Engineers working on neural prosthetics say a solution will be available within four years.
Although medical research has snatched patients from the brink of death, and reinstated abilities to those who had lost them, there are many areas over which doctors are only able to have the merest influence. Over time however, more and more concepts are becoming science fact over science fiction, thanks to technology-based research offering brand new insights and solutions.
One area that scientists and engineers from the DEMAR team - a joint project between INRIA, CNRS, and the universities of Montpellier 1 & 2 - are researching is that of modelling and controlling the human sensory-motor system. This will in time lead to a better understanding of movement disorders and the ability to offer solutions to movement deficiencies.
Our research team is currently working on another project in conjunction with this, to create a working neuroprosthetic solution. This research could lead to positive life-changing technology for many disabled or limbless people.
Functional Electrical Stimulation (FES) technology for movement restoration is currently under research and continues to evolve. At the beginning, only surface stimulation was possible and thus only used in a clinical context due to the low reliability of electrode placements. In the early 1980s, implanted FES appeared through well-known applications such as the cochlear implant, and, more recently Deep Brain Stimulation, but the complexity of the system for movement restoration is such that few commercial applications readily arose.
Even though the original idea of FES is still the same: activating the motor-neurone axons with impulse current generators, the stimulus waveform and its parameters have drastically evolved and the electrode placements have become various, such as transcutaneous, percutaneous, implanted epimysial stimulation at the muscle's motor point, neural stimulation on the nerve and Sacral roots' stimulation near the spinal cord etc. This knowledge can now be included in the next generation of implanted neuroprosthetic devices allowing a wide variety of features.
Moreover, FES is currently the only way to restore motor function for some diseases, even though biological solutions are being studied, because the research has not yet successfully been tested on humans. FES should not be considered as a competitor of spinal cord regeneration as it might be that both approaches could be complementary in the end.
Few teams carry out research on implanted FES dedicated to complex movement restoration, and the functional results remain poor with regard to the complex and heavy surgery required. For example, fatigue, muscle's selectivity are not yet properly managed.
Nevertheless, the technique has proved to be useable and needs enhancing, which our research team; DEMAR (Deambulation Et Mouvement ARtiﬁciel) is addressing. In particular, complex electrode geometries associated with complex stimulus waveforms provide a way to perform fibre type selectivity and spatial localisation of the stimuli in the nerves. These features are not yet being implemented and demand new hardware and software architectures.
Advanced control theory
Such a complex system needs advanced control theory tools coupled with a deep under-standing of the underlying neurophysiological processes. This major area of research is also an important part of the DEMAR objectives. Few teams work on this topic because it needs considerable interaction between completely different disciplines such as neuro-physiology, biomechanics, microelectronics, industrial informatics, automatic control theory, and advanced signal processing, but we are rising to the challenge.
Still in its youth, the DEMAR project is just four years old, and one of our biggest focuses is to truly understand in detail the body's sensory motor network. This requires accurately modelling all the network's parts, aside from the brain, which is what we are working towards.
We have shied away from the brain at this stage because our knowledge and expertise lie in other areas, so we have chosen to stop at the spinal cord, focusing mainly on patients with spinal cord injuries or pathologies where peripheral stimulation may be applied, such as stroke patients.
The first part of the work is to model and identify the muscles and spinal cord loops such as reflexes. Our next big goal is to identify the model parameters, which we have just begun to work on. This work will in time help diagnosis of motor disorders, but will also help our secondary research work, which is to design neuroprosthetic devices, again leading to restoring motor and sensitivity functions though implanted FES and neural signal sensing.
We now have a mathematical model of the muscles in place and can simulate many things, but we must now consider how we can accurately estimate the parameters of these models on a given patient. There is the question of experimental protocols of course; not only what kind of measurements we have to do, but also how to optimise the mathematical equations to get the right estimation of the parameters.
We've begun work on what is coded by the system as regards natural sensors and are trying to understand how the information is coded, what kind of information is coded and understand and determine how to interpret neural sensor activity to estimate muscle state that can be used to feedback closed loop controllers.
Furthermore, sensing the different pathways such as those originating from muscles' spindles, will be used to provide a closed loop control of FES through natural sensing and then a complete implanted solution. Sensing the neural system may provide great enhancements in some other complex motor controls such as bladder control. Antagonist muscles' contractions, and sensory feedbacks interfere with FES when applied directly on the sacral root nerve concerned. Thus, enhanced activation waveforms and sensing feedback or feedforward signals are needed to perform a highly selective stimulation.
