M. Jawad Khan

Decoding of four movement directions using hybrid NIRS-EEG brain-computer interface.

The hybrid brain-computer interface (BCI)s multimodal technology enables precision brain-signal classification that can be used in the formulation of control commands. In the present study, an experimental hybrid near-infrared spectroscopy-electroencephalography (NIRS-EEG) technique was used to extract and decode four different types of brain signals. The NIRS setup was positioned over the prefrontal brain region, and the EEG over the left and right motor cortex regions. Twelve subjects participating in the experiment were shown four direction symbols, namely, “forward,” “backward,” “left,” and “right.” The control commands for forward and backward movement were estimated by performing arithmetic mental tasks related to oxy-hemoglobin (HbO) changes. The left and right directions commands were associated with right and left hand tapping, respectively. The high classification accuracies achieved showed that the four different control signals can be accurately estimated using the hybrid NIRS-EEG technology.

Passive BCI based on drowsiness detection: an fNIRS study

We use functional near-infrared spectroscopy (fNIRS) to discriminate the alert and drowsy states for a passive brain-computer interface (BCI). The passive brain signals for the drowsy state are acquired from the prefrontal and dorsolateral prefrontal cortex. The experiment is performed on 13 healthy subjects using a driving simulator, and their brain activity is recorded using a continuous-wave fNIRS system. Linear discriminant analysis (LDA) is employed for training and testing, using the data from the prefrontal, left- and right-dorsolateral prefrontal regions. For classification, eight features are tested: mean oxyhemoglobin, mean deoxyhemoglobin, skewness, kurtosis, signal slope, number of peaks, sum of peaks, and signal peak, in 0~5, 0~10, and 0~15 second time windows, respectively. The results show that the best performance for classification is achieved using mean oxyhemoglobin, the signal peak, and the sum of peaks as features. The average accuracies in the right dorsolateral prefrontal cortex (83.1, 83.4 and 84.9% in the 0~5, 0~10 and 0~15 second time windows, respectively) show that the proposed method has an effective utility for detection of drowsiness for a passive BCI.

A hybrid EEG-fNIRS BCI: Motor imagery for EEG and mental arithmetic for fNIRS

In this paper, we have combined electroencephalography (EEG) and functional near-infrared spectroscopy (fNRIS) to make a hybrid EEG-NIRS based system for brain-computer interface (BCI). The EEG electrodes were placed on the motor cortex region and the NIRS optodes were set on the prefrontal region. The data of four subjects was acquired using mental arithmetic tasks and motor imageries of the left- and right-hand. The EEG data were band-pass filtered to obtain the activity (8~18 Hz). The modified Beer-Lambert law (MBLL) was used to convert the fNIRS data into oxy- and deoxy-hemoglobin (HbO and HbR), respectively. A common threshold between the two modalities was established to define a common resting state. The support vector machines (SVM) was used for data classification. Three control commands were generated using the prefrontal and motor cortex data. The results show that EEG and fNIRS can be combined for better brain signal acquisition and classification for BCI.

Hybrid EEG-NIRS based BCI for quadcopter control

In this paper, we have proposed a novel control strategy for a quadcopter control using brain signals. A brain-computer interface (BCI) technology is developed using hybrid electroencephalography - near-infrared spectroscopy (EEG-NIRS) system and two commands are used to operate the quadcopter. An active brain signal upon the users own will is generated using a motor imagery task and a reactive brain signal is generated by visual flickering of light. The reactive command is used for the triggering control of the quadcopter and the active command is used to navigate the quadcopter in the forward direction. Linear discriminant analysis is used to classify the brain activity in offline environment. The results indicate that the proposed scheme is suitable for the BCI control applications.

Hemodynamic signals based lie detection using a new wireless NIRS system

In this paper, we have demonstrated the ability of a new multi-channel near-infrared spectroscopy (NIRS) system to detect deception in a concealed information test (CIT) paradigm. Brain signals from prefrontal cortex area of four healthy male subjects are collected using two different machines for comparison. Oxy and deoxy-hemoglobin (HbO and HbR) signals are used to define the features and then data is classified using linear discriminant analysis (LDA). Results show that the difference in HbO and HbR signals during lie and truth can be detected by the new wireless NIRS system.

