Real-time prediction of bladder urine leakage using fuzzy inference system and dual Kalman filtering in cats.

The use of electrical stimulation devices to manage bladder incontinence relies on the application of continuous inhibitory stimulation. However, continuous stimulation can result in tissue fatigue and increased delivered charge. Here, we employ a real-time algorithm to provide a short-time prediction of urine leakage using the high-resolution power spectrum of the bladder pressure during the presence of non-voiding contractions (NVC) in normal and overactive bladder (OAB) cats. The proposed method is threshold-free and does not require pre-training. The analysis revealed that there is a significant difference between voiding contraction (VC) and NVC pressures as well as band powers (0.5-5 Hz) during both normal and OAB conditions. Also, most of the first leakage points occurred after the maximum VC pressure, while all of them were observed subsequent to the maximum VC spectral power. Kalman-Fuzzy method predicted urine leakage on average 2.2 s and 1.6 s before its occurrence and an average of 2.0 s and 1.1 s after the contraction started with success rates of 94.2% and 100% in normal and OAB cats, respectively. This work presents a promising approach for developing a neuroprosthesis device, with on-demand stimulation to control bladder incontinence.

A Distributed Automatic Control Framework for Simultaneous Control of Torque and Cadence in Functional Electrical Stimulation Cycling.

One of the major challenges facing functional electrical stimulation (FES) cycling is the design of an automatic control system that addresses the problem of disturbance with unknown bound and time-varying behavior of the muscular system. The previous methods for FES-cycling are based on the system modeling and require pre-adjustment of the control parameters which are based on the model parameters. These will degrade the FES-cycling performance and limit the clinical application of the methods. In this paper, a distributed cooperative control framework, which is based on an adaptive higher-order sliding mode (AHOSM) controller, is proposed for simultaneous control of torque and cadence in FES-cycling. The proposed control system is free-model which does not require any pre-adjustment of the control parameters and does not need the boundary of the disturbance to be known. Another major issue in FES-cycling is the stimulation pattern. In the paper, an automatic pattern generator is proposed which is capable of providing not only the regions of the crank angle in which each muscle group should be stimulated but also a specific gain for each muscle group. The results of the simulation studies and experiments on three spinal cord injuries showed that the proposed control strategy significantly increases the efficiency and tracking accuracy of motor-assisted FES-cycling in paraplegic patients and decreases the power consumption compared to HOSM controller with the fixed stimulation pattern. Reducing power consumption can slow down muscle fatigue and consequently increase cycling endurance. The average of cadence and torque tracking errors over three subjects using the proposed method are 5.77± 0.5% and 5.23± 0.8%, respectively.

Decoding Bilateral Hindlimb Kinematics From Cat Spinal Signals Using Three-Dimensional Convolutional Neural Network.

To date, decoding limb kinematic information mostly relies on neural signals recorded from the peripheral nerve, dorsal root ganglia (DRG), ventral roots, spinal cord gray matter, and the sensorimotor cortex. In the current study, we demonstrated that the neural signals recorded from the lateral and dorsal columns within the spinal cord have the potential to decode hindlimb kinematics during locomotion. Experiments were conducted using intact cats. The cats were trained to walk on a moving belt in a hindlimb-only condition, while their forelimbs were kept on the front body of the treadmill. The bilateral hindlimb joint angles were decoded using local field potential signals recorded using a microelectrode array implanted in the dorsal and lateral columns of both the left and right sides of the cat spinal cord. The results show that contralateral hindlimb kinematics can be decoded as accurately as ipsilateral kinematics. Interestingly, hindlimb kinematics of both legs can be accurately decoded from the lateral columns within one side of the spinal cord during hindlimb-only locomotion. The results indicated that there was no significant difference between the decoding performances obtained using neural signals recorded from the dorsal and lateral columns. The results of the time-frequency analysis show that event-related synchronization (ERS) and event-related desynchronization (ERD) patterns in all frequency bands could reveal the dynamics of the neural signals during movement. The onset and offset of the movement can be clearly identified by the ERD/ERS patterns. The results of the mutual information (MI) analysis showed that the theta frequency band contained significantly more limb kinematics information than the other frequency bands. Moreover, the theta power increased with a higher locomotion speed.

A general framework for automatic closed-loop control of bladder voiding induced by intraspinal microstimulation in rats.

