WO2008142027A1

WO2008142027A1 - Method and apparatus for stimulating activity in the peripheral nervous system

Info

Publication number: WO2008142027A1
Application number: PCT/EP2008/056061
Authority: WO
Inventors: Iasonas Triantis; Christopher Toumazou
Original assignee: Imperial Innovations Limited
Priority date: 2007-05-17
Filing date: 2008-05-16
Publication date: 2008-11-27
Also published as: GB0709542D0

Abstract

A method and apparatus for stimulating activity in the peripheral nervous system comprises a nerve cuff (20) having a pair of segmented stimulation electrodes (36, 38). In order to stimulate a fascicle (22) within the cuff, a voltage is applied to opposing electrode segments, one segment being on the first electrode and the other on the second. This is then repeated for the next pair in sequence, and so on, thereby generating a rotating dipole which causes a relatively high stimulation current at an interior focal point (50) while at the same time keeping the current at the periphery low.

Description

METHOD AND APPARATUS FOR STIMULATING ACTIVITY IN THE PERIPHERAL NERVOUS SYSEM

The present invention relates to a method and apparatus for stimulating activity in the peripheral nervous system.

The preferred interface is potentially fully-implantable, non-invasive to the nerve and could be eventually developed for chronic neural stimulation and monitoring. One specific application relates to the use of a "cuff electrode" interface, placed around the Vagus nerve to detect and control activity in that nerve prior to and during a seizure. In such an application, the aim is to identify a specific pattern that predicts the onset of a seizure, and on the occurrence of that pattern to stimulate the Vagus nerve to prevent the seizure from occurring.

Electrical stimulation of the Vagus nerve (Vagus Nerve Stimulation or VNS) has been reported as a method for treating epilepsy and more recently depression and is under investigation for treatment of Alzheimer's disease, anxiety and bulimia. VNS combined with normal use of anti-epilepsy drugs has been shown to reduce seizure frequency and to improve the condition of nearly 50% of patients, almost irrespective of age, Still, the application of the method is based on empirical studies, and it is uncertain exactly in what way VNS affects seizures and when it should be applied.

Typically, electrical stimulation of the Vagus nerve is carried out using the commercially available "helix-cuff electrode, coupled to an implanted electrical stimulator that operates periodically with the option of manual operation in case the patient senses a seizure that has not been suppressed. Existing VNS schemes suffer from a number of serious drawbacks. In particular, periodic stimulation is irrelevant to the possible occurrence of a seizure and power drain is high since the stimulator is operated continually. The device therefore requires battery change - and therefore surgery - every few years. Moreover, the interface used is unsophisticated, as it features no recording capability that could allow monitoring of seizure-related neural activity and no advanced stimulation techniques can be employed, as would be the case with a conventional tripolar cuff.

In addition to being used to stimulate nerves, cuff electrodes have also been used to measure and record activity within the peripheral nervous system. One example is described in Struijk J. J., Thomsen, M., Larsen, J. O. and Sinkjaer,

T. "Cuff electrodes for long-term recording of natural sensory information" lEEE Eng. Med. Biol, vol. 18, 91-98, 1999. Theoretically, the same cuff could be used for both recording and stimulation, although not simultaneously, as large stimulus currents will interfere with detection of the smaller neural electrical signals.

Further existing approaches to stimulation are described in the following publications: • Deurloo K. E. L, Holsheimer, J., and Bergveld, P. "Nerve Stimulation

With A Multi-Contact Cuff Electrode: Validation Of Model Predictions" Archives of Physiology and Biochemistry, vol. 108, 349-359, 2000.

• Deurloo K. E. L, Holsheimer, J., and Bergveld, P. "Fascicular Selectivity in Transverse Stimulation with a Nerve Cuff Electrode: A Theoretical Approach" Neuromodulation, vol. 6, 258-269, 2003.

• Veraart C, Grill, W.M., and Mortimer, J.T. "Selective control of muscle activation with a multipolar nerve cuff electrode" IEEE Trans. Biomed. Eng., vol. 40, 640-653, 1993. • Choi A. Q., Cavanaugh, IK. , and Durand, D.M. "Selectivity of multiple-contact nerve cuff electrodes: a simulation analysis" IEEE Trans. Biomed. Eng., vol. 48, 165-172, 2001.

