A tensor-based scheme for stroke patients’ motor imagery EEG analysis in BCI-FES rehabilitation training

Higher-dimension Tensor Completion via Low-rank Tensor Ring Decomposition

“…We compare our algorithm with some tensor completion algorithms which are similar to our algorithm, e.g., TT-WOPT [24] and CP-WOPT [1]. Moreover, some other state-of-the-art algorithms, e.g., FBCP [26] and FaLRTC [16] are also tested in the next experiments.…”

“…Tensor has been studied for more than a century, many methodologies have been proposed [14]. Moreover, tensor has been applied in various research field such as signal processing [6], machine learning [2], data completion [1], brain-computer interface (BCI) [16], etc.. CANDECOMP/PARAFAC (CP) decomposition [3] and Tucker decomposition [22] are the most popular and classic tensor decomposition models, of which many theories and applications have been proposed [21]. In recent years, a new theory system named tensor network has drawn people's attention and becomes a promising aspect of tensor methodology [5], [7].…”

Use of Electroencephalography Brain‐Computer Interface Systems as a Rehabilitative Approach for Upper Limb Function After a Stroke: A Systematic Review

“…A total of 248 articles were retrieved, but only 45 were selected for further review and critical reading, according to the previously established selection procedure. Finally, 13 of the 45 articles met the inclusion criteria [8,51‐62], and 32 were excluded [29,63‐93] (Table 4) (Figure 1). The methodological quality of the included articles, measured with the Critical Review Form, ranged between 6 and 15 (Table 5).…”

“…Exclusion Criteria Niazi et al [29] The sample was composed of healthy subjects Tam et al [63] Intervention effectiveness was not analyzed; authors studied channel selection parameters on the classification accuracy of BCI Tan et al [64] The intervention effectiveness was not analyzed with motor outcome measures; authors intended only to demonstrate that movement intention was detectable during a training session for using BCI Cincotti et al [65] This study did not use motor outcome measures; the authors analyzed only the BCI systems ability to induce cortical plasticity Kasashima et al [66] This study analyzed the stroke patients' ability to use EEG-based BCI to detect motor imagery Ang et al [67] Motor outcome measures were not used to analyze intervention effectiveness; the authors considered only the changes in cortical excitability Gómez-Rodriguez et al [68] Motor outcome measures were not used; this paper demonstrated that artificially closing the sensoriomotor feedback loop facilitates decoding of movement intention by means of a BCI system Lew et al [69] This study demonstrates successful single-trial detection of movement intention from EEG Arvaneh et al [70] This article presents research on the need for a calibration session for long-term BCI users Arvaneh et al [71] The authors propose a novel algorithm for EEG-BCI to extract features of EEG signals Bundy et al [72] This study sought to evaluate whether stroke survivors could achieve BCI control with motor activity from their unaffected hemisphere Aono et al [73] This study presented research regarding the relationship between ERD magnitude and cortical excitability in healthy subjects and subjects with stroke Ang et al [74] This study presented research on the feasibility of using EEG calibration data from passive movement to detect motor imagery Leamy et al [75] Motor outcome measures were not used; this study did not analyze the intervention effectiveness for functional improvements, only neuroplastic changes associated with the recovery process Liu et al [76] Motor outcome measures were not used; the authors developed a scheme to detect motor imagery EEG patterns Petti et al [77] The authors proposed an analysis based on data acquired from stroke patients, using an EEG-BCI based on motor imagery Schreuder et al [78] This study did not develop an intervention; the authors researched how to develop a suitable user-centered BCI design Takemi et al [79] This was a research study on the relationship between ERD magnitude and cortical excitability in healthy subjects Bermudez et al [80] The sample was composed of healthy subjects Ang et al [81] ...…”

Multi-phase cycle coding for SSVEP based brain-computer interfaces

“…It is an effective neural source compared to those in other BCI systems (e.g. motor imagery [2, 3], P300 [4, 5], and slow cortical potentials [6]), since it can achieve higher information transfer rate (ITR) with shorter time response [7, 8]. …”