An Adaptive SSVEP-Based Brain-Computer Interface to Compensate Fatigue-Induced Decline of Performance in Practical Application.

Brain-computer interfaces based on steady-state visual evoked potentials are promising communication systems for people with speech and motor disabilities. However, reliable SSVEP response requires user's attention, which degrades over time due to significant eye-fatigue when low-frequency visual stimuli (5-15 Hz) are used. Previous studies have shown that eye-fatigue can be reduced using high-frequency flickering stimuli (>25 Hz). Here, it is quantitatively demonstrated that the performance of a high-frequency SSVEP BCI decreases over time, but this amount of decrease can be compensated effectively by using two proposed adaptive algorithms. This leaded to a robust alternative communication system for practical applications. The asynchronous spelling system implemented in this study uses a threshold-based version of LASSO algorithm for frequency recognition. In long online experiments, when participants typed a sentence with the BCI system for 16 times, accuracy of the system was close to its maximum along the experiment. However, regression analysis on typing speed of each sentence demonstrated a significant decrease in all 7 subjects ( ) when thresholds obtained from a calibration test were kept fixed over the experiment. In comparison, no significant change in typing speed was observed when the proposed adaptive algorithms were used. The analysis of variances revealed that the average typing speed of the last four sentences when using adaptive relational algorithm (8.7 char/min) was significantly higher than the tolerance-based algorithm (8.1 char/min) and significantly above 6 char/min when the fixed thresholds were used. Therefore, the relational algorithm proposed in this paper could successfully compensate for the effect of fatigue on performance of the SSVEP BCI system.

An Efficient Asynchronous High-Frequency Steady-State Visual Evoked Potential-Based Brain-Computer Interface speller: The Problem of Individual Differences.

BACKGROUND: Brain-computer interfaces (BCIs) based on steady-state visual evoked potentials (SSVEPs) provide high rates of accuracy and information transfer rate, but need user's attention to flickering visual stimuli. This quickly leads to eye-fatigue when the flickering frequency is in the low-frequency range. High-frequency flickering stimuli (>30 Hz) have been proposed with significantly lower eye-fatigue. However, SSVEP responses in this frequency range are remarkably weaker, leading to doubts about usability of high-frequency stimuli to develop efficient BCI systems. The purpose of this study was to evaluate if a practical SSVEP Speller can be developed with Repetitive Visual Stimuli in the high-frequency range. METHODS: An asynchronous high-frequency (35-40 Hz) speller for typing in Persian language was developed using five flickering visual stimuli. Least absolute shrinkage and selection operator algorithm with two user-calibrated thresholds was used to detect the user's selections. A total of 14 volunteers evaluated the system in an ordinary office environment to type 9 sentences consist of 81 characters with a multistage virtual keyboard. RESULTS: Despite very high performance of 6.9 chars/min overall typing speed, average accuracy of 98.3%, and information transfer rate of 64.9 bpm for eight of the participants, the other six participants had serious difficulty in spelling with the system and could not complete the typing experiment. CONCLUSIONS: The results of this study in accordance with some previous studies suggest that high rate of illiteracy in high-frequency SSVEP-based BCI systems may be a major burden for their practical application.

Performance Improvement of the Goertzel Algorithm in Estimating of Protein Coding Regions Using Modified Anti-notch Filter and Linear Predictive Coding Model.

The aim of this paper is to improve the performance of the conventional Goertzel algorithm in determining the protein coding regions in deoxyribonucleic acid (DNA) sequences. First, the symbolic DNA sequences are converted into numerical signals using electron ion interaction potential method. Then by combining the modified anti-notch filter and linear predictive coding model, we proposed an efficient algorithm to achieve the performance improvement in the Goertzel algorithm for estimating genetic regions. Finally, a thresholding method is applied to precisely identify the exon and intron regions. The proposed algorithm is applied to several genes, including genes available in databases BG570 and HMR195 and the results are compared to other methods based on the nucleotide level evaluation criteria. Results demonstrate that our proposed method reduces the number of incorrect nucleotides which are estimated to be in the noncoding region. In addition, the area under the receiver operating characteristic curve has improved by the factor of 1.35 and 1.12 in HMR195 and BG570 datasets respectively, in comparison with the conventional Goertzel algorithm.