Wide application area
The possibility to interface the sensory motor system, both activating neural structure with implanted FES, and sensing through implanted neural signal recordings opens a wide application area for us. This includes restoring motor function, such as grasping for quadriplegic patients and standing and walking for paraplegic and hemiplegic patients. These applications can be used in a clinical environment to provide physiotherapists with a new efficient FES-based therapy (using surface electrodes) in the rehabilitation process. With a more sophisticated technology such as implanted neuroprostheses, systems can be used at home by the patients themselves, without the need for clinical staff.
While many different studies into the modelling and control areas are currently underway within our research group, we're also aiming high, hoping eventually to create an implemented system to monitor and stimulate neural structures, giving back mobility to patients. There have been some really great advances in this field. In the past all the neural prosthetics used for movement restoration and rehabilitation were based on centralised electronics, meaning you have a centralised implant system with all the electronics generating all the stimuli, which requires cables to go through the body to the nerves and the electrodes. There could be something like eight metres of wire weaving it way through the implanted patient's body.
This makes surgery very tricky and increases the risk of a global infection, so our aim is to provide a completely wireless stimulation unit. This requires electronics that permit the stimulation generator to be placed close to the electrodes; the electronics also need to communicate with all the units and with the global movement controller.
There is only one other team in the world to have tried this; The Alfred Mann Foundation in the US with the Bion product. Teams there developed an independent stimulation unit, injected using a needle in the muscle itself or close to the motor nerve, stimulating the muscle through an RF link.
One of the early outcomes of our research has been the development of a dedicated protocol to communicate through the network, a protocol which is now patented and used by two industrial partners in the medical field - we hope two more will be using it soon.
Another outcome is the development of a prototype stimulation unit, which should become implantable later this year. The prototype unit is able to stimulate a nerve through multipolar stimulation, meaning that, not only can you generate stimulus, but you can generate a complex stimulus on the multipolar electrodes. The idea is that you can focus the current line in the nerve, so that you can selectively stimulate part of the nerve, and not the whole nerve. To explain this in greater detail; to overcome the limitations of the present FES centralised architecture, a new FES architecture was proposed according to the SENIS (Stimulation Electrique Neurale dIStribuée) concept: the distribution of, firstly, the stimulation unit with its control near its activator, i.e. its associated multipolar neural electrode, and, secondly, the implanted neural sensor coupled to a multipolar recording electrode with its embedded signal processing.
FES was thus performed by means of distributed small stimulation units which are driven by an external controller in charge of the coordination of stimulation sequences. Each stimulation unit (called DSU, Distributed Stimulation Unit) is in charge of the execution of the stimulation pattern, applied to the muscle by means of a neural multipolar electrode, with each DSU being composed of analogue and digital parts.
The SENIS architecture therefore relies on a set of DSUs that communicates with an external controller. We studied the communication architecture and defined an adequate protocol, assuming firstly that the communication should be performed on a wireless medium and secondly, that this architecture can also contain distributed measurement units (DMU for sensors).
We have also developed a dedicated processor for stimulation - a 'heart' with its own compact, specific 'language' - to tell the nerves what to do. It is fully developed in VHDL (VHSIC Hardware Description Language where VHSIC stands for Very High Speed Integrated Circuit) so that it is embedded on the DSU ASIC (Application Specific Integrated Circuit). This is the real advance for us, with the next step being able to provide a final, complete, implant for humans.
The active part of the stimulation device is made up of a control bus that acts as the interface with the digital (control) block, a digital to analogue converter (DAC) generating the required current for the stimulation, and a reversible current divider allowing the duplication and distribution of the current between the 12 poles of a multipolar cuff electrode.
The overall shape of the current wave form is defined by the DAC output. The current divider block is in charge of the replication and distribution of this current within the 12 poles of the electrode and guarantees the current ratios accurately between the poles. Security aspects will soon be integrated within the DSU, for example, by means of embedded reference models ensuring the respect of physiological constraints. To allow for remote monitoring of the DSU, we have added the ability to remotely access the (internal) DSU state, in addition to already having the possibility to access downloaded micro-programs, network configuration, etc.
The biggest aim of the project is to provide neuroprosthetics for humans, which I hope we will achieve within the next four years. The length of time is due to the work that will need to take place post-creation, which includes usability and clinical trials.
Once these are complete, we can then say we have helped people to rehabilitate and regain some basic movement functionality. The work is still in its early stages but with every passing year we are moving the science and technology on in leaps and bounds. Already we've discovered much and learnt ways to make technology work for us as humans, but there's still considerably more to uncover. We may not be there yet, but if we can continue to progress as we have so far, I believe there will be a viable neural prosthetic solution for the thousands of patients who can - and will benefit from such a breakthrough.