Upper limb motor assessment for stroke with force, muscle activation and interhemispheric balance indices based on sEMG and fNIRS.

Introduction: Upper limb rehabilitation assessment plays a pivotal role in the recovery process of stroke patients. The current clinical assessment tools often rely on subjective judgments of healthcare professionals. Some existing research studies have utilized physiological signals for quantitative assessments. However, most studies used single index to assess the motor functions of upper limb. The fusion of surface electromyography (sEMG) and functional near-infrared spectroscopy (fNIRS) presents an innovative approach, offering simultaneous insights into the central and peripheral nervous systems. Methods: We concurrently collected sEMG signals and brain hemodynamic signals during bilateral elbow flexion in 15 stroke patients with subacute and chronic stages and 15 healthy control subjects. The sEMG signals were analyzed to obtain muscle synergy based indexes including synergy stability index (SSI), closeness of individual vector (CV) and closeness of time profile (CT). The fNIRS signals were calculated to extract laterality index (LI). Results: The primary findings were that CV, SSI and LI in posterior motor cortex (PMC) and primary motor cortex (M1) on the affected hemisphere of stroke patients were significantly lower than those in the control group (p < 0.05). Moreover, CV, SSI and LI in PMC were also significantly different between affected and unaffected upper limb movements (p < 0.05). Furthermore, a linear regression model was used to predict the value of the Fugl-Meyer score of upper limb (FMul) (R2 = 0.860, p < 0.001). Discussion: This study established a linear regression model using force, CV, and LI features to predict FMul scale values, which suggests that the combination of force, sEMG and fNIRS hold promise as a novel method for assessing stroke rehabilitation.

Using different matrix factorization approaches to identify muscle synergy in stroke survivors.

Over the past several decades, many scholars have investigated muscle synergy as a promising tool for evaluating motor function. However, it is challenging to obtain favorable robustness using the general muscle synergy identification algorithms, namely non-negative matrix factorization (NMF), independent component analysis (ICA), and factor analysis (FA). Some scholars have proposed improved muscle synergy identification algorithms to overcome the shortcomings of these approaches, such as singular value decomposition NMF (SVD-NMF), sparse NMF (S-NMF), and multivariate curve resolution-alternating least squares (MCR-ALS). However, performance comparisons of these algorithms are seldom conducted. In this study, experimental electromyography (EMG) data collected from healthy individuals and stroke survivors were applied to assess the repeatability and intra-subject consistency of NMF, SVD-NMF, S-NMF, ICA, FA, and MCR-ALS. MCR-ALS presented higher repeatability and intra-subject consistencies than the other algorithms. More synergies and lower intra-subject consistencies were observed in stroke survivors than in healthy individuals. Thus, MCR-ALS is considered a favorable muscle synergy identification algorithm for patients with neural system disorders.

MCR-ALS-based muscle synergy extraction method combined with LSTM neural network for motion intention detection.

Introduction: The time-varying and individual variability of surface electromyographic signals (sEMG) can lead to poorer motor intention detection results from different subjects and longer temporal intervals between training and testing datasets. The consistency of using muscle synergy between the same tasks may be beneficial to improve the detection accuracy over long time ranges. However, the conventional muscle synergy extraction methods, such as non-negative matrix factorization (NMF) and principal component analysis (PCA) have some limitations in the field of motor intention detection, especially in the continuous estimation of upper limb joint angles. Methods: In this study, we proposed a reliable multivariate curve-resolved-alternating least squares (MCR-ALS) muscle synergy extraction method combined with long-short term memory neural network (LSTM) to estimate continuous elbow joint motion by using the sEMG datasets from different subjects and different days. The pre-processed sEMG signals were then decomposed into muscle synergies by MCR-ALS, NMF and PCA methods, and the decomposed muscle activation matrices were used as sEMG features. The sEMG features and elbow joint angular signals were input to LSTM to establish a neural network model. Finally, the established neural network models were tested by using sEMG dataset from different subjects and different days, and the detection accuracy was measured by correlation coefficient. Results: The detection accuracy of elbow joint angle was more than 85% by using the proposed method. This result was significantly higher than the detection accuracies obtained by using NMF and PCA methods. The results showed that the proposed method can improve the accuracy of motor intention detection results from different subjects and different acquisition timepoints. Discussion: This study successfully improves the robustness of sEMG signals in neural network applications using an innovative muscle synergy extraction method. It contributes to the application of human physiological signals in human-machine interaction.

