Continuous Force Decoding from Local Field Potentials of the Primary Motor Cortex in Freely Moving Rats

Degradation-aware neural imputation: Advancing decoding stability in brain machine interfaces

Neural signal degradation poses a significant challenge in maintaining stable performance when decoding motor tasks using multiunit activity (MUA) and local field potential (LFP) signals in the implantable brain machine interface (iBMI) applications. Effective methods for imputing degraded or missing signals are essential to restore neural signal integrity, thereby improving decoding accuracy and system robustness over long-term recordings with fluctuating signal quality. This study introduces a confidence-weighted Bayesian linear regression (CW-BLR) approach to impute neural signals affected by degradation, enhancing the robustness and consistency of decoding. The performance of CW-BLR was compared to traditional methods-mean imputation (Mean-imp) and Gaussian-mixture-model-based expectation-maximization (GMM-EM)-using a kernel-sliced inverse regression (kSIR) decoder to evaluate decoding outcomes. Four Wistar rats were trained to perform a forelimb-reaching task while neural activity (MUA and LFPs) was recorded over 27 days. CW-BLR imputed signals degraded during days 8-27. Decoding performance was evaluated using kSIR and compared with Mean-imp and GMM-EM. CW-BLR demonstrated superior performance by effectively preserving both temporal and spatial dependencies within the neural signals. CW-BLR-imputed data significantly improved decoding accuracy over traditional imputation methods, with the kSIR decoder showing consistently higher performance, particularly in maintaining signal quality from the degraded period. CW-BLR offers a robust and effective imputation framework for iBMI applications, addressing signal degradation challenges and maintaining accurate decoding over prolonged recordings. By utilizing confidence-based quality metrics, CW-BLR surpasses traditional methods, providing stable neural decoding across fluctuating signal quality scenarios.

Decoding Single-Pellet Retrieval Task From Local Field Potentials in Pre- and Post-Stroke Motor Areas: Insights Into Interhemispheric Connectivity Difference

OBJECTIVE

Intracortical brain-machine interfaces (iBMIs) hold promise for restoring communication and movement in stroke-paralyzed individuals. Recent studies have demonstrated the potential of using local field potentials (LFPs) for decoding single-pellet retrieval (SPR) tasks in iBMIs. However, most research has relied on LFPs from healthy rats rather than those affected by stroke. This study aimed to investigate the feasibility of utilizing LFPs from both the right and left (stroke) cortical forelimb areas (CFAs) for the SPR tasks decoding under both pre- and post-stroke conditions.

METHODS

LFPs were recorded via microelectrode arrays implanted into CFAs of eight rats trained to perform the SPR tasks. The relative spectral power method was used to represent frequency information, and random forest classification differentiated SPR tasks from resting states. We also assessed interhemispheric connectivity, including correlation, coherence, and phase-amplitude coupling (PAC), to compare differences between the SPR tasks and the resting states under both pre- and post-stroke conditions.

RESULTS

Our findings indicated that the relative PS method with LFPs achieves 87.10% 9.2% accuracy in post-stoke SPR decoding, where high gamma is crucial. Additionally, we observed changes in PACs from the right to the left sensorimotor cortex post-stroke during the SPR tasks compared to the resting states.

SIGNIFICANCE

Our work provides a comprehensive insight into the role of different frequency band from LFPs in motor function recovery mechanisms, highlighting the importance of the high gamma in motor function. This research lays the foundation for developing post-stoke SPR-related BMIs.

Human cortical high-gamma power scales with movement rate in healthy participants and stroke survivors

Motor cortical high-gamma oscillations (60-90 Hz) occur at movement onset and are spatially focused over the contralateral primary motor cortex. Although high-gamma oscillations are widely recognized for their significance in human motor control, their precise function on a cortical level remains elusive. Importantly, their relevance in human stroke pathophysiology is unknown. Because motor deficits are fundamental determinants of symptom burden after stroke, understanding the neurophysiological processes of motor coding could be an important step in improving stroke rehabilitation. We recorded magnetoencephalography data during a thumb movement rate task in 14 chronic stroke survivors, 15 age-matched control participants and 29 healthy young participants. Motor cortical high-gamma oscillations showed a strong relation with movement rate as trials with higher movement rate were associated with greater high-gamma power. Although stroke survivors showed reduced cortical high-gamma power, this reduction primarily reflected the scaling of high-gamma power with movement rate, yet after matching movement rate in stroke survivors and age-matched controls, the reduction of high-gamma power exceeded the effect of their decreased movement rate alone. Even though motor skill acquisition was evident in all three groups, it was not linked to high-gamma power. Our study quantifies high-gamma oscillations after stroke, revealing a reduction in movement-related high-gamma power. Moreover, we provide strong evidence for a pivotal role of motor cortical high-gamma oscillations in encoding movement rate. KEY POINTS: Neural oscillations in the high-gamma frequency range (60-90 Hz) emerge in the human motor cortex during movement. The precise function of these oscillations in motor control remains unclear, and they have never been characterized in stroke survivors. In a magnetoencephalography study, we demonstrate that high-gamma oscillations in motor cortical areas scale with movement rate, and we further explore their temporal and spatial characteristics. Stroke survivors exhibit lower high-gamma power during movement than age-matched control participants, even after matching for movement rate. The results contribute to the understanding of the role of high-gamma oscillations in motor control and have important implications for neuromodulation in stroke rehabilitation.

