Adaptive senso-motor encoder for neuroprostheses

P:OPER\RABklientt\misc * dbw\in(elligent implants ISPA spec doc-18/03/02

- 1 ADAPTIVE SENSORY-MOTOR ENCODER FOR NEUROPROSTHESES

The invention concerns an adaptive sensory-motor encoder as well as a spinal implant and a cranial implant.

There are several precursor systems of spinal implants known, that, for example, are employed in cases of transverse lesions of the spinal cord with paraplega for control of urinary tract functions and for guidance of ambulating movements or grasping movements that work by means of stimulation contacts in the form of implants or that work 10 transcutaneously (see Eckmiller and colleagues., Neurotechnology Report, 1994, and 1995).

The spinal implant precursors currently available or those in development have diverse limitations, for example, no adaptation, no functional increase of number of 15 microcontacts and no bi-directional, perception-based control by the implant carrier.

In particular, the currently developed microcontact structures and signal and energy transfer systems operate unidirectionally from the external encoder to the implanted stimulator and therefore offer no possibility of ongoing monitoring of neural impulse 20 activity of the stimulated neurons. Thus, the stimulation pulse sequence cannot be adapted to the spontaneous activity of the neurons. Furthermore, triggering of neurobiological impulses by stimulation pulses can not be monitored directly. Moreover, an assured impulse monitoring opportunity for possible temporal tuning and synchronization of the impulse sequences of multiple neurons is also lacking.

There are isolated rationales for development of implanted, active substance applicators that are controlled by need, for example for insulin, but there have been to date no cranial implants that have been successfully implemented. Cranial implants that, for example, are urgently needed for local, event-triggered administration of active substances for suppression of onset of epileptic events, are not available.

<img file="AU747686B2_D0001.tif" />

R'OPERXRAB'clientsVmisc · dbw\inielligent implants ISPA specdoc-18/03/02

-2It is desired to alleviate one or more difficulties of the prior art, or at least provide a useful alternative.

In accordance with the present invention, there is provided an adaptive sensory5 motor encoder for use as a neuroprosthesis in the central nervous system within the cranium or the spinal cord or within the peripheral nervous system, said encoder comprising:

a central control unit, the central control unit carrying out at least one of signal processing functions, monitoring functions, checking functions, and external interventional 10 functions, the central control unit being in bi-directional communication with a microstructure capable of being implanted for at least one of stimulation of nerve or glial tissue and monitoring of brain or other neural functions, said bi-directional communication being carried out via a bi-directional interface coupling said microstructure with the encoder; and at least two adaptive spatio-temporal filters, the spatio-temporal filters carrying out at least one of sensory or motor stimulation functions and monitoring functions.

The present invention also provides a neuroprosthesis capable of being used in the central nervous system within the cranium or the spinal cord or within the peripheral 20 nervous system, comprising:

an adaptive sensory-motor encoder comprising:

a central control unit, the central control unit carrying out at least one of signal processing functions, monitoring functions, checking functions, and external interventional functions; and at least two adaptive spatio-temporal filters, the spatio-temporal filters carrying out at least one of sensory or motor stimulation functions and monitoring functions;

a microstructure capable of being implanted for at least one of stimulation of nerve or glial tissue and monitoring of brain or other neural functions; and .30 a bi-directional interface for coupling the implanted microstructure with the

<img file="AU747686B2_D0002.tif" />

V- x'encoder.

RVOI'ERXRABklienuVnisc · dbwMnlelhgent implants I SPA spec.doc-18/03/02

-3The present invention also provides a method of operating a neuroprosthesis, the neuroprosthesis including an encoder that includes at least a central control unit and at least one adaptive spatio-temporal filter, a microstructure capable of being implanted in an implant carrier, and a bi-directional interface, said method comprising the steps of:

performing, using the central control unit, at least one of signal processing functions, monitoring functions, checking functions, and external interventional functions;

performing, using the at least one adaptive spatio-temporal filter, at least one of sensory or motor functions;

communicating bi-directionally between the encoder and the microstructure via the bi-directional interface; and performing, using the microstructure, at least one of stimulation and monitoring of nerve or glial tissue.