• Riso R., Dalmose, A., Stefania, D., and Schuttler, M. "Addition of an intrafascicular electrode at the site of application of a multipolar nerve cuff enhances the opportunity for selective fascicular activation" 23^rd Ann. Int. Con/. IEEEEMBS, pp. 711-714, Istanbul, Turkey, 2001.

• Tarler M. D., and Mortimer, J.T. "Selective and independent activation of four motor fascicles using a four contact nerve-cuff electrode" IEEE Trans. Rehab. Eng., vol. 12, 251-257, 2004.

• Rijkhoff N. J., Holsheimer, J., Koldewijn, EX., Struijk, IJ., van Kerrebroeck, P.E., Debruyne, F.M., and Wijkstra, H. "Selective stimulation of sacral nerve roots for bladder control: a study by computer modelling" IEEE Trans. Biomed. Eng., vol. 41 , 413-424, 1994.

• Vuckovic A., Rijkhoff _j N. J. M,, and Struijk, JJ. "Different pulse shapes to obtain small fiber selective activation by anodal blocking : a stimulation study" IEEE Trans. Biomed. Eng., vol. 51 , 698-706, 2004.

These prior art approaches display a variety of limitations, including lack of sensitivity: it is often difficult to achieve adequate stimulation for "deep" fascicles within the nerve bundle. If one tries to increase selectivity and penetration by increasing the current, one tends to find that the higher currents generate unwanted activation of fascicles close to the periphery of the bundle.

In addition, of course, increasing the current or using multiple current sources also increases power consumption. That is extremely undesirable, particularly for implanted devices which take their power from an implanted battery. According to a first aspect of the present invention there is provided a nerve cuff having first and second parallel segmented ring electrodes for stimulating a nerve within the cuff, and a controller for repeatedly activating a selected segment of the first electrode and a corresponding segment of the second electrode on an opposite side of the nerve from that of the first segment.

Each individually controllable segment is preferably activated by applying an appropriate voltage to it, with one of the electrodes acting as anode and the other as a cathode. The summation of currents due to subsequent activation of different segments causes current flows within the volume surrounded by the cuff, with relatively high aggregate current in the centre and relatively low currents around the periphery. This allows fascicles which are deep within the nerve bundle to be stimulated without at the same time stimulating other fascicles on the periphery of the bundle.

The controller preferably comprises on-board electronics such as a circuit board offering programmable amplitude, timing and shape to the output stimulus pulse. Moreover the circuit provides the means for signal transmission to an external device (not shown here) and reception of command signals from the external device, if necessary.

According to a second aspect of the invention there is provided a method of stimulating a nerve comprising surrounding the nerve with a cuff having first and second parallel segmented ring electrodes, and repeatedly activating a selected segment of the first electrode and a corresponding segment of the second electrode on an opposite side of the nerve from that of the first segment.

According to a third aspect there is provided a method of treatment comprising surrounding a nerve of the peripheral nervous system with a cuff having first and second parallel segmented ring electrodes, monitoring the activity of the nerve and, in dependence upon said activity, stimulating the nerve by repeatedly activating a selected segment of the first electrode and a corresponding segment of the second electrode on an opposite side of the nerve from that of the first segment.

The invention extends to a method of multiplexing a number of stimulus electrode pairs, quickly enough so that the nerve reacts as if all were stimulating simultaneously, while the power consumed corresponds to only one electrode pair applying stimulation. It also extends to a method of recording neural signals for diagnosis and treatment comprising surrounding a nerve of the peripheral nervous system with a cuff, preferably tripolar, having a middle segmented ring electrode, and connecting subsequent groups of segments together sequentially to scan the circumference of the nerve.