Protocol for state-based decoding of hand movement parameters using neural signals

We present a protocol for decoding kinematic and kinetic parameters from the primary somatosensory cortex during active and passive hand movements in a center-out reaching task using state-based and conventional decoders. We describe steps for preparing data and using the state-based model to classify movement directions into states via feature extraction and predict parameters with regression models (partial least squares and multilinear regression) trained per state. This state-based approach outperforms conventional methods, enhancing accuracy for brain-computer interface applications. For complete details on the use and execution of this protocol, please refer to Mirfathollahi et al.1.

Depletion of complement factor 3 delays the neuroinflammatory response to intracortical microelectrodes

The neuroinflammatory response to intracortical microelectrodes (IMEs) used with brain-machine interfacing (BMI) applications is regarded as the primary contributor to poor chronic performance. Recent developments in high-plex gene expression technologies have allowed for an evolution in the investigation of individual proteins or genes to be able to identify specific pathways of upregulated genes that may contribute to the neuroinflammatory response. Several key pathways that are upregulated following IME implantation are involved with the complement system. The complement system is part of the innate immune system involved in recognizing and eliminating pathogens - a significant contributor to the foreign body response against biomaterials. Specifically, we have identified Complement 3 (C3) as a gene of interest because it is the intersection of several key complement pathways. In this study, we investigated the role of C3 in the IME inflammatory response by comparing the neuroinflammatory gene expression at the microelectrode implant site between C3 knockout (C3-/-) and wild-type (WT) mice. We have found that, like in WT mice, implantation of intracortical microelectrodes in C3-/- mice yields a dramatic increase in the neuroinflammatory gene expression at all post-surgery time points investigated. However, compared to WT mice, C3 depletion showed reduced expression of many neuroinflammatory genes pre-surgery and 4 weeks post-surgery. Conversely, depletion of C3 increased the expression of many neuroinflammatory genes at 8 weeks and 16 weeks post-surgery, compared to WT mice. Our results suggest that C3 depletion may be a promising therapeutic target for acute, but not chronic, relief of the neuroinflammatory response to IME implantation. Additional compensatory targets may also be required for comprehensive long-term reduction of the neuroinflammatory response for improved intracortical microelectrode performance.

Therapeutic Viewpoint on Rat Models of Locomotion Abnormalities and Neurobiological Indicators in Parkinson's Disease

BACKGROUND

Locomotion problems in Parkinson's syndrome are still a research and treatment difficulty. With the recent introduction of brain stimulation or neuromodulation equipment that is sufficient to monitor activity in the brain using electrodes placed on the scalp, new locomotion investigations in patients having the capacity to move freely have sprung up.

OBJECTIVE

This study aimed to find rat models and locomotion-connected neuronal indicators and use them all over a closed-loop system to enhance the future and present treatment options available for Parkinson's disease.

METHODS

Various publications on locomotor abnormalities, Parkinson's disease, animal models, and other topics have been searched using several search engines, such as Google Scholar, Web of Science, Research Gate, and PubMed.

RESULTS

Based on the literature, we can conclude that animal models are used for further investigating the locomotion connectivity deficiencies of many biological measuring devices and attempting to address unanswered concerns from clinical and non-clinical research. However, translational validity is required for rat models to contribute to the improvement of upcoming neurostimulation-based medicines. This review discusses the most successful methods for modelling Parkinson's locomotion in rats.

CONCLUSION

This review article has examined how scientific clinical experiments lead to localised central nervous system injuries in rats, as well as how the associated motor deficits and connection oscillations reflect this. This evolutionary process of therapeutic interventions may help to improve locomotion- based treatment and management of Parkinson's syndrome in the upcoming years.