Because the encoder of the preferred embodiment is bi-directionally coupled with implanted microcontacts, monitoring of the neural impulse activity of individual neurons to be stimulated and other signals and the execution of quasi-autonomous actions can be realized. The functions can be optimized either self-actuating by the neural network or in dialogue with the implant carrier. The number of the selectively addressable stimulation sites can be functionally increased and the neural activity of individual neurons monitored. The implanted structure can operate sensory-motor quasi-autonomously by using appropriate sensory and action components and an adaptive control system. Components and processes of the adaptive information processing system are implemented in various combinations, particularly for spinal implants in bi-directional contact with the spinal cord or the peripheral nervous system and for cranial implants in bi-directional contact with the structures of the central nervous system within the cranium.

Furthermore, the encoder of the preferred embodiment allows the number of selectively reachable stimulation sites to be functionally increased and also subsequently to

<img file="AU747686B2_D0003.tif" />

adapt itself to new stimulation conditions. The encoder described herein can (on the basis \^of its structure and function as a group of adaptive spatio-temporal filters) in addition to

P:VOPER\RABklients\mi$c · dbvAjnlelligent implants ISPA spec.doc-18/03/02

-4the stimulation function, also perform monitoring an evaluation of the neural activity of the neurons to be stimulated. The spatio-temporal filters associated with the individual microcontacts, to the extent possible, are tuned to optimum function individually in the dialogue between the encoder and the implant carrier.

In contrast with an encoder with static pre-processing; that is, without the possibility of individual adjustment, the encoder of the preferred embodiment allows, on the basis of the single relevant criterion; namely the specific functional enhancement of the given area of the nervous system, adjustment of the single spatio-temporal filters as 10 separate encoder channels. This advantage includes the possibility that subsequent function changes, for example, as a result of shift of micro-contacts by corresponding adaptations of the spatio-temporal filter function, can be compensated for. An advantage of the tuning of the spatio-temporal filter function in the dialogue with the implant carrier or with an area of the carrier’s nervous system is in the consideration of functional aspects, 15 that only the actual implant carrier can incorporate into the optimization process and only in implicit form; namely, for example, by subjective assessment of his perception or by evaluation and function monitoring of his nervous system and their use in the encoder adjustment.

The asynchronous impulse sequences of the individual spatio-temporal filter outputs of the functionally separated encoder channels, as stimulation signals, selective stimulation sites are selectively tuned to one another in the dialogue with the implant carrier, in consideration of the neural impulses recorded at the stimulation site.

Because it is presumed that time courses and locale distributions of the stimulation signals that have been reciprocally tuned by superposition, have been suitably selected by an adaptation process and their field distributions effected at several microcontacts will, as stimulation foci, trigger local and temporal selective neural impulse excitations, the number of selectively addressable stimulation sites and their definition or cross-talk

<img file="AU747686B2_D0004.tif" />

suppression will be functionally enhanced with fixed number of implanted microcontacts.

R\OI'ER\RABklienU\nusc · dbwMntetligent implants I SPA specdoc-18/03/02

-5With a given, relatively low number of implanted and permanently functional microcontacts, whose position relative to the neurons can not be modified, it is of particular advantage, functionally; that is, by generation of suitable signals, to increase the number of selectively reachable stimulation sites or neurons and thus, at the same time, 5 increase the number of separately accessible encoder channels with an adequate reserve of spatio-temporal filters. This advantage effects an improvement of the quality of the respective function.

The control or relief of defective functions of the spinal cord or peripheral nervous 10 system with the aid of a partially implanted neuroprosthesis in the closest possible sensory and motor coupling with the implant carrier and by using quasi-autonomous sensory-motor functions of the implanted structure is thus made possible.

The present invention also provides an adaptive spinal implant for alleviation of 15 neural functional impairment in the spinal cord or peripheral nervous system, comprising:

an adaptive encoder;

a sensory-motor structure capable of being implanted and that functions in a quasiautonomous fashion; and a bi-directional signal and energy transfer system, wherein the structure capable of being implanted and the transfer system function so as to facilitate bi-directional contact between the encoder and parts of at least one of the spinal cord, the peripheral nervous system, and a muscle group.

Using adaptive spinal implants the quality of the relief from functional impairments in the spinal cord or peripheral nervous system fundamentally improved and, with respect to diverse applications, is possible for the first time.