Allocation of cognitive resources in cognitive processing of rhythmic visual stimuli before gait-related motor initiation.

Rhythmic visual cues can affect the allocation of cognitive resources during gait initiation (GI) and motor preparation. However, it is unclear how the input of rhythmic visual information modulates the allocation of cognitive resources and affects GI. The purpose of this study was to explore the effect of rhythmic visual cues on the dynamic allocation of cognitive resources by recording electroencephalographic (EEG) activity during exposure to visual stimuli. This study assessed event-related potentials (ERPs), event-related synchronization/desynchronization (ERS/ERD), and EEG microstates at 32 electrodes during presentation of non-rhythmic and rhythmic visual stimuli in 20 healthy participants. The ERP results showed that the amplitude of the C1 component was positive under exposure to rhythmic visual stimuli, while the amplitude of the N1 component was higher under exposure to rhythmic visual stimuli compared to their non-rhythmic counterparts. Within the first 200 ms of the onset of rhythmic visual stimuli, ERS in the theta band was highly pronounced in all brain regions analyzed. The results of microstate analysis showed that rhythmic visual stimuli were associated with an increase in cognitive processing over time, while non-rhythmic visual stimuli were associated with a decrease. Overall, these findings indicated that, under exposure to rhythmic visual stimuli, consumption of cognitive resources is lower during the first 200 ms of visual cognitive processing, but the consumption of cognitive resources gradually increases over time. After approximately 300 ms, cognitive processing of rhythmic visual stimuli consumes more cognitive resources than processing of stimuli in the non-rhythmic condition. This indicates that the former is more conducive to the completion of gait-related motor preparation activities, based on processing of rhythmic visual information during the later stages. This finding indicates that the dynamic allocation of cognitive resources is the key to improving gait-related movement based on rhythmic visual cues.

A Novel Muscle Synergy Extraction Method Used for Motor Function Evaluation of Stroke Patients: A Pilot Study.

In this paper, we present a novel muscle synergy extraction method based on multivariate curve resolution-alternating least squares (MCR-ALS) to overcome the limitation of the nonnegative matrix factorization (NMF) method for extracting non-sparse muscle synergy, and we study its potential application for evaluating motor function of stroke survivors. Nonnegative matrix factorization (NMF) is the most widely used method for muscle synergy extraction. However, NMF is susceptible to components' sparseness and usually provides inferior reliability, which significantly limits the promotion of muscle synergy. In this study, MCR-ALS was employed to extract muscle synergy from electromyography (EMG) data. Its performance was compared with two other matrix factorization algorithms, NMF and self-modeling mixture analysis (SMMA). Simulated data sets were utilized to explore the influences of the sparseness and noise on the extracted synergies. As a result, the synergies estimated by MCR-ALS were the most similar to true synergies as compared with SMMA and NMF. MCR-ALS was used to analyze the muscle synergy characteristics of upper limb movements performed by healthy (n = 11) and stroke (n = 5) subjects. The repeatability and intra-subject consistency were used to evaluate the performance of MCR-ALS. As a result, MCR-ALS provided much higher repeatability and intra-subject consistency as compared with NMF, which were important for the reliability of the motor function evaluation. The stroke subjects had lower intra-subject consistency and seemingly had more synergies as compared with the healthy subjects. Thus, MCR-ALS is a promising muscle synergy analysis method for motor function evaluation of stroke patients.

Effects of rhythmic visual cues on cognitive resources allocation characterized by electroencephalographic (EEG) features during human gait initiation.