The effect of a Mn(III)tetrakis(4-benzoic acid)porphyrin (MnTBAP) coating on the chronic recording performance of planar silicon intracortical microelectrode arrays

Intracortical microelectrode arrays (MEAs) are used to record neural activity. However, their implantation initiates a neuroinflammatory cascade, involving the accumulation of reactive oxygen species, leading to interface failure. Here, we coated commercially-available MEAs with Mn(III)tetrakis(4-benzoic acid)porphyrin (MnTBAP), to mitigate oxidative stress. First, we assessed the in vitro cytotoxicity of modified sample substrates. Then, we implanted 36 rats with uncoated, MnTBAP-coated ("Coated"), or (3-Aminopropyl)triethoxysilane (APTES)-coated devices - an intermediate step in the coating process. We assessed electrode performance during the acute (1-5 weeks), sub-chronic (6-11 weeks), and chronic (12-16 weeks) phases after implantation. Three subsets of animals were euthanized at different time points to assess the acute, sub-chronic and chronic immunohistological responses. Results showed that MnTBAP coatings were not cytotoxic in vitro, and their implantation in vivo improved the proportion of electrodes during the sub-chronic and chronic phases; APTES coatings resulted in failure of the neural interface during the chronic phase. In addition, MnTBAP coatings improved the quality of the signal throughout the study and reduced the neuroinflammatory response around the implant as early as two weeks, an effect that remained consistent for months post-implantation. Together, these results suggest that MnTBAP coatings are a potentially useful modification to improve MEA reliability.

Decoding hand kinetics and kinematics using somatosensory cortex activity in active and passive movement

Area 2 of the primary somatosensory cortex (S1), encodes proprioceptive information of limbs. Several studies investigated the encoding of movement parameters in this area. However, the single-trial decoding of these parameters, which can provide additional knowledge about the amount of information available in sub-regions of this area about instantaneous limb movement, has not been well investigated. We decoded kinematic and kinetic parameters of active and passive hand movement during center-out task using conventional and state-based decoders. Our results show that this area can be used to accurately decode position, velocity, force, moment, and joint angles of hand. Kinematics had higher accuracies compared to kinetics and active trials were decoded more accurately than passive trials. Although the state-based decoder outperformed the conventional decoder in the active task, it was the opposite in the passive task. These results can be used in intracortical micro-stimulation procedures to provide proprioceptive feedback to BCI subjects.

Differential expression of genes involved in the chronic response to intracortical microelectrodes

Brain-Machine Interface systems (BMIs) are clinically valuable devices that can provide functional restoration for patients with spinal cord injury or improved integration for patients requiring prostheses. Intracortical microelectrodes can record neuronal action potentials at a resolution necessary for precisely controlling BMIs. However, intracortical microelectrodes have a demonstrated history of progressive decline in the recording performance with time, inhibiting their usefulness. One major contributor to decreased performance is the neuroinflammatory response to the implanted microelectrodes. The neuroinflammatory response can lead to neurodegeneration and the formation of a glial scar at the implant site. Historically, histological imaging of relatively few known cellular and protein markers has characterized the neuroinflammatory response to implanted microelectrode arrays. However, neuroinflammation requires many molecular players to coordinate the response - meaning traditional methods could result in an incomplete understanding. Taking advantage of recent advancements in tools to characterize the relative or absolute DNA/RNA expression levels, a few groups have begun to explore gene expression at the microelectrode-tissue interface. We have utilized a custom panel of ∼813 neuroinflammatory-specific genes developed with NanoString for bulk tissue analysis at the microelectrode-tissue interface. Our previous studies characterized the acute innate immune response to intracortical microelectrodes. Here we investigated the gene expression at the microelectrode-tissue interface in wild-type (WT) mice chronically implanted with nonfunctioning probes. We found 28 differentially expressed genes at chronic time points (4WK, 8WK, and 16WK), many in the complement and extracellular matrix system. Further, the expression levels were relatively stable over time. Genes identified here represent chronic molecular players at the microelectrode implant sites and potential therapeutic targets for the long-term integration of microelectrodes. STATEMENT OF SIGNIFICANCE: Intracortical microelectrodes can record neuronal action potentials at a resolution necessary for the precise control of Brain-Machine Interface systems (BMIs). However, intracortical microelectrodes have a demonstrated history of progressive declines in the recording performance with time, inhibiting their usefulness. One major contributor to the decline in these devices is the neuroinflammatory response against the implanted microelectrodes. Historically, neuroinflammation to implanted microelectrode arrays has been characterized by histological imaging of relatively few known cellular and protein markers. Few studies have begun to develop a more in-depth understanding of the molecular pathways facilitating device-mediated neuroinflammation. Here, we are among the first to identify genetic pathways that could represent targets to improve the host response to intracortical microelectrodes, and ultimately device performance.