The present invention also provides an adaptive cranial implant apparatus for alleviation of neural functional impairments of the central nervous system, comprising:

an adaptive encoder that includes:

a central control unit, the central control unit performing at least one of

<img file="AU747686B2_D0005.tif" />

P:VOI*ER\RAB\clienU\mi$c * dbw\inielligcni implants 1SPA spec.doc-18/03/02

-6signal processing functions, monitoring functions, checking functions, and external intervention functions;

a structure capable of being implanted, functioning in quasi-autonomous fashion, and which may also receive control inputs from the adaptive encoder, comprising:

at least two microcontacts;

at least two detectors;

at least one active substance depot; and means for administering the active substance from the active substance depot; and a signal and energy transfer system for facilitating bi-directional contact between the adaptive encoder and the structure capable of being implanted, and thus with the brain.

Alleviation of neural functional impairments of the central nervous system within the cranium is made possible, particularly for the purpose of reducing undesirable sensory, 15 motor, or cognitive effects for a number of groups of neurological or psychiatric patients using an implanted structure with an active substance applicator and quasi-autonomous, sensory-motor functions in coupling with control and monitoring functions of the implant carrier.

For the first time, using adaptive cranial implants the quality of the relief of neural functional impairments in the central nervous system within the cranium is possible in diverse applications.

Preferred embodiments of the present invention are hereinafter described, by way 25 of example only, with reference to the accompanying drawings, wherein:

Figure 1 is a schematic illustration of a preferred embodiment of a spinal implant in the area of the nerve tissue with a control unit and a command module activated by head movements;

Figure 2 is a schematic illustration of a preferred embodiment of an implanted microcontact structure for stimulation of nerve tissue not directly contacted;

Figure 3 is a schematic illustration of a preferred embodiment of a cranial implant

<img file="AU747686B2_D0006.tif" />

P:\OPER\RAB\dientsVnisc-dbwMntelligenl implants 1SPA spec doc-18/03/02

-7 with the various modules in a block circuit diagram;

Figure 4 is a block diagram of a digital signal processor-based implementation of the encoder according to a first preferred embodiment of the invention; and

Figure 5 is a block diagram of an implementation of the encoder according to a 5 second preferred embodiment of the invention.

A cranial implant used as a grasp prosthesis is illustrated in Figure 1. This prosthesis is used when arm movements can still be performed in a paraplegic, but gripping movements with the hand are not possible.

The grasp prosthesis includes an implant (1) that communicates with a transcutaneous module (2). The module (2) is, in turn, in bi-directional, wireless connection with an adaptive central control unit (3). The control unit receives commands from a command module (4) that is guided by a head movement sensor (5).

The implant (1) is in contact with nerve tissue (8) via micro-contacts (6, 7). In addition, several sensors (9), that pick up measurement values for the grip process in the area of the terminal thumb and one other finger, are connected to the implant (1).

The system described so far operates in the following manner to produce and control a grip process. Initially the user moves the hand in the direction of an object that he intends to grasp. For example, the object can be a glass. When the hand is in the right position for grasping the glass, the grasp process is triggered by a particular head movement that is recognized by the head movement sensor (5). Then, the head movement sensor (5) communicates its signals to the command module (4). The command module (4) generates the appropriate control commands, which are then communicated to the control unit (3) either wirelessly or by signal conduction. The command interpreter (11), which also provides movement monitoring, is situated in an initial area (10) of the control unit (3). A second area (11) of the control unit (3) includes the position and strength control. The grasp command coming from the command module (4) is interpreted in the unit (10) and an appropriate movement stimulation pattern is determined for the action.

<img file="AU747686B2_D0007.tif" />

RVDI'ER\RABk;lienls\misc * dbw’iintelligenl implants ISPA spec.doe-18/03/02

- 8 The unit (11) in the central control unit (3) then issues positioning commands to the transcutaneous module (2), which then issues the command to the implant (1) and in particular to a stimulation unit (12) therein for stimulation of the appropriate neural pathways (8) via the micro-contacts (6, 7). The neural pathways (8) control the electrical 5 stimulation of the hand muscles required for the grasping process, which then contract and trigger the grasping process.

