
Electronics and Communication Engineering, BIT Mesra, Ranchi, Jharkhand 835215, India
13
Vol.:(0123456789)
M. K. Ojha, M. K. Mukul
The SSVEP is an evoked signal that flows through normal electroencephalogram (EEG) signals when a subject concentrates the visual stimulus which flickers at a particular frequency [9, 10]. The SSVEP signal appears in the recorded EEG signal when the neuronal circuits over the posterior region of the brain resonate with the stimulation frequency. Consequently, the component of flickering stimulus frequency and its harmonics flows with the spontaneous EEG signal. The recorded EEG signal over the surface of the head is contaminated with various artifacts [11] too. Thus, the presence of artifacts and other background EEG signals creates the challenge in the accurate detection of flickering frequency from the recorded brain signals for the development of SSVEP based BCI inference system [12–14]. Therefore, t
Data Loading...
