Multi-feature fusion method based on WOSF and MSE for four-class MI EEG identification

A multi-feature fusion decoding study for unilateral upper-limb fine motor imagery

To address the fact that the classical motor imagination paradigm has no noticeable effect on the rehabilitation training of upper limbs in patients after stroke and the corresponding feature extraction algorithm is limited to a single domain, this paper describes the design of a unilateral upper-limb fine motor imagination paradigm and the collection of data from 20 healthy people. It presents a feature extraction algorithm for multi-domain fusion and compares the common spatial pattern (CSP), improved multiscale permutation entropy (IMPE) and multi-domain fusion features of all participants through the use of decision tree, linear discriminant analysis, naive Bayes, a support vector machine, k-nearest neighbor and ensemble classification precision algorithms in the ensemble classifier. For the same subject, the average classification accuracy improvement of the same classifier for multi-domain feature extraction relative to CSP feature results went up by 1.52%. The average classification accuracy improvement of the same classifier went up by 32.87% relative to the IMPE feature classification results. This study's unilateral fine motor imagery paradigm and multi-domain feature fusion algorithm provide new ideas for upper limb rehabilitation after stroke.

A Novel In-Home Sleep Monitoring System Based on Fully Integrated Multichannel Front-End Chip and Its Multilevel Analyses

OBJECTIVE

A novel in-home sleep monitoring system with an 8-channel biopotential acquisition front-end chip is presented and validated via multilevel data analyses and comparision with advanced polysomnography.

METHODS AND PROCEDURES

The chip includes a cascaded low-noise programmable gain amplifier (PGA) and 24-bit [Formula: see text]-[Formula: see text] analog-to-digital converter (ADC). The PGA is based on three op-amp structure while the ADC adopts cascade of integrator feedforward and feedback (CIFF-B) architecture. An innovative chopper-modulated input-scaling-down technique enhances the dynamic range. The proposed system and commercial polysomnography were used for in-home sleep monitoring of 20 healthy participants. The consistency and significance of the two groups' data were analyzed.

RESULTS

Fabricated in 180 nm BCD technology, the input-referred noise, input impedance, common-mode rejection ratio, and dynamic range of the acquisition front-end chip were [Formula: see text]Vpp, 1.25 GN), 113.9 dB, and 119.8 dB. The kappa coefficients between the sleep stage labels of the three scorers were 0.80, 0.76, and 0.79. The consistency of the slowing index, multiscale entropy, and percentile features between the two devices reached 0.958, 0.885, and 0.834. The macro sleep architecture characteristics of the two devices were not significantly different (all p [Formula: see text] 0.05).

CONCLUSION

The proposed chip was applied to develop an in-home sleep monitoring system with significantly reduced size, power, and cost. Multilevel analyses demonstrated that this system collects stable and accurate in-home sleep data.

CLINICAL IMPACT

The proposed system can be applied for long-term in-home sleep monitoring outside of laboratory environments and sleep disorders screening that with low cost.

Multi-Hierarchical Fusion to Capture the Latent Invariance for Calibration-Free Brain-Computer Interfaces

Brain-computer interfaces (BCI) based motor imagery (MI) has become a research hotspot for establishing a flexible communication channel for patients with apoplexy or degenerative pathologies. Accurate decoding of motor imagery electroencephalography (MI-EEG) signals, while essential for effective BCI systems, is still challenging due to the significant noise inherent in the EEG signals and the lack of informative correlation between the signals and brain activities. The application of deep learning for EEG feature representation has been rarely investigated, nevertheless bringing improvements to the performance of motor imagery classification. This paper proposes a deep learning decoding method based on multi-hierarchical representation fusion (MHRF) on MI-EEG. It consists of a concurrent framework constructed of bidirectional LSTM (Bi-LSTM) and convolutional neural network (CNN) to fully capture the contextual correlations of MI-EEG and the spectral feature. Also, the stacked sparse autoencoder (SSAE) is employed to concentrate these two domain features into a high-level representation for cross-session and subject training guidance. The experimental analysis demonstrated the efficacy and practicality of the proposed approach using a public dataset from BCI competition IV and a private one collected by our MI task. The proposed approach can serve as a robust and competitive method to improve inter-session and inter-subject transferability, adding anticipation and prospective thoughts to the practical implementation of a calibration-free BCI system.