Individuals with spinal cord injury or neurological disorders have problems in voiding function due to the dyssynergic contraction of the urethral sphincter. Here, we introduce a closed-loop control of intraspinal microstimulation (ISMS) for efficient bladder voiding. The strategy is based on asynchronous two-electrode ISMS with combined pulse-amplitude and pulse-frequency modulation without requiring rhizotomy, neurotomy, or high-frequency blocking. Intermittent stimulation is alternately applied to the two electrodes that are implanted in the S2 lateral ventral horn and S1 dorsal gray commissure, to excite the bladder motoneurons and to inhibit the urethral sphincter motoneurons. Asynchronous stimulation would lead to reduce the net electric field and to maximize the selective stimulation. The proposed closed-loop system attains a highly voiding efficiency of 77.2-100%, with an average of 91.28 ± 8.4%. This work represents a promising approach to the development of a natural and robust motor neuroprosthesis device for restoring bladder functions.

Decoding hindlimb kinematics from descending and ascending neural signals during cat locomotion.

Objective. The main objective of this research is to record both sensory and motor information from the ascending and descending tracts within the spinal cord to decode the hindlimb kinematics during walking on a treadmill.Approach. Two different experimental paradigms (i.e. active and passive) were used in the current study. During active experiments, five cats were trained to walk bipedally while their hands were kept on the front frame of the treadmill for balance or to walk quadrupedally. During passive experiments, the limb was passively moved by the experimenter. Local field potential (LFP) activity was recorded using a microwire array implanted in the dorsal column (DC) and lateral column (LC) of the L3-L4 spinal segments. The amplitude and frequency components of the LFP formed the feature set, and the elastic net regularization was used to decode the hindlimb joint angles.Main results.The results show that there is no significant difference between the information content of the signals recorded from the DC and LC regions during walking on the treadmill, but the information content of the DC is significantly higher than that of the LC during passively applied movement of the hindlimb in the anesthetized cats. Moreover, the decoding performance obtained using the recorded signals from the DC is comparable with that from the LC during locomotion. However, the decoding performance obtained using the recording channels in the DC is significantly better than that obtained using the signals recorded from the LC. The long-term analysis shows that robust decoding performance can be achieved over 2-3 months without a significant decrease in performance.Significance. This work presents a promising approach to developing a natural and robust motor neuroprosthesis device using descending neural signals to execute the movement and ascending neural signals as the feedback information to control the movement.

A state-based probabilistic method for decoding hand position during movement from ECoG signals in non-human primate.

OBJECTIVE: In this study, we proposed a state-based probabilistic method for decoding hand positions during unilateral and bilateral movements using the ECoG signals recorded from the brain of Rhesus monkey. APPROACH: A customized electrode array was implanted subdurally in the right hemisphere of the brain covering from the primary motor cortex to the frontal cortex. Three different experimental paradigms were considered: ipsilateral, contralateral, and bilateral movements. During unilateral movement, the monkey was trained to get food with one hand, while during bilateral movement, the monkey used its left and right hands alternately to get food. To estimate the hand positions, a state-based probabilistic method was introduced which was based on the conditional probability of the hand movement state (i.e. idle, right hand movement, and left hand movement) and the conditional expectation of the hand position for each state. Moreover, a hybrid feature extraction method based on linear discriminant analysis and partial least squares (PLS) was introduced. MAIN RESULTS: The proposed method could successfully decode the hand positions during ipsilateral, contralateral, and bilateral movements and significantly improved the decoding performance compared to the conventional Kalman and PLS regression methods [Formula: see text]. The proposed hybrid feature extraction method was found to outperform both the PLS and PCA methods [Formula: see text]. Investigating the kinematic information of each frequency band shows that more informative frequency bands were [Formula: see text] (15-30 Hz) and [Formula: see text](50-100 Hz) for ipsilateral and [Formula: see text] and [Formula: see text] (100-200 Hz) for contralateral movements. It is observed that ipsilateral movement was decoded better than contralateral movement for [Formula: see text] (5-15 Hz) and [Formula: see text] bands, while contralateral movements was decoded better for [Formula: see text] (30-200 Hz) and hfECoG (200-400 Hz) bands. SIGNIFICANCE: Accurate decoding the bilateral movement using the ECoG recorded from one brain hemisphere is an important issue toward real-life applications of the brain-machine interface technologies.

Estimation of Bladder Pressure and Volume from the Neural Activity of Lumbosacral Dorsal Horn Using a Long-Short-Term-Memory-based Deep Neural Network.