Enhancing Prediction of Forelimb Movement Trajectory through a Calibrating-Feedback Paradigm Incorporating RAT Primary Motor and Agranular Cortical Ensemble Activity in the Goal-Directed Reaching Task

Complete reaching movements involve target sensing, motor planning, and arm movement execution, and this process requires the integration and communication of various brain regions. Previously, reaching movements have been decoded successfully from the motor cortex (M1) and applied to prosthetic control. However, most studies attempted to decode neural activities from a single brain region, resulting in reduced decoding accuracy during visually guided reaching motions. To enhance the decoding accuracy of visually guided forelimb reaching movements, we propose a parallel computing neural network using both M1 and medial agranular cortex (AGm) neural activities of rats to predict forelimb-reaching movements. The proposed network decodes M1 neural activities into the primary components of the forelimb movement and decodes AGm neural activities into internal feedforward information to calibrate the forelimb movement in a goal-reaching movement. We demonstrate that using AGm neural activity to calibrate M1 predicted forelimb movement can improve decoding performance significantly compared to neural decoders without calibration. We also show that the M1 and AGm neural activities contribute to controlling forelimb movement during goal-reaching movements, and we report an increase in the power of the local field potential (LFP) in beta and gamma bands over AGm in response to a change in the target distance, which may involve sensorimotor transformation and communication between the visual cortex and AGm when preparing for an upcoming reaching movement. The proposed parallel computing neural network with the internal feedback model improves prediction accuracy for goal-reaching movements.

Interhemispheric neural characteristics of noxious mechano-nociceptive stimulation in the anterior cingulate cortex

Background

Pain is an unpleasant sensory and emotional experience. One of the most critical regions of the brain for pain processing is the anterior cingulate cortex (ACC). Several studies have examined the role of this region in thermal nociceptive pain. However, studies on mechanical nociceptive pain have been very limited to date. Although several studies have investigated pain, the interactions between the two hemispheres are still not clear. This study aimed to investigate nociceptive mechanical pain in the ACC bilaterally.

Methods

Local field potential (LFP) signals were recorded from seven male Wistar rats' ACC regions of both hemispheres. Mechanical stimulations with two intensities, high-intensity noxious (HN) and non-noxious (NN) were applied to the left hind paw. At the same time, the LFP signals were recorded bilaterally from awake and freely moving rats. The recorded signals were analyzed from different perspectives, including spectral analysis, intensity classification, evoked potential (EP) analysis, and synchrony and similarity of two hemispheres.

Results

By using spectro-temporal features and support vector machine (SVM) classifier, HN vs. no-stimulation (NS), NN vs. NS, and HN vs. NN were classified with accuracies of 89.6, 71.1, and 84.7%, respectively. Analyses of the signals from the two hemispheres showed that the EPs in the two hemispheres were very similar and occurred simultaneously; however, the correlation and phase locking value (PLV) between the two hemispheres changed significantly after HN stimulation. These variations persisted for up to 4 s after the stimulation. In contrast, variations in the PLV and correlation for NN stimulation were not significant.

Conclusions

This study showed that the ACC area was able to distinguish the intensity of mechanical stimulation based on the power activities of neural responses. In addition, our results suggest that the ACC region is activated bilaterally due to nociceptive mechanical pain. Additionally, stimulations above the pain threshold (HN) significantly affect the synchronicity and correlation between the two hemispheres compared to non-noxious stimuli.

Continuous Decoding of Hand Movement From EEG Signals Using Phase-Based Connectivity Features

The principal goal of the brain-computer interface (BCI) is to translate brain signals into meaningful commands to control external devices or neuroprostheses to restore lost functions of patients with severe motor disabilities. The invasive recording of brain signals involves numerous health issues. Therefore, BCIs based on non-invasive recording modalities such as electroencephalography (EEG) are safer and more comfortable for the patients. The BCI requires reconstructing continuous movement parameters such as position or velocity for practical application of neuroprostheses. The BCI studies in continuous decoding have extensively relied on extracting features from the amplitude of brain signals, whereas the brain connectivity features have rarely been explored. This study aims to investigate the feasibility of using phase-based connectivity features in decoding continuous hand movements from EEG signals. To this end, the EEG data were collected from seven healthy subjects performing a 2D center-out hand movement task in four orthogonal directions. The phase-locking value (PLV) and magnitude-squared coherence (MSC) are exploited as connectivity features along with multiple linear regression (MLR) for decoding hand positions. A brute-force search approach is employed to find the best channel pairs for extracting features related to hand movements. The results reveal that the regression models based on PLV and MSC features achieve the average Pearson correlations of 0.43 ± 0.03 and 0.42 ± 0.06, respectively, between predicted and actual trajectories over all subjects. The delta and alpha band features have the most contribution in regression analysis. The results also demonstrate that both PLV and MSC decoding models lead to superior results on our data compared to two recently proposed feature extraction methods solely based on the amplitude or phase of recording signals (p < 0.05). This study verifies the ability of PLV and MSC features in the continuous decoding of hand movements with linear regression. Thus, our findings suggest that extracting features based on brain connectivity can improve the accuracy of trajectory decoder BCIs.

