Gan Huang

A novel approach to predict subjective pain perception from single-trial laser-evoked potentials

Pain is a subjective first-person experience, and self-report is the gold standard for pain assessment in clinical practice. However, self-report of pain is not available in some vulnerable populations (e.g., patients with disorders of consciousness), which leads to an inadequate or suboptimal treatment of pain. Therefore, the availability of a physiology-based and objective assessment of pain that complements the self-report would be of great importance in various applications. Here, we aimed to develop a novel and practice-oriented approach to predict pain perception from single-trial laser-evoked potentials (LEPs). We applied a novel single-trial analysis approach that combined common spatial pattern and multiple linear regression to automatically and reliably estimate single-trial LEP features. Further, we adopted a Naïve Bayes classifier to discretely predict low and high pain and a multiple linear prediction model to continuously predict the intensity of pain perception from single-trial LEP features, at both within- and cross-individual levels. Our results showed that the proposed approach provided a binary prediction of pain (classification of low pain and high pain) with an accuracy of 86.3 ± 8.4% (within-individual) and 80.3 ± 8.5% (cross-individual), and a continuous prediction of pain (regression on a continuous scale from 0 to 10) with a mean absolute error of 1.031 ± 0.136 (within-individual) and 1.821 ± 0.202 (cross-individual). Thus, the proposed approach may help establish a fast and reliable tool for automated prediction of pain, which could be potentially adopted in various basic and clinical applications.

Interactions between two neural populations: A mechanism of chaos and oscillation in neural mass model

Neural mass model developed by Lopes da Silva et al. is able to describe limit cycle behavior in Electroencephalography (EEG) of alpha rhythm and exhibit complex dynamics between cortical areas. In this work, we extend Grimbert and Faugerass work to study the dynamical behavior caused by interaction of cortical areas. The model is developed with the coupling of two neural populations. We show that various attractors, including equilibrium points, periodic solutions and chaotic strange attractors, could coexist in different ways with different value of the connectivity parameters. The main findings are that: (1) The stable equilibrium points only appear with a small value of the parameter. (2) While the alpha activities always exist for both two populations with proper initial conditions. Interestingly, the coexistence of the multiple alpha-to-epileptic activities implies the multiple coupling ways for these activities in phase. Two neuronal populations with epileptic activities could interact with multiple rhythms depending on their connectivity. (3) For particular interest, chaotic behaviors are identified in four regions divided by the connectivity parameter with the positive maximal Lyapunov exponent. The four types of chaotic attractors have their own structures, but all of them are related to the epileptic activities.

Model based generalization analysis of common spatial pattern in brain computer interfaces

In the motor imagery based Brain Computer Interface (BCI) research, Common Spatial Pattern (CSP) algorithm is used widely as a spatial filter on multi-channel electroencephalogram (EEG) recordings. Recently the overfitting effect of CSP has been gradually noticed, but what influence the overfitting is still unclear. In this work, the generalization of CSP is investigated by a simple linear mixing model. Several factors in this model are discussed, and the simulation results indicate that channel numbers and the correlation between signals influence the generalization of CSP significantly. A larger number of training trials and a longer time length of the trial would prevent overfitting. The experiments on real data also verify our conclusion.

Automated selecting subset of channels based on CSP in motor imagery brain-computer interface system

The Common Spatial Pattern (CSP) algorithm is a popular method for efficiently calculating spatial filters. However, several previous studies show that CSPs performance deteriorates especially when the number of channels is large compared to small number of training datasets. As a result, it is necessary to choose an optimal subset of the whole channels to save computational time and retain high classification accuracy. In this paper, we propose a novel heuristic algorithm to select the optimal channels for CSP. The CSP procedure is applied to training datasets firstly and then a channel score based on l\ norm is defined for each channel. Finally, channels with larger scores are retained for further CSP processing. This approach utilizes CSP procedure twice to select channels and extract features, respectively; hence the complex optimization problem of channel selection for CSP is solved heuristically. We apply our method and other two existing methods to datasets from BCI competition 2005 for comparison and the experiment results show this method provides an effective way to accomplish the task of channel selection.

Optimizing spatial spectral patterns jointly with channel configuration for brain-computer interface

The power of common spatial pattern (CSP) has been widely validated in electroencephalogram (EEG) based brain-computer interface (BCI). However, its effectiveness is highly dependent on subject-specific time segment, channel configuration and frequency band. Hence, the preprocessing procedure of CSP algorithm is critical to enhance the performance of BCI system. This paper proposes a feature extraction and selection method based on common spatial and spectral pattern for motor imagery brain-computer interface (BCI). We formulate the optimization of spatial spectral patterns, channel configuration and time segment as maximizing the proposed criterions including mutual information algorithm, Fisher ratio algorithm and wrapper method. The proposed method is evaluated on single trial EEG from dataset IVa of BCI competition III. The results show that best features are selected by a wrapper method and these features in cross-validation yield better performance compared to most of the reported results.

