David Guiraud

An implantable neuroprosthesis for standing and walking in paraplegia: 5-year patient follow-up

We present the results of a 5-year patient follow-up after implantation of an original neuroprosthesis. The system is able to stimulate both epimysial and neural electrodes in such a way that the complete flexor-extensor chain of the lower limb can be activated without using the withdrawal reflex. We demonstrate that standing and assisted walking are possible, and the results have remained stable for 5 years. Nevertheless, some problems were noted, particularly regarding the muscle response on the epimysial channels. Analysis of the electrical behaviour and thresholds indicated that the surgical phase is crucial because of the sensitivity of the functional responses to electrode placement. Neural stimulation proved to be more efficient and more stable over time. This mode requires less energy and provides more selective stimulation. This FES system can be improved to enable balanced standing and less fatiguing gait, but this will require feedback on event detection to trigger transitions between stimulation sequences, as well as feedback to the patient about the state of his lower limbs.

Mathematical muscle model for functional electrical stimulation control strategies

In paraplegic patients with upper motor neuron lesions the signal path from the central nervous system to muscles is interrupted. Functional Electrical Stimulation (FES) applied to the lower motor neurons can replace the lacking signals. A neuroprosthesis may be used to restore motor function in paraplegic patients on the basis of FES. The neuroprosthesic implant allows muscles to be controlled with high accuracy, high selectivity and the repeatability of the muscles response can be achieved. The SUAW project succeeded in the implantation of an advanced neuroprosthetic device on two patients, but the movement generation remains open loop and is tuned empirically. The system is thus insufficient to enhance significantly the daily-life of the patient, nevertheless, the good results obtained give us the opportunity to envisage the system evolves towards the automatic synthesis of the stimulation patterns generating the desired movement and closed loop control. To achieve this goal, some preliminary researches have to be carried out; starting with a specific modeling that can be used in the contest of FES. The main issues concern muscle modeling including FES parameters as inputs, fatigue, the interaction with the skeleton, and the identification of parameters. This paper describes the mathematical modeling of the skeletal muscle.

A distributed architecture for activating the peripheral nervous system

We present a new system for functional electrical stimulation (FES) applications based on networked stimulation units. They embed an advanced analog circuit, which provides multipolar and multiphasic stimulation profiles, and digital circuits, which ensure safety, locally executed programmed profiles, and communication with the master controller. This architecture is thus based on distributed stimulation units (DSU) that need only a two-wire bus to communicate, regardless of the number of poles of each DSU-driven electrode. This structure minimizes the required bandwidth between master and distributed units, increases the safety and stimulation features and decreases the complexity of the surgical approach. We have successfully tested this network-based stimulation architecture on benchtop stimulators. This original approach allows broad exploration of all possible methods to stimulate peripheral nerves, particularly in the goal of restoring the motor function. It provides a powerful research device to determine the optimal, least aggressive and the most efficient way to activate the peripheral nervous system using an implanted FES system that is less invasive than other existing devices.

FES-Induced Torque Prediction With Evoked EMG Sensing for Muscle Fatigue Tracking

This paper investigates a torque estimation method for muscle fatigue tracking, using stimulus evoked electromyography (eEMG) in the context of a functional electrical stimulation (FES) rehabilitation system. Although FES is able to effectively restore motor function in spinal cord injured (SCI) individuals, its application is inevitably restricted by muscle fatigue. In addition, the sensory feedback indicating fatigue is missing in such patients. Therefore, torque estimation is essential to provide feedback or feedforward signal for adaptive FES control. In this paper, a fatigue-inducing protocol is conducted on five SCI subjects via transcutaneous electrodes under isometric condition, and eEMG signals are collected by surface electrodes. A myoelectrical mechanical muscle model based on the Hammerstein structure with eEMG as model input is employed to capture muscle contraction dynamics. It is demonstrated that the correlation between eEMG and torque is time varying during muscle fatigue. Compared to conventional fixed-parameter models, the adapted-parameter model shows better torque prediction performance in fatiguing muscles. It motivates us to use a Kalman filter with forgetting factor for estimating the time-varying parameters and for tracking muscle fatigue. The assessment with experimental data reveals that the identified eEMG-to-torque model properly predicts fatiguing muscle behavior. Furthermore, the performance of the time-varying parameter estimation is efficient, suggesting that real-time tracking is feasible with a Kalman filter and driven by eEMG sensing in the application of FES.