The sensors (9) now detect the grip of the hand around the objects by picking up information regarding the position, the pressure applied, and slippage of the object. The 10 goal of a grasp process is exertion of pressure that is as slight as possible in order not to damage the object (e.g., a raw egg), on the other hand, however, also to provide a grasp that is slippage free and so prevent dropping the object.

The sensor (9) signals are transferred to a sensor unit (13) of the implant (1). The 15 signals received are transferred by the sensor unit (13) to the transcutaneous module (2), which then delivers the values determined after signal processing to the checking unit (11) of the control unit (3), where position and strength regulation for the grasping process is done. In the individual case, the stimulation of the nerve tract (8) via the micro-contacts (6) is guided in such a manner that an optimal grasp event results. The user can, for 20 example, after grasping the glass, issue the information, by another head movement, to the head movement sensor (5) and thus to the command module (4) that the grasp process can now be sustained independently. The control unit (3) then effects the autonomous control of the grasp process. As an adaptive control unit (3) the stimulation is retro-coupled on a regular basis by way of the micro-contacts (6, 7) in the area of the nerve tract (8) in that the 25 activity of the nerve tract (8) via the sensors (9) is picked up in the sensor unit (13) of the implant (1). The control unit (3) integrates this adaptability via a neural network that so regulates the site, the strength and the time course of stimulation of the nerve tract (8) that those nerve tracts are stimulated as precisely as possible and in the appropriate strength that is required for the grasp process. Thus, after an adaptive phase, an optimal grasp

<img file="AU747686B2_D0008.tif" />

process is possible.

Pi\01*ER\RAB&gt;cliails\nusc -dbw\inlelligen( implants ISPA spec.doc-18/03/02

-8AFigure 2 shows an illustrative example of micro-contacts (6) that impinge into nerve tissue (8). In the present example three micro-contacts (16, 17, 18) are implanted into the nerve tissue (8) and are positioned more or less randomly near certain nerve cells. The micro-contact structure (6, 16, 17, 18) is in each case essentially coarser than the 5 matrix of the nerve cells (8). The micro-contacts (16, 17, 18) are supplied with signals (SI, S2, S3) via the stimulator (12).

In order to create a targeted neural stimulation, for example, a stimulation focus F must be attained that can not be directly affected by a micro-contact. The stimulation 10 focus F can, however be attained, by conducting the signals (SI, S2, and S3) using different strengths, time courses and, above all, temporal intervals to the electrodes (16, 17, 18). The overlap of the signals generated can then be so arranged that the convergence of the signals in the area of the intended stimulation focus F exceeds the stimulation threshold of single or a few nerve cells, while the addition of signal progressions in the remaining 15 area of the nerve tissue remains below the excitation threshold.

By changing the temporal sequence and the temporal signal progression of the various reciprocally tuned signals the stimulation focus can also be shifted from F to F’. For balancing of the stimulation functions that reach a stimulation focus which is not 20 directly in conjunction with electrodes, an adaptive process is required. Since it is not exactly known which stimulation focus F, F’ must be addressed for a particular neural stimulation, the adaptive sensory-motor control unit can offer only a certain signal pattern which the implant carrier then assess via a sensory perception or another sensor data evaluation. A second signal pattern, which is modified with respect to the first, is then 25 likewise evaluated as to whether the intended neural stimulation is reached or not. The user needs only to say whether the later signal pattern better or worse suited is than the previous one. Using the check mechanism of a neural network and in the course of the checking process, an optimal signal time function is determined for the electrodes (16, 17,

18) for the purpose of excitation of the stimulation focus F. Figure 3 depicts a cranial

<img file="AU747686B2_D0009.tif" />

implant for monitoring and influencing a region of nerve tissue in the cortex. A neural tissue (20) that is to be monitored is provided with micro-contacts (22) in the area of a

P:VOPER\RAEPdients\mi$c * dbwMntelligcnt implants ISPA spec.doc-18/03/02

-8Bregion (21). The region (21) can, for example, be that region whose impaired function results in epileptic events. In such a case, a micro-contact structure (22) includes both the micro-contacts for stimulation of the nerve cells and also microsensors for monitoring the nerve cell activity and other biophysical parameters. The micro-contact structure (22) 5 itself is connected to a signal path (23) with an adaptive processor (24) that has available to it a bi-directionally operating transmission and reception system. Otherwise, the microcontact structure (22) is in contact with a signal tract (25) with an active substance reservoir (26). The active substance reservoir (26) is prepared for controlled local release of small quantities of an active substance, for example, in the nanoliter range. Finally, the 10 processor (24) is connected via another signal tract (27) directly to the active substance reservoir (26).