Effect of anodal tDCS on human prefrontal cortex observed by fNIRS

Transcranial direct current stimulation (tDCS) is one of the noninvasive brain stimulation methods that have been used to study many neuropsychiatric and neurological disorders in humans. tDCS can excite or inhibit the neurons depending upon its polarity. In this study, we have investigated the effect of anodal tDCS on human prefrontal cortex using functional near-infrared spectroscopy (fNIRS), which is a noninvasive neuroimaging technique. We have developed a new wireless fNIRS system compatible with EEG, and also developed a pad-type tDCS with variable current limits. Our wireless fNIRS system is composed of a microcontroller, an optical probe, tri-wavelength light emitting diodes (LEDs), photodiodes, WiFi communication module and battery. The developed tDCS system can generate the current in the range of 0.8 ~ 2.2 mA. To test the functionality of the systems, fNIRS data was recorded before and after the tDCS stimulation. The results of this study show that the anodal tDCS excites the neurons in the region of interest and this excitability is monitored using the fNIRS system.

Drowsiness detection in dorsolateral-prefrontal cortex using fNIRS for a passive-BCI

In this paper, we have investigated the feasibility of detecting drowsiness using hemodynamic brain signals for a passive brain-computer interface (BCI). Functional near-infrared spectroscopy (fNIRS) is used to measure the right dorsolateral-prefrontal brain region in order to investigate the hemodynamic changes corresponding to drowsy and alert states. The data is recorded using five drowsy subjects during a simulated car driving task. The recoded data are converted into oxy- and deoxy-hemoglobin (HBO and HbR) using the modified Beer-Lambert law (MBLL) for feature extraction and classification. Signal mean and signal slope are extracted using the spatio-temporal time windows as features. Linear discriminant analysis (LDA) and support vector machines (SVM) are used for the training and testing of the brain data. The classification accuracy obtained using offline analyses is 74% and 77% respectively. The results show that drowsy and alert states are distinguishable from the right dorsolateral prefrontal brain region. Also, fNIRS modality can be used for drowsiness detection for a passive BCI.

Initial dip detection based on both HbO and HbR vector-based phase analysis

A new threshold circle to minimize possible misclassification of initial dips in the functional near-infrared spectroscopy (fNIRS) signals using the vector-based phase analysis is investigated. In contrast to the work in [20] (i.e., the square root of the sum of the squares of oxy- and deoxy-hemoglobins), the peak value of oxy- or deoxy-hemoglobin during the resting state is used. The experiment was performed on five healthy subjects. The activity was measured using a frequency domain fNIRS system during the mental arithmetic task from the prefrontal cortex. With the new criterion, the radius of the circle becomes smaller than that in [20], and earlier detection of initial dips are possible. For the given arithmetic task, the channels of detecting the initial dips were not the same over the subjects, which reflects that the activated brain region in association with the performed task spreads in the prefrontal cortex. Also, the new method can further reduce the misinterpretation of large variations in the resting state and even during the task period.

Drowsiness detection using fNIRS in different time windows for a passive BCI

In this research, we have investigated the detection of drowsiness activity in dorsolateral-prefrontal cortex in three different time windows (0~3 sec, 0~4 sec and 0~5 sec) using functional near-infrared spectroscopy (fNIRS). Five drowsy subjects participated in a simulated driving task while their brain activity is monitored using fNIRS. The recorded brain activity is segmented into three windows for the acquisition of signal mean, signal slope and number of peaks as features. The data in each window is classified using linear discriminant analysis to find best window size. The results show that the best accuracy is obtained using 0~5 sec window after classification. Although the classification accuracy in 0~4 sec window is lower than in 0~5 sec window, both accuracies are suitable for brain-computer interface applications (i.e. accuracy>70%). The accuracy in 0~3 sec window is less than 70% for two subjects. For driver drowsiness detection, high accuracy with quick detection time is required, therefore we propose drowsiness detection in 0~4 sec window using fNIRS monitoring.

Motor imagery performance evaluation using hybrid EEG-NIRS for BCI

In this paper, we have evaluated the performance of motor imagery (MI), before and after training by a rehabilitation robot, for brain-computer interface (BCI). A hybrid electroencephalography and near-infrared spectroscopy (EEG-NIRS) system is used to detect the MI by placing the electrodes and optodes around the motor cortex region. Five healthy subjects have participated in the experiment. The subjects are assisted by a rehabilitation robot in an arm movement paradigm during the training session. The MI activity of the subjects is recorded before and after the training sessions. The brain signals from the motor cortex are recorded simultaneously using EEG-NIRS. We found a significant improvement in the MI performance after training. Linear discriminant analysis is used to classify the acquired activity in an offline analysis. The data analysis shows that the hybrid EEG-NIRS can detect better motor activity than individual modality. The average classification accuracy of the subjects has increased from 66% to 94% after training. We propose that the training of the motor cortex by a rehabilitation robot can improve the MI performance for BCI.