The invention may be carried into practice in a number of ways and one specific embodiment will now be described, by way of example, with reference to the accompanying drawings, in which:

Figure 1 shows a cross section of a typical nerve bundle illustrating the way in which the nerve fibres are grouped into fascicles;

Figure 2 gives an indication of possible paths of ionic currents flowing between fascicles;

Figure 3 shows a nerve cuff in accordance with an embodiment of the invention;

Figure 4 shows the middle segmented ring electrode of the recording configuration, and how the segments are connected together to reduce electrode noise; Figure 5 shows exemplary electronics providing a convenient method for sweeping between electrode combinations; and

Figure 6 illustrates the radial selectivity that has been obtained in an experimental setup in which stimulation was provided in accordance with an embodiment of the invention.

A typical nerve bundle around which the cuff is to be located is shown in Figure 1. The bundle 20 consists of several fascicles 22, each fascicle containing many individual nerve fibres 24. When an individual fibre or neuron fires, it exchanges ions with the surrounding extracellular space, through specific so-called "ion pumps" or "ion-channels". The firing of a group of fibres in one or more fascicles 22 creates ionic currents 32, as shown in Figure 2, that flow between the fascicles to the extracellular space. Placing a recording cuff around the bundle allows for the detection of the corresponding action potentials. As action potentials propagate along nerve fibres, ionic currents flow through the medium between the nerve bundle and the surrounding cuff, which can then be detected. The space restriction between the nerve bundle and the cuff increases the amplitude of the signal detected extraneurally. Thus, the local ionic flow that takes place during an action potential occurrence is concentrated inside the recording cuff.

An implantable cuff according to an embodiment of the present invention is shown schematically in Figure 3. The device comprises a cuff body 30 which in use is positioned around a nerve bundle 20 within which are individual fascicles to be stimulated. On the body there are first and second ring-shaped end electrodes 32, 34, along with a pair of central electrodes 36, 38. Each central electrode is split up into a number of small segments, each segment being individually controllable. All of the electrodes are coupled to an attached circuit board 40 which receives and analyses the signals, as well as controlling data transmission to/from an external unit (not shown). Power may be provided via wires (not shown) to an external or implanted battery, or alternatively power may be supplied wirelessly through the skin of the patient.

In order to stimulate a desired fascicle 22 within the interior of the nerve bundle 20, appropriately controlled voltages are selectively applied to the individual segments of the central electrodes 36, 38. The aim is to create a focussing effect, so as to stimulate a particular chosen "focus point" 50 within the bundle, while avoid stimulation of other fascicles. One way of doing this is to create a rotating dipole by simultaneously applying voltage to opposing segments 100, 102, the first of these being on one of the central electrodes and the second on the other. The dipole is then rotated by repeatedly activating similar opposing pairs of segments, segment by segment. Depending on the focus point location relative to the centre, the segments forming a pair need not necessarily be diametrically opposed.

Each individual stimulus created by the pair is chosen to be adequately low in amplitude, so that it does not stimulate any fibres on its own. Such a stimulus current just decreases the activation membrane threshold of the fibres it transverses. Altering stimulus pulse duration affects the stimulation of fibres in different distances from the electrodes, as reported in Grill, W.M., and Mortimer, J.T. "The Effect of Stimulus Pulse Duration on Selectivity of Neural Stimulation" IEEE Transactions on Biomedical Engineering ,vol. 43, no. 2, 1996. Similarly, by appropriately timing the "on" duration of each pair of segments the charge accumulated in the "focus point" causes the membrane thresholds of fibres close to it to exceed the so-called "firing" value and the targeted fibres are then activated, generating action potentials. Stimulus pulse shape and timing adjustments also allow stimulation to be selective to specific fibre diameters, as reported in Vuckovic A., Rijkhoff, N. J. M., and Struijk, JJ. "Different pulse shapes to obtain small fiber selective activation by anodal blocking : a stimulation study" IEEE Trans. Biomed, Eng., vol. 51, 698-706, 2004.

Depending upon the particular application requirements, it may be desirable to activate not just a single segment on each electrode, but several consecutive segments. One might, for example, activate five consecutive segments on one electrode along with five opposing segments of the other electrode, and then cause those blocks to rotate by repeatedly switching off the segment at the trailing end and switching on a new segment at the leading end. In some applications it may even be desirable to activate almost all of the segments of each electrode at the same time, leaving only a small gap of non-activated electrodes. That gap can then be rotated in an analogous manner.