The three modules (22, 24, and 26) of the cranial implant function cooperatively to prevent epileptic events in the following way: The microsensors monitor the spontaneous 15 nerve cell activity in the region (21) and communicate their measurement signals over the signal tract (23) to the processor (24) which then on the one hand evaluates the signals itself and, on the other hand, communicates a status report to an external encoder. With the occurrence of suspect characteristic nerve cell activity, which can, for example, be expressed in synchronous, neural activity within the region (21), the processor (24) 20 identifies the existence of a stimulation pattern that can result in an epileptic event. It can then communicate on the one hand over the signal tract (23) stimulation pulse sequences to the micro-sensor (22), which counteract such synchronous, neural stimulation. When this purely electrical intervention is inadequate, the processor (24) can communicate an instruction over the signal tract (27) to the active substance reservoir (26). The reservoir 25 administers a precisely metered dose of a pharmacological active substance, which in turn is capable of lowering the synchronous nerve cell activity to a normal level.

The type of control process, its temporal sequence, and the power of the respective effect on the nerve tissue region (21) by means of the micro-contact structure (22) and of the active substance reservoir is set and optimized at an external encoder, such that in the ideal scenario the patient is unaware of the artificial preventive intervention. In the case of

P:\Ol’ER\RABkljenU\misc · dbwVintelligent implants ISPA spec.doc-18/03/02

- 8C physiological changes in the patient, the mode of the cranial implant is also further adapted. When this is done the encoder has the fundamental functions, that have already been described in connection with the design examples in Figure 1 and Figure 2.

An advantageous form of the structure and function of the encoders is shown in

Figure 4 and includes a digital signal processor (DSP) (30), for example, the Texas Instruments C80 model, which is combined with a pre-processing module (not shown), a pulse signal emitter and receiver (not shown) for bi-directional communication with the implanted structure, several signal interfaces, particularly for communication with the 10 evaluation input unit and the external monitoring and control (i.e., checking) system (not shown). The various adaptive information-processing functions are realized, particularly for spatio-temporal filters, dialogue module and pattern identification, in the DSP (30) with a central control unit (31). As is inherent in such DSPs, there is associated memory (32) that stores software implementing the various functions implemented by the DSP (30), the 15 software being executed by central control unit (31) to implement its functions. The user receives, on the one hand, signals as stimulation pulses or sensory perceptions from the encoder and sends, on the other hand, the evaluation input of biophysical parameters and neural activity to the encoder. Furthermore, there is an exemplary form of the encoder in which the encoder, because of the bi-directional wireless signal and energy transfer, can be 20 attached to the body or to a body-remote location. Further, an advantageous form of the encoder is where the spatio-temporal function space of the spatio-temporal filters for use as an encoder includes the information processing properties of the various classes of neurons contacted.

While the encoder may be implemented by means of a DSP (30), as discussed above, this is merely an exemplary embodiment, and other implementations are contemplated. For example, as shown in Figure 5, an encoder (40) may include separate modules implementing central control unit (41) (i.e., except for the adaptive spatiotemporal filters) and adaptive spatio-temporal filters (42). Each of these may be implemented, for example, using a DSP or by any other known implementational method.

<img file="AU747686B2_D0010.tif" />

RVOI’ERUlABklientttnusc · dbwMntclligeni implants 1SPA specdoc-18/03/02

-8DWith the encoder, a direct communication is established with part of the nervous system that, on the one hand, is already spontaneously active. Thus, neural impulses from single neurons are generated without technical stimulation. For optimal adaptation to the stimulation pulse sequences to the respective spontaneous activity, for precise 5 determination of the stimulation parameters for assured and, at the same time, biologically compatible 1:1 conversion of stimulation pulses into neural impulses, and for improved optimization of the temporal tuning and synchronization of the neural activity of several neurons, monitoring of the neural activity of single neurons to be stimulated is of considerable benefit.