In this paper, we propose a deep recurrent neural network (DRNN) for the estimation of bladder pressure and volume from neural activity recorded directly from spinal cord gray matter neurons. The model was based on the Long Short-Term Memory (LSTM) architecture, which has emerged as a general and effective model for capturing long-term temporal dependencies with good generalization performance. In this way, training the network with the data recorded from one rat could lead to estimating the bladder status of different rats. We combined modeling of spiking and local field potential (LFP) activity into a unified framework to estimate the pressure and volume of the bladder. Moreover, we investigated the effect of two-electrode recording on decoding performance. The results show that the two-electrode recordings significantly improve the decoding performance compared to single-electrode recordings. The proposed framework could estimate bladder pressure and volume with an average normalized root-mean-squared (NRMS) error of 14.9 ± 4.8% and 19.7 ± 4.7% and a correlation coefficient (CC) of 83.2 ± 3.2% and 74.2 ± 6.2%, respectively. This work represents a promising approach to the real-time estimation of bladder pressure/volume in the closed-loop control of bladder function using functional electrical stimulation.

A probabilistic recurrent neural network for decoding hind limb kinematics from multi-segment recordings of the dorsal horn neurons.

OBJECTIVE: Providing accurate and robust estimates of limb kinematics from recorded neural activities is prominent in closed-loop control of functional electrical stimulation (FES). A major issue in providing accurate decoding the limb kinematics is the decoding model. The primary goal of this study is to develop a decoding approach to model the dynamic interactions of neural systems for accurate decoding. Another critical issue is to find reliable recording sites. Up to now, neural recordings from spinal neural activities were investigated. In this paper, the neural recordings from different vertebrae in decoding limb kinematics are investigated. APPROACH: In the current study, a new generative probabilistic model with explicit considering the joint density is developed. Then, an adaptive discriminative learning algorithm is proposed for learning the model. It will be shown that the proposed generative process can be implemented by a recurrent neural network (RNN) with specific structure. We record the neural activities from dorsal horn neurons by using three electrodes placed in the L4, L5, and L6 vertebrae in anesthetized cats. MAIN RESULTS: Information theoretic analysis on single-joint movement and multi-segment recordings implies the rostrocaudal distribution of kinematic information. It is demonstrated that during hip movement, best decoding performance is achieved by L4 recordings. For knee and ankle movements, best decodings are achieved by L5, and L6 recordings respectively. It is also shown that the decoding accuracy using multi-segment recordings outperforms decoding accuracy obtained by single-segment recording in multi-joint movement. The results also confirm the superiority of the proposed probabilistic recurrent neural network (PRNN) over the conventional recurrent neural network and Kalman filter ([Formula: see text]). SIGNIFICANCE: Multi-segment recordings from dorsal horn neurons as well as the proposed probabilistic recurrent network model provide a promising approach for robust and accurate decoding limb kinematics.

Automatic Seizure Detection Based on Nonlinear Dynamical Analysis of EEG Signals and Mutual Information.

INTRODUCTION: In this paper, nonlinear dynamical analysis based on Recurrence Quantification Analysis (RQA) is employed to characterize the nonlinear EEG dynamics. RQA can provide useful quantitative information on the regular, chaotic, or stochastic property of the underlying dynamics. METHODS: We use the RQA-based measures as the quantitative features of the nonlinear EEG dynamics. Mutual Information (MI) was used to find the most relevant feature subset out of RQA-based features. The selected features were fed into an artificial neural network for grouping of EEG recordings to detect ictal, interictal, and healthy states. The performance of the proposed procedure was evaluated using a database for different classification cases. RESULTS: The combination of five selected features based on MI achieved 100% accuracy, which demonstrates the superiority of the proposed method. CONCLUSION: The results showed that the nonlinear dynamical analysis based on Rcurrence Quantification Analysis (RQA) can be employed as a suitable approach for characterizing the nonlinear EEG dynamics and detecting the seizure.

Block-based robust control of stepping using intraspinal microstimulation.