An EMG-based handwriting recognition through dynamic time warping

In this paper, an electromyography (EMG)-based handwriting recognition method was proposed for a latent tendency of natural user interface. The subjects wrote the characters at a normal speed, and six channels of EMG signals were recorded from forearm muscles. The dynamic time warping (DTW) algorithm was used to eliminate the time axis variance during writing. The process for template making and matching was illustrated diagrammatically. The results showed that no more than ten training trials per character could make an accuracy of above 90%. The recognition performance was compared in three character sets: digits, Chinese characters and capital letters.

A frequency-weighted method combined with Common Spatial Patterns for electroencephalogram classification in brain–computer interface

Abstract Common Spatial Patterns (CSP) has been proven to be a powerful and successful method in the detection of event-related desynchronization (ERD) and ERD based brain–computer interface (BCI). However, frequency optimization combined with CSP has only been investigated by a few groups. In this paper, a frequency-weighted method (FWM) is proposed to optimize the frequency spectrum of surface electroencephalogram (EEG) signals for a two-class mental task classification. This straightforward method computes a weight value for each frequency component according to its importance for the discrimination task and reforms the spectrum with the computed weights. The off-line analysis shows that the proposed method achieves an improvement of about 4% (averaged over 24 datasets) in terms of cross-validation accuracy over the basic CSP.

Spatio-spectral filters for low-density surface electromyographic signal classification

In this paper, we proposed to utilize a novel spatio-spectral filter, common spatio-spectral pattern (CSSP), to improve the classification accuracy in identifying intended motions based on low-density surface electromyography (EMG). Five able-bodied subjects and a transradial amputee participated in an experiment of eight-task wrist and hand motion recognition. Low-density (six channels) surface EMG signals were collected on forearms. Since surface EMG signals are contaminated by large amount of noises from various sources, the performance of the conventional time-domain feature extraction method is limited. The CSSP method is a classification-oriented optimal spatio-spectral filter, which is capable of separating discriminative information from noise and, thus, leads to better classification accuracy. The substantially improved classification accuracy of the CSSP method over the time-domain and other methods is observed in all five able-bodied subjects and verified via the cross-validation. The CSSP method can also achieve better classification accuracy in the amputee, which shows its potential use for functional prosthetic control.

An automated and fast approach to detect single-trial visual evoked potentials with application to brain–computer interface

OBJECTIVE This study aims (1) to develop an automated and fast approach for detecting visual evoked potentials (VEPs) in single trials and (2) to apply the single-trial VEP detection approach in designing a real-time and high-performance brain-computer interface (BCI) system. METHODS The single-trial VEP detection approach uses common spatial pattern (CSP) as a spatial filter and wavelet filtering (WF) a temporal-spectral filter to jointly enhance the signal-to-noise ratio (SNR) of single-trial VEPs. The performance of the joint spatial-temporal-spectral filtering approach was assessed in a four-command VEP-based BCI system. RESULTS The offline classification accuracy of the BCI system was significantly improved from 67.6±12.5% (raw data) to 97.3±2.1% (data filtered by CSP and WF). The proposed approach was successfully implemented in an online BCI system, where subjects could make 20 decisions in one minute with classification accuracy of 90%. CONCLUSIONS The proposed single-trial detection approach is able to obtain robust and reliable VEP waveform in an automatic and fast way and it is applicable in VEP based online BCI systems. SIGNIFICANCE This approach provides a real-time and automated solution for single-trial detection of evoked potentials or event-related potentials (EPs/ERPs) in various paradigms, which could benefit many applications such as BCI and intraoperative monitoring.

Single-trial detection of visual evoked potentials by common spatial patterns and wavelet filtering for brain-computer interface

Event-related potentials (ERPs) are widely used in brain-computer interface (BCI) systems as input signals conveying a subjects intention. A fast and reliable single-trial ERP detection method can be used to develop a BCI system with both high speed and high accuracy. However, most of single-trial ERP detection methods are developed for offline EEG analysis and thus have a high computational complexity and need manual operations. Therefore, they are not applicable to practical BCI systems, which require a low-complexity and automatic ERP detection method. This work presents a joint spatial-time-frequency filter that combines common spatial patterns (CSP) and wavelet filtering (WF) for improving the signal-to-noise (SNR) of visual evoked potentials (VEP), which can lead to a single-trial ERP-based BCI.