Appropriate stimulus waveforms, controlled by the circuit board 40, may be chosen to eliminate or at least reduce unwanted neural activity.

The current drawn may where appropriate be further reduced using multiplexing techniques. Thus, instead of simultaneously activating two or more pairs of electrodes to create a "focused" current, these pairs may be activated sequentially, quickly enough that the ion channels respond to the accumulated rather than to the instantaneous charge. If x number of pairs were activated at the same time, each with current amplitude A, the overall current would be x*A. By using multiplexing in this way, a single source may suffice, reducing the overall power by x.

For additional flexibility, the two central electrodes may be slightly offset from a central position between the end electrodes 32, 34. Such an arrangement makes it easier to stimulate in the sensory or motor direction, rather than producing a stimulus pulse that propagates in both directions along the nerves.

It is envisaged that in alternative embodiments (not shown) more than two central electrodes may be provided. This will provide additional flexibility in control, and may further reduce the current density at the periphery of the bundle 20, while maintaining the necessary high current density at the focal point 50.

When not being used for stimulation, the device described above may also be used for recording purposes, in a similar manner to a conventional tripolar recording cuff as described for example in Struijk J. J., Haugland, M.K., and Thomsen, M. "Fascicle Selective Recording with a Nerve Cuff Electrode" 18th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Amsterdam, 1996; and Rozman J., Zorko, B., Bunc, M., Mikac, U._; and Tegou, E. "Recording of ENGs from the nerves innervating the pancreas of a dog during the intravenous glucose tolerance test", Physiological Measurement, vol. 23, 695-705, 2002.

However, the small area of the electrode segments would translate into higher electrode noise relative to full ring electrodes, as reported in Koch K. P., Schuettler, M., and Stieglitz, T. "Considerations on Noise of Amplifiers and Electrodes for Bioelectrical Signal Recording" Proc. 7th An. Conf. IFESS, pp. 202-204, Ljubljana, Slovenia, 2002. In combination with the small amplitude of the neural signals, this would degrade the quality of recordings. Reducing the noise by increasing the width of the electrode segments would reduce the amplitude of the recorded neural signal, according to a theory presented in Struijk J. L, and Thomsen, M. "Tripolar Nerve Cuff Recording: Stimulus Artifact, EMG, and the Recorded Nerve Signal" IEEE - EMBS 17th Ann. Int. Conf., Montreal, 1995.

However, these disadvantages can be mitigated or overcome by multiplexing. As shown in Figure 4, multiplexing can be applied to multiple segments 61 using sampling frequencies sufficiently higher than the upper limit of the ENG bandwidth (eg 1 OkHz), to combine segments for selective recording. That makes possible a configuration 62 where, instead of scanning the nerve perimeter segment-by-segment, the scanning process involves multiple segments connected together, leaving a recording "gap". In that fashion, the recorded signals from a complete 360° rotation can be accumulated resulting in a signal peaking close to the side of the most active fascicle. Even if a deep fascicle is active, the signal amplitude will still be sufficiently high, as the electrode will always cover most of the periphery. Using this method, electrode noise is significantly lower, as the area of the recording electrode is much greater than in the case of a single segment.

The overall system may be configured automatically to stimulate the nerve on the measurement of a particular pattern of chemical activity. For the Vagus nerve, stimulation may be provided when the detected pattern of activity indicates that a seizure may be imminent. In this case, stimulation will not be used for activation of muscle groups or organs but rather for neuromodulation. This is a method of essentially stopping neural activity, thus blocking seizure- related motor signals when measurements indicate they are about to occur.

As an alternative to multiplexing, switching circuits can be used, such as the programmable circuitry shown in figure 5. This particular setup is an example that can be used for sweeping between electrode combinations in a tripole cuff with an 8-segment middle electrode. It consists of a PIC 16F818 microprocessor programmable to 16 different electrode settings that uploads the switching strategy to two 74HC595 shift registers that control four TS3A5018 four-SPST CMOS switch ICs. The circuit gives the potential for QT, TT and cross segment bipolar and tripolar stimulation.