Muscle Fatigue Tracking with Evoked EMG via Recurrent Neural Network: Toward Personalized Neuroprosthetics

One of the challenging issues in computational rehabilitation is that there is a large variety of patient situations depending on the type of neurological disorder. Human characteristics are basically subject specific and time variant; for instance, neuromuscular dynamics may vary due to muscle fatigue. To tackle such patient specificity and time-varying characteristics, a robust bio-signal processing and a precise model-based control which can manage the nonlinearity and time variance of the system, would bring break-through and new modality toward computational intelligence (CI) based rehabilitation technology and personalized neuroprosthetics. Functional electrical stimulation (FES) is a useful technique to assist restoring motor capability of spinal cord injured (SCI) patients by delivering electrical pulses to paralyzed muscles. However, muscle fatigue constraints the application of FES as it results in the time-variant muscle response. To perform adaptive closedloop FES control with actual muscle response feedback taken into account, muscular torque is essential to be estimated accurately. However, inadequacy of the implantable torque sensor limits the direct measurement of the time-variant torque at the joint. This motivates the development of methods to estimate muscle torque from bio-signals that can be measured. Evoked electromyogram (eEMG) has been found to be highly correlated with FES-induced torque under various muscle conditions, indicating that it can be used for torque/force prediction. A nonlinear ARX (NARX) type model is preferred to track the relationship between eEMG and stimulated muscular torque. This paper presents a NARX recurrent neural network (NARX-RNN) model for identification/prediction of FES-induced muscular dynamics with eEMG. The NARX-RNN model may possess novelty of robust prediction performance. Due to the difficulty of choosing a proper forgetting factor of Kalman filter for predicting time-variant torque with eEMG, the presented NARX-RNN could be considered as an alternative muscular torque predictor. Data collected from five SCI patients is used to evaluate the proposed NARX-RNN model, and the results show promising estimation performances. In addition, the general importance regarding CI-based motor function modeling is introduced along with its potential impact in the rehabilitation domain. The issue toward personalized neuroprosthetics is discussed in detail with the potential role of CI-based identification and the benefit for motor-impaired patient community.

Kinetics of neuromuscular changes during low-frequency electrical stimulation.

The purpose of the study was to examine the time course of neuromuscular fatigue components during a low‐frequency electrostimulation (ES) session. Three bouts of 17 trains of stimulation at 30 HZ (4 s on, 6 s off) were used to electrically induce fatigue in the plantar flexor muscles. Before and after every 17‐train bout, torque, electromyographic activity [expressed as root mean square (RMS) and median frequency (MF) values], evoked potentials (M‐wave and H‐reflex), and the level of voluntary activation (LOA, using twitch interpolation technique) were assessed. Torque during maximal voluntary contraction decreased significantly from the very first stimulation bout (−6.6 ± 1.11%, P < 0.001) and throughout the session (−10.32 ± 1.68% and −11.53 ± 1.27%, for the second and third bouts, respectively). The LOA and RMS/Mmax values were significantly decreased during the ES session (−2.9 ± 1.07% and −17.5 ± 6.14%, P < 0.01 and P< 0.001, respectively, at the end of the protocol), while MF showed no changes. The Hmax/Mmax ratio and Mmax were not significantly modified during the session. All twitch parameters were significantly potentiated after the first bout and throughout the session (P < 0.001). The maximal torque decrease was evident from the early phase of a low‐frequency ES protocol, with no concomitant inhibition of motoneuron excitability or depression of muscle contractile properties. These results are consistent with an early failure of the central drive to the muscle. Muscle Nerve, 2010