OBJECTIVE: The problem of motor control using intraspinal microstimulation (ISMS) can be approached at two levels of the motor system: individual muscles (motor pools) and motor primitives. The major challenges of direct ISMS at the level of individual muscle are the number of electrodes that are required to be implanted in order to recruit all muscles involving the motion and muscle selectivity. One solution to cope with these problems is the control of movement generated by appropriate combination of the movement primitives. In this paper, we proposed a robust control framework using primitives for fully automatic block-based control of the motion through ISMS. APPROACH: The control framework is based on an adaptive fuzzy terminal sliding mode control. The biggest advantage of the controller is the fast convergence compared to the conventional sliding mode control. MAIN RESULTS: The experiments were conducted on spinally-intact anesthetized cats. Based on electromyography activity of the hindlimbs muscles, different movement blocks were defined. The results of block-based air-stepping control show that the proposed control framework could generate the gait cycle with good tracking performance. The averages of tracking error, over five cats, were 9.3%, 11.2%, and 16.1%, for the ankle, knee, and hip joints, respectively. The results of walking control on the moving treadmill demonstrated that the gait cycle can be generated only with two movement blocks for each leg. SIGNIFICANCE: The results of the current study demonstrated that the normal gait pattern can be achieved by tracking control of the movement blocks using ISMS, while the controller requires no offline learning phase and no pre-adjustment of the stimulation level. The controller is able to automatically regulate the interactions between movement blocks without any preprogrammed block activities.

A piecewise probabilistic regression model to decode hand movement trajectories from epidural and subdural ECoG signals.

OBJECTIVE: The primary concern of this study is to develop a probabilistic regression method that would improve the decoding of the hand movement trajectories from epidural ECoG as well as from subdural ECoG signals. APPROACH: The model is characterized by the conditional expectation of the hand position given the ECoG signals. The conditional expectation of the hand position is then modeled by a linear combination of the conditional probability density functions defined for each segment of the movement. Moreover, a spatial linear filter is proposed for reducing the dimension of the feature space. The spatial linear filter is applied to each frequency band of the ECoG signals and extract the features with highest decoding performance. MAIN RESULTS: For evaluating the proposed method, a dataset including 28 ECoG recordings from four adult Japanese macaques is used. The results show that the proposed decoding method outperforms the results with respect to the state of the art methods using this dataset. The relative kinematic information of each frequency band is also investigated using mutual information and decoding performance. The decoding performance shows that the best performance was obtained for high gamma bands from 50 to 200 Hz as well as high frequency ECoG band from 200 to 400 Hz for subdural recordings. However, the decoding performance was decreased for these frequency bands using epidural recordings. The mutual information shows that, on average, the high gamma band from 50 to 200 Hz and high frequency ECoG band from 200 to 400 Hz contain significantly more information than the average of the rest of the frequency bands [Formula: see text] for both subdural and epidural recordings. The results of high resolution time-frequency analysis show that ERD/ERS patterns in all frequency bands could reveal the dynamics of the ECoG responses during the movement. The onset and offset of the movement can be clearly identified by the ERD/ERS patterns. SIGNIFICANCE: Reliable decoding the kinematic information from the brain signals paves the way for robust control of external devices.

Decoding hind limb kinematics from neuronal activity of the dorsal horn neurons using multiple level learning algorithm.

Decoding continuous hind limb joint angles from sensory recordings of neural system provides a feedback for closed-loop control of hind limb movement using functional electrical stimulation. So far, many attempts have been done to extract sensory information from dorsal root ganglia and sensory nerves. In this work, we examine decoding joint angles trajectories from the single-electrode extracellular recording of dorsal horn gray matter of the spinal cord during passive limb movement in anesthetized cats. In this study, a processing framework based on ensemble learning approach is propose to combine firing rate (FR) and interspike interval (ISI) information of the neuronal activity. For this purpose, a stacked generalization approach based on recurrent neural network is proposed to enhance decoding accuracy of the movement kinematics. The results show that the high precision neural decoding of limb movement can be achieved even with a single electrode implanted in the spinal cord gray matter.

A modular robust control framework for control of movement elicited by multi-electrode intraspinal microstimulation.