Rhythmic visual cues are beneficial in gait initiation (GI) in Parkinson's disease patients with freezing of gait (FOG), however, the underlying neurophysiological mechanism remains poorly understood. The cognitive control modulated by visual cues during GI has been investigated and considered as a potential factor influencing automatic motor actions, but it is unclear how rhythmic visual cues affect cognitive resources demands during GI. The purpose of this study was to explore the effect of rhythmic visual cues on cognitive resources allocation by recording the anticipatory cerebral cortex electroencephalographic (EEG) activity during GI. Twenty healthy participants initiated gait in response to the rhythmic and non-rhythmic visual cues of stimulus presentation. We assessed the contingent negative variation (CNV) of averaged EEG data over 32 electrode positions during GI preparation, the results of which showed that the CNV was induced over prefrontal, frontal, central, and parietal regions in both rhythmic conditions and non-rhythmic conditions. Overall, different visual cues modulated the amplitude of CNV in the early and late stages of the GI preparation. Compared with the non-rhythmic condition, the CNV amplitude was lower in rhythmic condition over displayed regions precede the GI onset. In the late stage of GI preparation, it showed significant differences between the two conditions in prefrontal, frontal, and central regions, and the amplitude of CNV was lower under rhythmic condition. More to the point, the differences were more obvious in the late stage of GI preparation between the two conditions, which was closely associated with the cognitive resources. Therefore, the results indicate that less cognitive resources allocation is required to trigger GI under rhythmic visual cues compared with non-rhythmic visual cues. This study may provide a new insight into why rhythmic visual cues are more effective in improving GI ability compared to non-rhythmic visual cues.

THz spectral data analysis and components unmixing based on non-negative matrix factorization methods.

In many situations the THz spectroscopic data observed from complex samples represent the integrated result of several interrelated variables or feature components acting together. The actual information contained in the original data might be overlapping and there is a necessity to investigate various approaches for model reduction and data unmixing. The development and use of low-rank approximate nonnegative matrix factorization (NMF) and smooth constraint NMF (CNMF) algorithms for feature components extraction and identification in the fields of terahertz time domain spectroscopy (THz-TDS) data analysis are presented. The evolution and convergence properties of NMF and CNMF methods based on sparseness, independence and smoothness constraints for the resulting nonnegative matrix factors are discussed. For general NMF, its cost function is nonconvex and the result is usually susceptible to initialization and noise corruption, and may fall into local minima and lead to unstable decomposition. To reduce these drawbacks, smoothness constraint is introduced to enhance the performance of NMF. The proposed algorithms are evaluated by several THz-TDS data decomposition experiments including a binary system and a ternary system simulating some applications such as medicine tablet inspection. Results show that CNMF is more capable of finding optimal solutions and more robust for random initialization in contrast to NMF. The investigated method is promising for THz data resolution contributing to unknown mixture identification.

Terahertz spectroscopic investigation of human gastric normal and tumor tissues.

Human dehydrated normal and cancerous gastric tissues were measured using transmission time-domain terahertz spectroscopy. Based on the obtained terahertz absorption spectra, the contrasts between the two kinds of tissue were investigated and techniques for automatic identification of cancerous tissue were studied. Distinctive differences were demonstrated in both the shape and amplitude of the absorption spectra between normal and tumor tissue. Additionally, some spectral features in the range of 0.2~0.5 THz and 1~1.5 THz were revealed for all cancerous gastric tissues. To systematically achieve the identification of gastric cancer, principal component analysis combined with t-test was used to extract valuable information indicating the best distinction between the two types. Two clustering approaches, K-means and support vector machine (SVM), were then performed to classify the processed terahertz data into normal and cancerous groups. SVM presented a satisfactory result with less false classification cases. The results of this study implicate the potential of the terahertz technique to detect gastric cancer. The applied data analysis methodology provides a suggestion for automatic discrimination of terahertz spectra in other applications.