Closed-loop stimulation using a multiregion brain-machine interface has analgesic effects in rodents

Effective treatments for chronic pain remain limited. Conceptually, a closed-loop neural interface combining sensory signal detection with therapeutic delivery could produce timely and effective pain relief. Such systems are challenging to develop because of difficulties in accurate pain detection and ultrafast analgesic delivery. Pain has sensory and affective components, encoded in large part by neural activities in the primary somatosensory cortex (S1) and anterior cingulate cortex (ACC), respectively. Meanwhile, studies show that stimulation of the prefrontal cortex (PFC) produces descending pain control. Here, we designed and tested a brain-machine interface (BMI) combining an automated pain detection arm, based on simultaneously recorded local field potential (LFP) signals from the S1 and ACC, with a treatment arm, based on optogenetic activation or electrical deep brain stimulation (DBS) of the PFC in freely behaving rats. Our multiregion neural interface accurately detected and treated acute evoked pain and chronic pain. This neural interface is activated rapidly, and its efficacy remained stable over time. Given the clinical feasibility of LFP recordings and DBS, our findings suggest that BMI is a promising approach for pain treatment.

Subthalamic nucleus deep brain stimulation driven by primary motor cortex γ2 activity in parkinsonian monkeys

In parkinsonism, subthalamic nucleus (STN) electrical deep brain stimulation (DBS) improves symptoms, but may be associated with side effects. Adaptive DBS (aDBS), which enables modulation of stimulation, may limit side effects, but limited information is available about clinical effectiveness and efficaciousness. We developed a brain-machine interface for aDBS, which enables modulation of stimulation parameters of STN-DBS in response to γ2 band activity (80-200 Hz) of local field potentials (LFPs) recorded from the primary motor cortex (M1), and tested its effectiveness in parkinsonian monkeys. We trained two monkeys to perform an upper limb reaching task and rendered them parkinsonian with 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine. Bipolar intracortical recording electrodes were implanted in the M1, and a recording chamber was attached to access the STN. In aDBS, the M1 LFPs were recorded, filtered into the γ2 band, and discretized into logic pulses by a window discriminator, and the pulses were used to modulate the interval and amplitude of DBS pulses. In constant DBS (cDBS), constant stimulus intervals and amplitudes were used. Reaction and movement times during the task were measured and compared between aDBS and cDBS. The M1-γ2 activities were increased before and during movements in parkinsonian monkeys and these activities modulated the aDBS pulse interval, amplitude, and dispersion. With aDBS and cDBS, reaction and movement times were significantly decreased in comparison to DBS-OFF. The electric charge delivered was lower with aDBS than cDBS. M1-γ2 aDBS in parkinsonian monkeys resulted in clinical benefits that did not exceed those from cDBS. However, M1-γ2 aDBS achieved this magnitude of benefit for only two thirds of the charge delivered by cDBS. In conclusion, M1-γ2 aDBS is an effective therapeutic approach which requires a lower electrical charge delivery than cDBS for comparable clinical benefits.

A Brain Biometric-based Identification Approach Using Local Field Potentials

Traditional biometrics such as face, iris and fingerprint have been applied widely nowadays. Nevertheless, with more and more potential problems being exposed, such as privacy leak and fabricate attack, it is urgent to find new secure biometrics to meet the needs. Identification based on brain signals is a promising option due to its unique advantages of confidentiality, anti-spoofing, continuity and cancelability. Among various types of brain signals, local field potential (LFP) has long term stability, high signal to noise ratio and high spatial resolution, which is suitable for identification. In this paper, we propose a novel biometric which is extracted from LFP signals with a deep neural network. The proposed biometric can be generated in a task-related manner thus is cancelable. Experiments with ten rats demonstrate that, the proposed biometric achieves a high identification accuracy of 94.47%, and the performance is stable over several days.

A New Spike Sorting Algorithm Based on Continuous Wavelet Transform and Investigating Its Effect on Improving Neural Decoding Accuracy

Spike sorting is an essential step in extracting neuronal discharge patterns which help to decode different activities in the neural system. Therefore, improving the spike sorting accuracy can improve neural decoding performance subsequently. Although many methods are suggested for spike sorting, few studies have evaluated their effect on neural decoding performance. In this paper, a method of spike sorting based on an optimized selection of the parameters in the continuous wavelet transform (CWT) is proposed. The proposed algorithm was tested on a simulated dataset and two publicly available benchmark datasets to evaluate its performance in spike sorting. To evaluate the effect of utilizing different spike sorting algorithms on neural decoding performance, real data was used in which the aim was to decode the force applied by the rat's hand to a pedal continuously from the intra-cortical data of the primary motor area of the cortex. The extracted neuronal firing rates by the spike sorting algorithms were applied to a partial least squares regression to decode the force signal. In the simulation study, the proposed spike sorting algorithm based on optimized wavelet parameter selection outperformed both the WaveClus spike sorting and traditional PCA-based spike sorting algorithms. The results showed the superiority of the spike sorting algorithm based on optimal wavelet parameters compared to classical discrete wavelet transform (DWT) or PCA-based spike sorting methods in decoding real intracortical data. Overall, the results indicate that it is possible to improve neural decoding performance by improving the spike sorting accuracy.