The graph of figure 6 illustrate the radial selectivity that can be obtained in an experimental setting using an 8-segment mid electrode tripole stimulation cuff. An 8-segment mid electrode tripole recording cuff was used for recording,

In the upper four grafts, the fixed pair of opposite segments was activated in the stimulating cuff, while recording from an adaptive tripole (AT), the middle connection of which was swept around the mid electrode ring. Graph (a) shows the use of a single swept recording electrode, graph (b) shows two swept recording electrodes, graph (c) four swept recording electrodes, and graph (d) seven swept recording electrodes.

Graphs (e) and (f) show an alternative stimulation arrangement, in which the stimulating electrodes are separated by two (for example, stimulation occurs between segments 2 and 4). Graph (e) illustrates the use of two swept recording electrodes and graph (f) seven swept recording electrodes.

It will be easily seen from the graphs that the presence of an air bubble effects the recordings when one and two recording electrodes are used (graphs (a), (b) and (e)). Combinations of more recording electrodes still exhibit radial selectivity, even though the signal is averaged across the shortened segments. It will be evident from a comparison between graph (b) and (e), and also between (d) and (f), that the use of multiple rotating stimulating electrodes improves selectivity. It will also be seen that the use of multiple recording electrodes also improves the output signal.

Claims

1. A nerve cuff having first (36) and second (38) parallel segmented ring electrodes for stimulating a nerve within the cuff, and a controller (40) for repeatedly activating a selected segment of the first electrode and a corresponding segment of the second electrode on an opposite side of the nerve from that of the first segment.

2. A nerve cuff as claimed in claim 1 in which the selected and opposing segments are activated simultaneously .

3. A nerve cuff as claimed in claim 1 in which the controller repeatedly changes the selected segment to generate a rotating field within the cuff.

4. A nerve cuff as claimed in any one of the preceding claims in which the controller repeatedly activates a selected block of segments of the first electrode and a corresponding block of segments on the second electrode.

5. A nerve cuff as claimed in any one of the preceding claims including a first end electrode (32) for monitoring electrical signals.

6. A nerve cuff as claimed in claim 5 including a second end electrode (34) for monitoring electrical signals, the first (36) and second (38) electrodes being located between the end electrodes.

7. A nerve cuff as claimed in claim 6 in which the first and second electrodes are centrally-positioned between the end electrodes.

8. A nerve cuff as claimed in claim 6 in which a central point between the first and second electrodes is offset from a central point between the end electrodes.

9. A nerve cuff as claimed in claim 4 in which the respective blocks of segments each comprise the majority of the respective electrode, with a gap between the block ends defined by a non-activated segment or segments.

10. A nerve cuff as claimed in claim 9 in which the controller causes the gap between the block ends to rotate about 360 degrees.

11. A nerve cuff as claimed in any one of the preceding claims in which, in operation, the controller (40) sequentially activates a first then a second segment of the first electrode, and also sequentially activates a first then a second segment of the second electrode, in each case sufficiently rapidly that the nerve responds to the cumulative rather than to the individual activations.

12. A method of stimulating a nerve comprising surrounding the nerve with a cuff having first (36) and second (38) parallel segmented ring electrodes, and repeatedly activating a selected segment of the first electrode and a corresponding of the second electrode on an opposite side of the nerve from that of the first segment.

13. A method of stimulating a nerve as claimed in claim 12 comprising sequentially activating a first then a second segment of the first electrode, and also sequentially activating a first then a second segment of the second electrode, in each case sufficiently rapidly that the nerve responds to the cumulative rather than to the individual activations.

14. A method of treatment comprising surrounding a nerve of the peripheral nervous system with a cuff having first (36) and second (38) parallel segmented ring electrodes, monitoring the activity of the nerve and, in dependence upon said activity, stimulating the nerve by repeatedly activating a selected segment of the first electrode and a corresponding segment of the second electrode on an opposite side of the nerve from that of the first segment.

15, A nerve cuff as claimed in claim 1 including programmable switching circuitry for controlling activation of said selected first and second electrode segments.