Chuang Lin

Enhanced Low-Latency Detection of Motor Intention From EEG for Closed-Loop Brain-Computer Interface Applications

In recent years, the detection of voluntary motor intentions from electroencephalogram (EEG) has been used for triggering external devices in closed-loop brain-computer interface (BCI) research. Movement-related cortical potentials (MRCP), a type of slow cortical potentials, have been recently used for detection. In order to enhance the efficacy of closed-loop BCI systems based on MRCPs, a manifold method called Locality Preserving Projection, followed by a linear discriminant analysis (LDA) classifier (LPP-LDA) is proposed in this paper to detect MRCPs from scalp EEG in real time. In an online experiment on nine healthy subjects, LPP-LDA statistically outperformed the classic matched filter approach with greater true positive rate (79 ± 11% versus 68 ± 10%; p = 0.007) and less false positives (1.4 ± 0.8/min versus 2.3 ± 1.1/min; p = 0.016). Moreover, the proposed system performed detections with significantly shorter latency (315 ± 165 ms versus 460 ± 123 ms; p = 0.013), which is a fundamental characteristics to induce neuroplastic changes in closed-loop BCIs, following the Hebbian principle. In conclusion, the proposed system works as a generic brain switch, with high accuracy, low latency, and easy online implementation. It can thus be used as a fundamental element of BCI systems for neuromodulation and motor function rehabilitation.

A Closed-Loop Brain–Computer Interface Triggering an Active Ankle–Foot Orthosis for Inducing Cortical Neural Plasticity

In this paper, we present a brain-computer interface (BCI) driven motorized ankle-foot orthosis (BCI-MAFO), intended for stroke rehabilitation, and we demonstrate its efficacy in inducing cortical neuroplasticity in healthy subjects with a short intervention procedure (~15 min). This system detects imaginary dorsiflexion movements within a short latency from scalp EEG through the analysis of movement-related cortical potentials (MRCPs). A manifold-based method, called locality preserving projection, detected the motor imagery online with a true positive rate of 73.0 ± 10.3%. Each detection triggered the MAFO to elicit a passive dorsiflexion. In nine healthy subjects, the size of the motorevoked potential (MEP) elicited by transcranial magnetic stimulation increased significantly immediately following and 30 min after the cessation of this BCI-MAFO intervention for ~15 min (p = 0.009 and p <; 0.001, respectively), indicating neural plasticity. In four subjects, the size of the short latency stretch reflex component did not change following the intervention, suggesting that the site of the induced plasticity was cortical. All but one subject also performed two control conditions where they either imagined only or where the MAFO was randomly triggered. Both of these control conditions resulted in no significant changes in MEP size (p = 0.38 and p = 0.15). The proposed system provides a fast and effective approach for inducing cortical plasticity through BCI and has potential in motor function rehabilitation following stroke.

Optimizing Kernel PCA Using Sparse Representation-Based Classifier for MSTAR SAR Image Target Recognition

Different kernels cause various class discriminations owing to their different geometrical structures of the data in the feature space. In this paper, a method of kernel optimization by maximizing a measure of class separability in the empirical feature space with sparse representation-based classifier (SRC) is proposed to solve the problem of automatically choosing kernel functions and their parameters in kernel learning. The proposed method first adopts a so-called data-dependent kernel to generate an efficient kernel optimization algorithm. Then, a constrained optimization function using general gradient descent method is created to find combination coefficients varied with the input data. After that, optimized kernel PCA (KOPCA) is obtained via combination coefficients to extract features. Finally, the sparse representation-based classifier is used to perform pattern classification task. Experimental results on MSTAR SAR images show the effectiveness of the proposed method.

Neighbourhood sensitive preserving embedding for pattern classification

Recently, a large family of supervised or unsupervised manifold learning algorithms that stem from statistical or geometrical theory has been designed to solve the problem of pattern classification. In this study, consider the fact that the data are usually sampled from a low-dimensional manifold space which resides in a high-dimensional Euclidean space, the authors propose a novel two-graph-based supervised linear classification algorithm called neighbourhood sensitive preserving embedding (NSPE). Different from local linear embedding (LLE) (or neighbourhood preserving embedding (NPE)) which preserves the local neighbourhood structure with one graph, NSPE can discover both the intrinsic and discriminant structure of the data manifold by constructing two graphs, that is, the within-class graph and the between-class graph. Thus, the data are mapped into a subspace where the nearby points with the same label are close to each other, whereas the nearby points with different labels are far apart. As a classification method, besides being defined on training samples, NSPE is also defined on testing samples. Experiments carried on the real-world face databases demonstrate that the results of all two-graph-based spectral methods are comparable and better than that of one-graph-based methods.

Discriminative manifold learning based detection of movement-related cortical potentials

The detection of voluntary motor intention from EEG has been applied to closed-loop brain-computer interfacing (BCI). The movement-related cortical potential (MRCP) is a low frequency component of the EEG signal, which represents movement intention, preparation, and execution. In this study, we aim at detecting MRCPs from single-trial EEG traces. For this purpose, we propose a detector based on a discriminant manifold learning method, called locality sensitive discriminant analysis (LSDA), and we test it in both online and offline experiments with executed and imagined movements. The online and offline experimental results demonstrated that the proposed LSDA approach for MRCP detection outperformed the Locality Preserving Projection (LPP) approach, which was previously shown to be the most accurate algorithm so far tested for MRCP detection. For example, in the online tests, the performance of LSDA was superior than LPP in terms of a significant reduction in false positives (FP) (passive FP: 1.6 ±0.9/min versus 2.9 ±1.0/min, p = 0.002, active FP: 2.2 ±0.8/min versus 2.7 ±0.6/min, p = 0.03), for a similar rate of true positives. In conclusion, the proposed LSDA based MRCP detection method is superior to previous approaches and is promising for developing patient-driven BCI systems for motor function rehabilitation as well as for neuroscience research.