OBJECTIVE: An important issue in restoring motor function through intraspinal microstimulation (ISMS) is the motor control. To provide a physiologically plausible motor control using ISMS, it should be able to control the individual motor unit which is the lowest functional unit of motor control. By focal stimulation only a small group of motor neurons (MNs) within a motor pool can be activated. Different groups of MNs within a motor pool can potentially be activated without involving adjacent motor pools by local stimulation of different parts of a motor pool via microelectrode array implanted into a motor pool. However, since the system has multiple inputs with single output during multi-electrode ISMS, it poses a challenge to movement control. In this paper, we proposed a modular robust control strategy for movement control, whereas multi-electrode array is implanted into each motor activation pool of a muscle. APPROACH: The controller was based on the combination of proportional-integral-derivative and adaptive fuzzy sliding mode control. The global stability of the controller was guaranteed. MAIN RESULTS: The results of the experiments on rat models showed that the multi-electrode control can provide a more robust control and accurate tracking performance than a single-electrode control. The control output can be pulse amplitude (pulse amplitude modulation, PAM) or pulse width (pulse width modulation, PWM) of the stimulation signal. The results demonstrated that the controller with PAM provided faster convergence rate and better tracking performance than the controller with PWM. SIGNIFICANCE: This work represents a promising control approach to the restoring motor functions using ISMS. The proposed controller requires no prior knowledge about the dynamics of the system to be controlled and no offline learning phase. The proposed control design is modular in the sense that each motor pool has an independent controller and each controller is able to control ISMS through an array of microelectrodes.

The Effects of Stimulation Strategy on Joint Movement Elicited by Intraspinal Microstimulation.

The goal of this study was to characterize the effects of stimulation parameters and multielectrode stimulation on selectivity, range of motion, recruitment characteristics, and fatigue during intraspinal microstimulation (ISMS). A custom-made multielectrode array was implanted into the activation pool of the rat dorsiflexor muscle where the stimulation produced the highest movement range on the ankle joint and the least effect on the other joints. The results show that the selectivity could be significantly enhanced using multielectrode stimulation strategy. Moreover, the fatigue was significantly reduced using multielectrode synchronous stimulation with respect to single-electrode stimulation. For a given charge, stimulation with higher current amplitude and shorter pulse duration produced greater range of motion than that with lower amplitude and longer pulse duration. However, the stimulation with shorter duration caused greater fatigue than that with longer. In addition, there was a significant difference in time constant of spinal response obtained with different pulse amplitudes during pulse width (PW) modulation. The time constant decreased with increasing pulse amplitude. However, there was no significant effect of pulse duration on time constant during pulse amplitude (PA) modulation. The results suggest that the motor neurons (MNs) within the spinal cord can be recruited according to size principle by appropriate selection of stimulation parameters. Based on these results an efficient stimulation strategy can be designed for control of movement performance (i.e., speed of movement, fatigue, range of motion, and selectivity) during ISMS.

Fully automatic control of paraplegic FES pedaling using higher-order sliding mode and fuzzy logic control.

In this paper, a fully automatic robust control strategy is proposed for control of paraplegic pedaling using functional electrical stimulation (FES). The method is based on higher-order sliding mode (HOSM) control and fuzzy logic control. In FES, the strength of muscle contraction can be altered either by varying the pulse width (PW) or by the pulse amplitude (PA) of the stimulation signal. The proposed control strategy regulates simultaneously both PA and PW (i.e., PA/PW modulation). A HOSM controller is designed for regulating the PW and a fuzzy logic controller for the PA. The proposed control scheme is free-model and does not require any offline training phase and subject-specific information. Simulation studies on a virtual patient and experiments on three paraplegic subjects demonstrate good tracking performance and robustness of the proposed control strategy against muscle fatigue and external disturbances during FES-induced pedaling. The results of simulation studies show that the power and cadence tracking errors are 5.4% and 4.8%, respectively. The experimental results indicate that the proposed controller can improve pedaling system efficacy and increase the endurance of FES pedaling. The average of power tracking error over three paraplegic subjects is 7.4±1.4% using PA/PW modulation, while the tracking error is 10.2±1.2% when PW modulation is used. The subjects could pedal for 15 min with about 4.1% power loss at the end of experiment using proposed control strategy, while the power loss is 14.3% using PW modulation. The controller could adjust the stimulation intensity to compensate the muscle fatigue during long period of FES pedaling.

Fuzzy logic control of ankle movement using multi-electrode intraspinal microstimulation.

In this paper, we propose a fuzzy logic control (FLC) for control of ankle movement using multi-electrode intraspinal microstimulation (ISMS). It has been demonstrated that ISMS via multi-electrode implanted into a given motor pool has several advantages over the single-electrode ISMS. In the current study, we investigate the closed-loop control of ankle movement using multi-electrode ISMS. For this purpose, a pair of electrodes was implanted into the each motor pool of dorsiflexor and plantar flexor muscles in the spinal cord. For each muscle, an independent FLC was designed. The response of neuromuscular system has a time-delay with respect to the input stimulation. To compensate the effect of time-delay, the future value of desired response was considered as the input of the FLC as well as the error signal. The results of experiments on animals show that the proposed control framework can provide a good tracking performance.