With a design example of a process for adjustment of the spatio-temporal filters of the encoder in dialogue with the user the spatio-temporal filters are realized as spatiotemporal filters whose spatial and temporal function parameters are modified, within a sufficiently large functional space, in approximation of the respective information 15 processing features of neurons; namely, by using externally accessible parameter exchange points placed at appropriate locations in the filter algorithm. An advantageous form of the adaptive process for the spatio-temporal filters is where a person, as a normal healthy person or as an implant carrier, in a perception-based dialogue with the encoder, the perceptual comparison between the desired pattern and the actual pattern; for example, by 20 using as the evaluation input unit a row of several sliding controls or in the form of particular encoded head movements, communicates with a technical neural network with non-monitored adaptation rules and that the neural network establishes the next parameter vector for the spatio-temporal filters as well as the next desired value, with the goal of reducing the perceived pattern difference in the next dialogue step. An advantageous form 25 of the search for the optimal parameter vectors for the spatio-temporal filters is where, in the dialogue module either from a neural network with non-monitored adaptation

<img file="AU747686B2_D0011.tif" />

08/04/99 WED 08:03 FAX 1 809 924 1811 _WQ0dt&gt;rldge&amp;A68OCOCi.aTe6 @042 parameter vectors are produced that result In a particular perception for a given pattern presented and are correspondingly subjectively evaluated or, where in a dialogue module with 5 another parameter adjusting system sequences of parameter vectors are uaad for the production of virtual motion in the function space of the spatio-temporal filters; for example, as continuous trajectories depending on the type of scanning or sweep processes, or as nonregular sequences, or as sequences of neurophysiological, particularly typical filter functions. In the course of this suitable timed sequence, the user casually reports the ‘sensible perceptions produced by the collaboration of the given pattern, the pre-processing module, the following spatio-temporal filters and that part of the nervous system coupled to the associated microcontact and. then. In the region of the function apace determined in this process, a more precise parameter optimization can be undertaken based on perception.

An exemplary form of the generation of asynchronous impulse sequences le where the 15 output signals of the Individual spatio-temporal filters are transformed into asynchronous impulse sequence, relevant to the activity of the neurons, by using suitable converter algorithms of the quasl-contlnuous time functions of the spatio-temporal filters and where Impulse sequence - time courses and points in time of the occurrence of single impulses can be shifted in the dialogue phase by using variable time delay elements.

An exemplary form of the process for temporal coupling of the asynchronous Impulse sequences generated by several spatio-temporal filters of the encoders for the purposes of triggering nerve cell impulses is where the transmission time point of the Individual impulse signals are varied by controllable time delay elements In such a way that temporal coupling up to precise asynchronous occurrence results, wherein the variation of the time delay la regulated 25 by the Implant carrier. Or, in the dialogue this occurs based on perception via a neural network, or Is externally regulated. In that the selection of the impulse groups that are to be temporally coupled can take into consideration the Impulses coming both from the spatio-temporal filters and those recorded In the Interface and that, in view of the very different momentary impulse rates of the different spatio-temporal filters, suitable criteria are established for inclusion of

Individual impulses In the impulse groups that are to be coupled.

An exemplary example of the process for functional increase of the numtier and definition of the selectively addressable stimulation sites In the case of a given number of stationary Implanted micro-contacts is where the Impulse signals from a given spatio-temporal filter are conducted to several, locally adjacent micro-contacts, whereby the characteristic time

<img file="AU747686B2_D0012.tif" />

ZVZO89T91120

I xejsBueBsnv 393Ί 017 = 171 66-6nV-6T

08/04/09 WED 09:04 FAX 1 609 924 1811__ WoodDrl&lt;Jge«iA880C0ciatee @043 courses - that have been established for each micro-contacts and corresponding to current amplitude, polarity, and phase length in the interface - of the electromagnetic field In the area ₅ of the neurons to be stimulated have the effect that said stimulation signals, which have been tuned to each other by superposing, trigger locally and temporally selective neural Impulse excitations In the field distributions at several micro-contacts and that the selective stimulation sites can b. rapidly changed by suitable variation of the superpositioned stimulation signals and that the respective variation of diverse parameters of the reciprocally tuned stimulation signals results in the perception-based dialogue with the Implant carrier via a neural network, or other signal variation process for the determination of as many as possible simulation sites leading to selective and defined neural excitation. Further, this advantageous form consists in the optimization of the etimulatlon-time functions is improved with respect to Intended single cell selectivity and lasting biocompatibility through the comparison of recorded neural impulses to the stimulation signals.