Graph Regularized Nonnegative Matrix Factorization with Sparse Coding

In this paper, we propose a sparseness constraint NMF method, named graph regularized matrix factorization with sparse coding (GRNMF_SC). By combining manifold learning and sparse coding techniques together, GRNMF_SC can efficiently extract the basic vectors from the data space, which preserves the intrinsic manifold structure and also the local features of original data. The target function of our method is easy to propose, while the solving procedures are really nontrivial; in the paper we gave the detailed derivation of solving the target function and also a strict proof of its convergence, which is a key contribution of the paper. Compared with sparseness constrained NMF and GNMF algorithms, GRNMF_SC can learn much sparser representation of the data and can also preserve the geometrical structure of the data, which endow it with powerful discriminating ability. Furthermore, the GRNMF_SC is generalized as supervised and unsupervised models to meet different demands. Experimental results demonstrate encouraging results of GRNMF_SC on image recognition and clustering when comparing with the other state-of-the-art NMF methods.

Graph regularized non-negative matrix factorization with sparse coding

Matrix factorization techniques have been frequently utilized in pattern recognition and machine learning. Among them, Non-negative Matrix Factorization (NMF) has received considerable attention because it represents the naturally occurring data by parts of it. On the other hand, from the geometric perspective, the data is usually sampled from a low dimensional manifold embedded in a high dimensional ambient space. One hopes then to find a compact representation which uncovers the intrinsic geometric structure. In this paper, we propose a novel method, called Graph Regularized Non-negative Matrix Factorization with Sparse Coding (GRNMF_SC), the new model can learn much sparser representation and more discriminating power via imposing sparse constraint and Laplacian regularization explicitly. Experimental results on the ORL and Yale databases demonstrate encouraging performance of the proposed algorithm when compared with the state-of-the-art algorithms.

Matrix Factorization with Column L0-Norm Constraint for Robust Multi-subspace Analysis

We aim to study the subspace structure of data approximately generated from multiple categories and remove errors (e.g., noise, corruptions, and outliers) in the data as well. Most previous methods for subspace analysis learn only one subspace, failing to discover the intrinsic complex structure, while state-of-the-art methods use data itself as the basis (self-expressiveness property), showing degraded performance when data contain errors. To tackle the problem, we propose a novel method, called Matrix Factorization with Column L0-norm constraint (MFC0), from the matrix factorization perspective. MFC0 simultaneously discovers the multi-subspace structure of either clean or contaminated data, and learns the basis for each subspace. Specifically, the learnt basis with the orthonormal constraint shows high robustness to errors by adding a regularization term. Owing to the column l0-norm constraint, the generated representation matrix can be (approximate) block-diagonal after reordering its columns, with each block characterizing one subspace. We develop an efficient first-order optimization scheme to stably solve the nonconvex and nonsmooth objective function of MFC0. Experimental results on synthetic data and real-world face datasets demonstrate the superiority over traditional and state-of-the-art methods on both representation learning, subspace recovery and clustering.

Discriminant Manifold Learning via Sparse Coding for Image Analysis

Traditional subspace learning methods directly calculate the statistical properties of the original input images, while ignoring different contributions of different image components. In fact, the noise (e.g., illumination, shadow) in the image often has a negative influence on learning the desired subspace and should have little contribution to image recognition. To tackle this problem, we propose a novel subspace learning method named Discriminant Manifold Learning via Sparse Coding (DML_SC). In our method, we first decompose the input image into several components via dictionary learning, and then regroup the components into a More Important Part (MIP) and a Less Important Part (LIP). The MIP can be regarded as the clean part of the original image residing on a nonlinear submanifold, while LIP as noise in the image. Finally, the MIP and LIP are incorporated into manifold learning to learn a desired discriminative subspace. The proposed method is able to deal with data with and without labels, yielding supervised and unsupervised DML SCs. Experimental results show that DML_SC achieves best performance on image recognition and clustering tasks compared with well-known subspace learning and sparse representation methods.

Orthogonal enhanced linear discriminant analysis for face recognition

From the intuition that natural face images lie on or near a low-dimensional submanifold, the authors propose a novel spectral graph based dimensionality reduction method, named orthogonal enhanced linear discriminant analysis (OELDA), for face recognition. OELDA is based on enhanced LDA (ELDA), which takes into account both the discriminative structure and geometrical structure of the face space, and generates non-orthogonal basis vectors. However, a significant fact is that eliminating the dependence of basis vectors can promote more effective recognition of unseen face images. For this purpose, the authors seek to improve the ELDA scheme by imposing orthogonal constraints on the basis vectors. Experimental results on real-world face datasets show that, benefitting from orthogonality, OELDA has more locality preserving power and discriminative power than LDA and ELDA, and achieves the highest recognition rates among compared methods.