Classification of brain signals associated with imagination of hand grasping, opening and reaching by means of wavelet-based common spatial pattern and mutual information.

An important issue in designing a practical brain-computer interface (BCI) is the selection of mental tasks to be imagined. Different types of mental tasks have been used in BCI including left, right, foot, and tongue motor imageries. However, the mental tasks are different from the actions to be controlled by the BCI. It is desirable to select a mental task to be consistent with the desired action to be performed by BCI. In this paper, we investigated the detecting the imagination of the hand grasping, hand opening, and hand reaching in one hand using electroencephalographic (EEG) signals. The results show that the ERD/ERS patterns, associated with the imagination of hand grasping, opening, and reaching are different. For classification of brain signals associated with these mental tasks and feature extraction, a method based on wavelet packet, regularized common spatial pattern (CSP), and mutual information is proposed. The results of an offline analysis on five subjects show that the two-class mental tasks can be classified with an average accuracy of 77.6% using proposed method. In addition, we examine the proposed method on datasets IVa from BCI Competition III and IIa from BCI Competition IV.

Dynamic optimization of walker-assisted FES-activated paraplegic walking: simulation and experimental studies.

In this paper, we propose a musculoskeletal model of walker-assisted FES-activated paraplegic walking for the generation of muscle stimulation patterns and characterization of the causal relationships between muscle excitations, multi-joint movement, and handle reaction force (HRF). The model consists of the lower extremities, trunk, hands, and a walker. The simulation of walking is performed using particle swarm optimization to minimize the tracking errors from the desired trajectories for the lower extremity joints, to reduce the stimulations of the muscle groups acting around the hip, knee, and ankle joints, and to minimize the HRF. The results of the simulation studies using data recorded from healthy subjects performing walker-assisted walking indicate that the model-generated muscle stimulation patterns are in agreement with the EMG patterns that have been reported in the literature. The experimental results on two paraplegic subjects demonstrate that the proposed methodology can improve walking performance, reduce HRF, and increase walking speed when compared to the conventional FES-activated paraplegic walking.

Restoring Motor Functions in Paralyzed Limbs through Intraspinal Multielectrode Microstimulation Using Fuzzy Logic Control and Lag Compensator.

In this paper, a control strategy is proposed for control of ankle movement on animals using intraspinal microstimulation (ISMS). The proposed method is based on fuzzy logic control. Fuzzy logic control is a methodology of intelligent control that mimics human decision making process. This type of control method can be very useful for the complex uncertain systems that their mathematical model is unknown. To increase the stability and speed of the system's response and reduce the steady-state error, we combine the FLC with a lead (lag) compensator. The experiments are conducted on five rats. Microelectrodes are implanted into the spinal cord to provide selective stimulation of plantarflexor and dorsiflexor. The results show that motor functions can be restored using ISMS. Despite the complexity of the spinal neuronal networks and simplicity of the proposed control strategy, our results show that the proposed strategy can provide acceptable tracking control with fast convergence.

A decentralized modular control framework for robust control of FES-activated walker-assisted paraplegic walking using terminal sliding mode and fuzzy logic control.

A major challenge to developing functional electrical stimulation (FES) systems for paraplegic walking and widespread acceptance of these systems is the design of a robust control strategy that provides satisfactory tracking performance. The systems need to be robust against time-varying properties of neuromusculoskeletal dynamics, day-to-day variations, subject-to-subject variations, external disturbances, and must be easily applied without requiring offline identification during different experimental sessions. Another major problem related to walker-assisted FES-activated walking concerns the high metabolic rate and upper body effort that limit the clinical applications of FES systems. In this paper, we present a novel decentralized modular control framework for robust control of walker-assisted FES-activated walking. For each muscle-joint dynamics, an independent module control is designed, and the dynamics of the plant are identified online. This process requires no prior knowledge about the dynamics of the plant to be controlled and no offline learning phase. The module is based on adaptive fuzzy terminal sliding mode control and fuzzy logic control. The module control adjusts both pulse-amplitude and pulsewidth of the stimulation signal in such a way that upper body effort is minimized and the lower extremity walking pattern lies within a defined boundary of the reference trajectory. The proposed control strategy has been evaluated on three paraplegic subjects. The results showed that accurate tracking performance and smooth walking pattern were achieved. This favorable performance was obtained without requiring offline identification, manual adjustments, and predefined ON/OFF timing of the muscles.