Evaluation of estimating missed answers in conners adult ADHD rating scale (screening version).

OBJECTIVE: Conners Adult ADHD Rating Scale (CAARS) is among the valid questionnaires for evaluating Attention-Deficit/Hyperactivity Disorder in adults. The aim of this paper is to evaluate the validity of the estimation of missed answers in scoring the screening version of the Conners questionnaire, and to extract its principal components. METHOD: This study was performed on 400 participants. Answer estimation was calculated for each question (assuming the answer was missed), and then a Kruskal-Wallis test was performed to evaluate the difference between the original answer and its estimation. In the next step, principal components of the questionnaire were extracted by means of Principal Component Analysis (PCA). Finally the evaluation of differences in the whole groups was provided using the Multiple Comparison Procedure (MCP). RESULTS: Findings indicated that a significant difference existed between the original and estimated answers for some particular questions. However, the results of MCP showed that this estimation, when evaluated in the whole group, did not show a significant difference with the original value in neither of the questionnaire subscales. The results of PCA revealed that there are eight principal components in the CAARS questionnaire. CONCLUSION: The obtained results can emphasize the fact that this questionnaire is mainly designed for screening purposes, and this estimation does not change the results of groups when a question is missed randomly. Notwithstanding this finding, more considerations should be paid when the missed question is a critical one.

Visual acuity classification using single trial visual evoked potentials.

Several researches have been done to identify visual system characteristics. Some of them are based on the processing of the brain signal recordings. Visual evoked potentials (VEPs) are electrical signals which are produced in response to the visual stimuli and recorded by means of electrodes placed on the head. These signals are usually characterized by the amplitude and latency of their peaks. Different types of visual stimuli and visual system characteristics can affect the shape and hence the characteristics of VEPs. In this paper, proper visual stimuli were used and VEPs were recorded in order to classify visual acuity. To achieve this goal, visual evoked potentials were recorded and processed in time, frequency and time-frequency domains. In order to preserve dynamics of the recorded signals, two algorithms for single-trial VEP extraction were used. The results of the classification of visual acuity in both average and single-trial VEPs are acceptable.

A new approach for EEG feature extraction in P300-based lie detection.

P300-based Guilty Knowledge Test (GKT) has been suggested as an alternative approach for conventional polygraphy. The purpose of this study was to extend a previously introduced pattern recognition method for the ERP assessment in this application. This extension was done by the further extending the feature set and also the employing a method for the selection of optimal features. For the evaluation of the method, several subjects went through the designed GKT paradigm and their respective brain signals were recorded. Next, a P300 detection approach based on some features and a statistical classifier was implemented. The optimal feature set was selected using a genetic algorithm from a primary feature set including some morphological, frequency and wavelet features and was used for the classification of the data. The rates of correct detection in guilty and innocent subjects were 86%, which was better than other previously used methods.

A comparison of methods for ERP assessment in a P300-based GKT.

P300-based GKT (guilty knowledge test) has been suggested as an alternative approach for conventional polygraphy. The purpose of this study is to evaluate three classifying methods for this approach and compare their performances in a lab analogue. Several subjects went through the designed GKT paradigm and their respective brain signals were recorded. For the analysis of signals, BAD (bootstrapped amplitude difference) and BCD (bootstrapped correlation difference) methods as two predefined methods alongside a new approach consisting of wavelet features and a statistical classifier were implemented. The rates of correct detection in guilty and innocent subjects were 74-80%. The results indicate the potential of P300-based GKT for detecting concealed information, although further research is required to increase its accuracy and precision and evaluating its vulnerability to countermeasures.