' A exemplary form cf the monitoring system for a partially sensory-motor autonomously functioning Implanted structure of the encoder Is where the Implanted micro-contacts are used both for stimulation and for recording of neural impulses, In that the recorded impulses and other physical or chemical signals from the Implanted Structure are reported through suitable 20 pre-ampitfiere and optical or electromagnetic transmitters to the encoder and there, the recorded neural signals are further processed for the various purposes of the encoder function.

An exemplary form of the partially autonomous sensory-motor actions of the implanted structure Is where, by the use of various sensors accessible In the Implanted structure that detect physical or chemical values, or using various actors such as, for example, electrical 25 stimulation electrodes, mechanical micro-actors, chemical active substance applicators, or thermally acting probes for healing or micro-surgical purposes, structures for the performance of chemical analysis and processes within the Implanted structure and using a partially neural, adaptive control system in communication with the encoder, quasi-autonomous. diverse sensory-motor actions can be executed, for example, rapid response effects on the local tissue 30 can be executed in response to the sensory data just detected. In thia case, the bl-directiona coupling with the encoder can be used as control or checking.

An exemplary form of the adaptive spinal Implant, for example for use In paraplegics for neural modulation of the urinary tract, for guidance of the grip or walking movements, or for the reduction of phantom pain following amputation le where the implanted microstructure is

<img file="AU747686B2_D0013.tif" />

TV'S

Zfr£O89T9rTZO

I xej.s6ue6snv 393Ί TirilrT 66-6πν-6ΐ

08/04/99 WED 09:07 FAZ 1 609 924 1811 wooapridge&amp;Assocociates ®044 .u _thc, nerlnheral nervous system, or in muscle groups, whereby the situated In the spinal cord, the peripheral nervous y . is earned „_lOTS1 mtapttv. =0=0«. tn with th.

signal recap . head or eye movements, An exemplary form of through manipulation or. for example, through head y _f the Implanted structure i. where quwletatonomou. sensory-motor actions such , example, a need-drlven administration of growth hormone or thermal effects ere perto .

An exemplary tom of the adaptive cranial Implant, for example, as need-triggered local admlnlshatlonofecUve substance for epileptic patients, parkinsonism p^errtt^orpsychiatnc patients, is where the Implanted structure — Including local ee or _Miwe neurophysiological and ionic, molecular, and active auMtanoe concentrations.

v, ♦ a daoots with th© possibility of simple external replenishment, local active su s substance depots wHh th© poeemiiny or p _intracranielly and dosers with control and adaptive Information processing module - .s srtuate pawns kTperwptlon-based dialogue can he optlmixod end monkored. An of the implanted structure is where the quaswutonomous senaory-motor action .

examplePneed-driven administration of growth hormone or locetlydefioiantneuroUanBm Itler for synaptic functions; In other woids. the Otologies, contact, between n«w 0 fundamentally involved In learning and adaptation and, for example, th. ne suppression of epileptic events or thermal effects are effected. .

<img file="AU747686B2_D0014.tif" />

ZV£O89T9TTZO

I XEj.s6up6snv 3331 TV-t^T 66-6ny-6T

P:\OPER\RAB\clicnts\misc · dbwtintelligeni implants I SPA spcc.doc· 19/03/02

Many modifications will be apparent to those skilled in the art without departing from the scope of the present invention as herein described with reference to the accompanying drawings.

- 11A5 Throughout this specification and claims which follow, unless the context requires otherwise, the word comprise, and variations such as comprises and comprising, will be understood to imply the inclusion of a stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integers or steps.

The reference to any prior art in this specification is not, and should not be taken as, an acknowledgment or any form of suggestion that that prior art forms part of the common general knowledge in Australia.