Force Decoding of Caudal Forelimb Area and Rostral Forelimb Area in Chronic Stroke Rats

Brain machine interfaces (BMIs) used for movement restoration primarily rely on studies of motor decoding. It has been proved that local field potentials (LFPs) from primary motor cortex and premotor cortex of normal rodents could be used for decoding motor signals. However, few studies have explored the decoding performance of these brain areas under motor cortex damage. In this work, we focus on force decoding performance of LFPs spectrum from both ipsilesional caudal forelimb area (CFA) and rostral forelimb area (RFA) of rodents with ischemia over CFA. After three months of ischemia induced by photothrombosis over CFA, the power of high-frequency bands (>120 Hz) from both CFA and RFA can decode force signals by Kalman filters. The fair performance of CFA indicates motor reorganization over penumbra. Further exploration of RFA decoding ability proves that at least four electrodes of RFA should be used on decoding and electrodes far from CFA of stroke rats could achieve almost as good results as those close to CFA of normal rats, which indicates the motor remapping. Experimental results show the long-term stability of PM LFPs decoding performance of stroke rats as the trained Kalman model could be used to accurately decode force some days later which provides a possibility for online decoding system. In conclusion, our work shows that even under CFA ischemia, high-frequency power of LFPs from RFA is still able to accurately decode force signals and has long stability, which provides the possibility of BMIs for motor function reconstruction of chronic stroke patients.

A stack LSTM structure for decoding continuous force from local field potential signal of primary motor cortex (M1)

Background

Brain Computer Interfaces (BCIs) translate the activity of the nervous system to a control signal which is interpretable for an external device. Using continuous motor BCIs, the user will be able to control a robotic arm or a disabled limb continuously. In addition to decoding the target position, accurate decoding of force amplitude is essential for designing BCI systems capable of performing fine movements like grasping. In this study, we proposed a stack Long Short-Term Memory (LSTM) neural network which was able to accurately predict the force amplitude applied by three freely moving rats using their Local Field Potential (LFP) signal.

Results

The performance of the network was compared with the Partial Least Square (PLS) method. The average coefficient of correlation (r) for three rats were 0.67 in PLS and 0.73 in LSTM based network and the coefficient of determination (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R^{2}$$\end{document}R2) were 0.45 and 0.54 for PLS and LSTM based network, respectively. The network was able to accurately decode the force values without explicitly using time lags in the input features. Additionally, the proposed method was able to predict zero-force values very accurately due to benefiting from an output nonlinearity.

Conclusion

The proposed stack LSTM structure was able to predict applied force from the LFP signal accurately. In addition to higher accuracy, these results were achieved without explicitly using time lags in input features which can lead to more accurate and faster BCI systems.

Regularized Partial Least Square Regression for Continuous Decoding in Brain-Computer Interfaces

Continuous decoding is a crucial step in many types of brain-computer interfaces (BCIs). Linear regression techniques have been widely used to determine a linear relation between the input and desired output. A serious issue in this technique is the over-fitting phenomenon. Partial least square (PLS) is a well-known and popular method which tries to overcome this problem. PLS calculates a set of latent variables which are maximally correlated to the output and determines a linear relation between a low-rank estimation of the input and output data. However, this method has shown its potential to overfit the training data in many cases. In this paper, a regularized version of PLS (RPLS) is proposed which tries to determine a linear relation between the latent vector of the input and desired output using the regularized least square instead of the ordinary one. This approach is able to control the effect of non-efficient and non-generalized latent vectors in prediction. We have shown that the proposed method outperforms Ridge regression (RR), PLS, and PLS with regularized weights (PLSRW) in estimating the output in two different real BCI datasets, Neurotycho public electrocorticogram (ECoG) dataset for decoding trajectory of hand movements in monkeys and our own local field potential (LFP) dataset for decoding applied force performed by rats. Furthermore, the results indicate that RPLS is more robust against the increase in the number of latent vectors compared to PLS and PLSRW. Next, we evaluated the resistance of our proposed method against the presence of different noise levels in a BCI application and compared it to other techniques using a semi-simulated dataset. This approach revealed that RPLS offered a higher performance compared with other techniques in all levels of noise. Finally, to illustrate the usability of RPLS in other type of data, we presented the application of this method in predicting relative active substance content of pharmaceutical tablets using near-infrared (NIR) transmittance spectroscopy data. This application showed a superior performance of the proposed method compared to other decoding methods.