Original electronic design to perform epimysial and neural stimulation in paraplegia

This paper presents an original electronic architecture to manage epimysial and neural stimulation using the same implantable device. All the muscles needed to achieve lower limb movements such as standing and walking can thus be activated. Mainly for surgical reasons, some muscles need to be stimulated through different inputs: epimysium or motor nerve. We developed an electronic solution, including the design of an application-specific integrated circuit, to meet the requirements of both types of stimulation. Five years after the successful implantation of the system, we were able to evaluate the systems performance. The patient is still using the system at home and no failure occurred during this 5-year period. We conclude that the electronic design not only provides a unique investigative tool for research, but that it can also be used to restore the motor function of the lower limb. This technology has an advantage over external stimulation because the patient can safely use the system at home. However, improvements such as lower power consumption, and thus greater autonomy, are needed. We further conclude that the modelling of the electrical behaviour of the electrodes is reliable and the estimated parameter values are homogeneous and consistent for the same type of electrode. Thus, the three parameters of the first-order model can be identified from an acute animal experiment and provide a means to optimize the design of the output stage of implanted stimulators.

Evoked EMG-based torque prediction under muscle fatigue in implanted neural stimulation

In patients with complete spinal cord injury, fatigue occurs rapidly and there is no proprioceptive feedback regarding the current muscle condition. Therefore, it is essential to monitor the muscle state and assess the expected muscle response to improve the current FES system toward adaptive force/torque control in the presence of muscle fatigue. Our team implanted neural and epimysial electrodes in a complete paraplegic patient in 1999. We carried out a case study, in the specific case of implanted stimulation, in order to verify the corresponding torque prediction based on stimulus evoked EMG (eEMG) when muscle fatigue is occurring during electrical stimulation. Indeed, in implanted stimulation, the relationship between stimulation parameters and output torques is more stable than external stimulation in which the electrode location strongly affects the quality of the recruitment. Thus, the assumption that changes in the stimulation-torque relationship would be mainly due to muscle fatigue can be made reasonably. The eEMG was proved to be correlated to the generated torque during the continuous stimulation while the frequency of eEMG also decreased during fatigue. The median frequency showed a similar variation trend to the mean absolute value of eEMG. Torque prediction during fatigue-inducing tests was performed based on eEMG in model cross-validation where the model was identified using recruitment test data. The torque prediction, apart from the potentiation period, showed acceptable tracking performances that would enable us to perform adaptive closed-loop control through implanted neural stimulation in the future.

New implantable stimulator for the FES of paralyzed muscles

We propose a new implantable circuit for the internal functional electrical stimulation (FES) of motor nerves for paraplegic people. The circuit is designed to deliver precise calibrated stimulation pulses to specific multipolar electrodes. Several original design features have been developed to respond to the particular specifications imposed by safety constraints. In particular, the DAC has been thought to be fully monotonic and the output stage to ensure a passive and secure discharge of the safety capacitor. Also some features have been added in order to improve the classical charge pump that generates on-chip high voltage.

Multiscale modeling of skeletal muscle properties and experimental validations in isometric conditions

In this article, we describe an approach to model the electromechanical behavior of the skeletal muscle based on the Huxley formulation. We propose a model that complies with a well established macroscopic behavior of striated muscles where force-length, force–velocity, and Mirsky–Parmley properties are taken into account. These properties are introduced at the microscopic scale and related to a tentative explanation of the phenomena. The method used integrates behavior ranging from the microscopic to the macroscopic scale, and allows the computation of the dynamics of the output force and stiffness controlled by EMG or stimulation parameters. The model can thus be used to simulate and carry out research to develop control strategies using electrical stimulation in the context of rehabilitation. Finally, through animal experiments, we estimated model parameters using a Sigma Point Kalman Filtering technique and dedicated experimental protocols in isometric conditions and demonstrated that the model can accurately simulate individual variations and thus take into account subject dependent behavior.