Regularized Kalman filter for brain-computer interfaces using local field potential signals

BACKGROUND

Brain-computer interfaces (BCIs) seek to establish a direct connection from brain to computer, to use in applications such as motor prosthesis control, control of a cursor on the monitor, and so on. Hence, the accuracy of movement decoding from brain signals in BCIs is crucial. The Kalman filter (KF) is often used in BCI systems to decode neural activity and estimate kinetic and kinematic parameters. To use the KF, the state transition matrix, the observation matrix and the covariance matrices of the process and measurement noises must be known in advance, however, in many applications these matrices are not known. Typically, to estimate these parameters, the ordinary least squares method and the sample covariance matrix estimator are used. Our purpose is to enhance the decoding performance of the KF in BCI systems by improving the estimation of the mentioned parameters.

NEW METHOD

Here, we propose the Regularized Kalman Filter (RKF) which implements two fundamental features: 1) Regularizing the regression estimate of the state equation to improve the estimation of the state transition matrix, and 2) Use of shrinkage method to improve the estimation of the unknown measurement noise covariance matrix. We validated the performance of the proposed method using two datasets of local field potentials obtained from motor cortex of a monkey (Estimation of kinematic parameters during hand movement) and three rats (Estimation of the amount of force applied by hand as a kinetic parameter).

RESULTS

The results demonstrate that the proposed method outperforms the conventional KF, the KF with feature selection, the Partial least squares, and the Ridge regression approaches.

Bimetallic Catalyzed N-arylation Used in Synthesis of Novel β-carbolines Derivatives

Background:

Natural occurring β-Carbolines alkaloids are abundant in the plant kingdom or other organisms, and they were found to possess good antitumor activity through multiple mechanisms. Based on previous summarized SARs of β-carboline derivatives, the modification on pyridine ring would have a great impact on their antitumor activities. Therefore, we plan to synthesized arylated β-carboline-3-amides to find more valuable β-Carbolines derivatives.

Methods:

A novel bimetallic Pd(OAc)2/AgOAc catalyst system was developed for the amidation of aryl iodides under acid condition. A series of N-arylated β-carbolines derivatives were synthesized using this method. The structures of these compounds were confirmed by 1H NMR, 13C NMR and HRMS, and their in vitro antiproliferative activity was investigated against HepG2 and Hela tumor cell lines by MTT assay.

Results:

Eleven N-arylated β-carboline-3-amides were synthesized using this bimetallic catalyzed method in 58-98% yields. These synthesized N-arylated compounds showed no antiproliferative activity at 20 μM.

Conclusion:

We have discovered an efficient and bimetallic catalytic system allowing the Narylation of secondary acyclic amides. This is the first report that N-arylation of aliphatic secondary acyclic amides under acid condition.

Force decoding using local field potentials in primary motor cortex: PLS or Kalman filter regression?

The development of brain-computer interface (BCI) systems is an important approach in brain studies. Control of communication devices and prostheses in real-world scenarios requires complex movement parameters. Decoding a variety of neural signals captured by micro-wire arrays is a potential applicant for extracting movement-related information. The present work was conducted to compare the functionality of partial least square (PLS) regression and Kalman filter to predict the force parameter from local field potential (LFP) signals of the primary motor cortex (M1). The signals were recorded using a 16-channel micro-wire array from the forelimb-related area of the M1 of three rats performing a behavioral task in which the force signal of the rat's forelimb paw was generated. Our results show that PLS regression and Kalman filters with the mean performance of 0.75 and 0.72 in terms of the correlation coefficient (CC) and 0.37 and 0.48 in terms of normalized mean square error (NMSE), respectively, are effective methods for decoding the force parameter from LFPs. Kalman filter underperforms PLS both in performance and speed. Although adding nonlinearity to the Kalman filter results in equally accurate CC performance as PLS, it has even more computational cost. Therefore, it is inferred that nonlinear methods do not necessarily have better functionality than linear ones and PLS, as a simple fast linear method could be an effectively applicable regression technique for BCIs.

Chronic stability of single-channel neurophysiological correlates of gross and fine reaching movements in the rat

While substantial task-related neural activity has been observed during motor tasks in rodent primary motor cortex and premotor cortex, the long-term stability of these responses in healthy rats is uncertain, limiting the interpretability of longitudinal changes in the specific patterns of neural activity associated with learning or motor recovery following injury. This study examined the stability of task-related neural activity associated with execution of two distinct reaching tasks in healthy rodents. A novel automated rodent behavioral apparatus was constructed and rats were trained to perform a reaching task combining a ‘gross’ lever press and a ‘fine’ pellet retrieval. In each animal, two chronic microelectrode arrays were implanted in motor cortex spanning the caudal forelimb area (rodent primary motor cortex) and the rostral forelimb area (rodent premotor cortex). We recorded multiunit spiking and local field potential activity from 10 days to 7–10 weeks post-implantation to characterize the patterns of neural activity observed during each task component and analyzed the consistency of channel-specific task-related neural activity. Task-related changes in neural activity were observed on the majority of channels. While the task-related changes in multi-unit spiking and local field potential spectral power were consistent over several weeks, spectral power changes were more stable, despite the trade-off of decreased spatial and temporal resolution. These results show that neural activity in rodent primary and premotor cortex is associated with specific phases of reaching movements with stable patterns of task-related activity across time, establishing the relevance of the rodent for future studies designed to examine changes in task-related neural activity during recovery from focal cortical lesions.

Decoding of muscle activity from the sensorimotor cortex in freely behaving monkeys

Remarkable advances have recently been made in the development of Brain-Machine Interface (BMI) technologies for restoring or enhancing motor function. However, the application of these technologies may be limited to patients in static conditions, as these developments have been largely based on studies of animals (e.g., non-human primates) in constrained movement conditions. The ultimate goal of BMI technology is to enable individuals to move their bodies naturally or control external devices without physical constraints. Here, we demonstrate accurate decoding of muscle activity from electrocorticogram (ECoG) signals in unrestrained, freely behaving monkeys. We recorded ECoG signals from the sensorimotor cortex as well as electromyogram signals from multiple muscles in the upper arm while monkeys performed two types of movements with no physical restraints, as follows: forced forelimb movement (lever-pull task) and natural whole-body movement (free movement within the cage). As in previous reports using restrained monkeys, we confirmed that muscle activity during forced forelimb movement was accurately predicted from simultaneously recorded ECoG data. More importantly, we demonstrated that accurate prediction of muscle activity from ECoG data was possible in monkeys performing natural whole-body movement. We found that high-gamma activity in the primary motor cortex primarily contributed to the prediction of muscle activity during natural whole-body movement as well as forced forelimb movement. In contrast, the contribution of high-gamma activity in the premotor and primary somatosensory cortices was significantly larger during natural whole-body movement. Thus, activity in a larger area of the sensorimotor cortex was needed to predict muscle activity during natural whole-body movement. Furthermore, decoding models obtained from forced forelimb movement could not be generalized to natural whole-body movement, which suggests that decoders should be built individually and according to different behavior types. These results contribute to the future application of BMI systems in unrestrained individuals.

Adaptive Artifact Removal From Intracortical Channels for Accurate Decoding of a Force Signal in Freely Moving Rats

Intracortical data recorded with multi-electrode arrays provide rich information about kinematic and kinetic states of movement in the brain–machine interface (BMI) systems. Direct estimation of kinetic information such as the force from cortical data has the same importance as kinematic information to make a functional BMI system. Various types of the information including single unit activity (SUA), multiunit activity (MUA) and local field potential (LFP) can be used as an input information to extract motor commands for control of the external devices in BMI. Here we combine LFP and MUA information to improve decoding accuracy of the force signal from the multi-channel intracortical data of freely moving rats. We suggest a weighted common average referencing (CAR) algorithm in order to valid interpretation of the force decoding from different data types. The proposed spatial filter adaptively identifies contribution of the common noise on the channels employing Kalman filter method. We evaluated the efficacy of the proposed artifact algorithm on both simulation and real data. In the simulation study, the average R² between the original and reconstructed signal of all channels after applying the proposed artifact removal method was computed for input SNRs in the range of −45 to 0 dB. Weighted CAR method can effectively reconstruct the original signal with average R² higher than 0.5 for input SNRs higher than −s10 dB in case of adding simulated outlier and motion artifacts. We also show that the proposed artifact removal algorithm 33% improves the accuracy of force decoding in terms of R² value compared to standard CAR filters.

Implicit Grasp Force Representation in Human Motor Cortical Recordings

In order for brain-computer interface (BCI) systems to maximize functionality, users will need to be able to accurately modulate grasp force to avoid dropping heavy objects while also being able to handle fragile items. We present a case-study consisting of two experiments designed to identify whether intracortical recordings from the motor cortex of a person with tetraplegia could predict intended grasp force. In the first task, we were able classify neural responses to attempted grasps of four objects, each of which required similar grasp kinematics but different implicit grasp force targets, with 69% accuracy. In the second task, the subject attempted to move a virtual robotic arm in space to grasp a simple virtual object. For each trial, the subject was asked to grasp the virtual object with the force appropriate for one of the four objects from the first experiment, with the goal of measuring an implicit representation of grasp force. While the subject knew the grasp force during all phases of the trial, accurate classification was only achieved during active grasping, not while the hand moved to, transported, or released the object. In both tasks, misclassifications were most often to the object with an adjacent force requirement. In addition to the implications for understanding the representation of grasp force in motor cortex, these results are a first step toward creating intelligent algorithms to help BCI users grasp and manipulate a variety of objects that will be encountered in daily life.

Clinical Trial Identifier: NCT01894802 https://clinicaltrials.gov/ct2/show/NCT01894802.