Abstract
Following a
spinal cord-injury, paraplegics suffer from bladder dysfunctions due to partial
or complete rupture of sensitive and motor innervations. Stimulation of the
sacral nerve root seems to be one of the most promising techniques for bladder
rehabilitation as it does not require any neurotomy or rhizotomy. A new
implantable neurostimulator system is reported. The implant prototype offers
Selective Stimulation to recover voluntary control of the micturition reflex
and Permanent Stimulation to reduce, or eliminate the undesirable hyperreflexia
of the detrusor. This paper gives an overview of the neurostimulator design
evolution through a review of the main architectures that led to this new
system.
1. INTRODUCTION
The spinal cord
is the unique neural pathway between the brain and all physiological and
anatomical systems. A normal individual feels the need to urinate as soon as
the bladder is full. The brain commands a voluntary micturition by contracting the
bladder muscle, the detrusor, and relaxing the sphincter [1, 2]. In addition,
the brain inhibits autonomously any reflex contractions of the detrusor that
may occur during the filling process. Following a spinal cord-injury,
paraplegics suffer from bladder dysfunctions due to partial or complete rupture
of sensitive and motor innervations [3]. Many attempts have been made to
recover voluntary control of the micturition reflex by means of electrical
stimulation at different sites of the urinary system [4]. Selective Stimulation
of the sacral nerve root seems to be one of the most promising techniques to induce
voiding without neurotomy or rhizotomy [5].
This paper
reports a new urinary implant prototype named Mixture Neurostimulator (MNS) for
the urinary system rehabilitation by means of electrical stimulation of sacral
nerves. It has been designed by the Urostim team within the Polystim lab in
partnership with Victhom Human Bionics. In order to undertake a new phase of
animal experiments, previous designs have been reviewed, modified and improved
taking into account all past experiments.
2. SYSTEM DESCRIPTION
The electronic
device, as shown in Fig.1, consists of a small implantable neurostimulator and
an external controller to communicate instructions and stimulation parameters
to the implant. In addition, the controller sends energy via a wireless
inductive link to provide enough power for high current stimulations.
Fig. 1 - Stimulation system block diagram
The
neurostimulator combines two types of functions. The first one, called
Selective Stimulation and illustrated in Fig.2, aims for voluntary voiding. It
is a bi-frequency, high amplitude stimulation that is launched on a time
limited basis by the external controller.
This stimulation uses the inductive power and will stop as soon as the
controller is removed.
|
|
Low frequency |
High frequency |
||||
|
Parameter |
LFA |
LFP |
LFW |
HFA |
HFP |
HFW |
|
Unit |
mA |
Hz |
ms |
mA |
Hz |
ms |
|
Range |
0 2000 |
20 2000 |
4 200 |
0 2000 |
300 2000 |
4 200 |
Fig. 2 - Selective Stimulation waveform
The second one, called
Permanent Stimulation and illustrated in Fig.3, aims for hyperreflexia
suppression. It is a low frequency, low amplitude pulse train that, once
launched, runs on a continuous time basis and in an autonomous mode. This
stimulation uses a long life embedded battery and may be stopped at anytime by
the controller.
|
Parameter |
Amp |
Freq |
PW |
Ton |
Toff |
|
Unit |
mA |
Hz |
ms |
sec |
sec |
|
Range |
0 1000 |
5 100 |
10 300 |
1 60 |
1 60 |
Fig. 3 - Permanent Stimulation waveform
The Selective and Permanent
Neurostimulator (SPN) was patented by the Polystim neurotechnologies laboratory
in 2002 [6,8]. Three years of chronic animal experiments demonstrated the
efficiency of micturition by means of selective stimulation [6,7]. The MNS is
an advanced version of the SPN and offers the choice of running the Permanent Stimulation
either from the embedded battery or from the inductive power. The external
controller has been improved as well to provide ease of handling and complete
flexibility over the stimulation parameters.
3. ARCHITECTURE REVIEW
This section
gives an overview of the neurostimulator design evolution through a review of
the main architectures [9]
that
led to the SPN and MNS systems.
The first
architecture (Fig.4) uses an FPGA (Field Programmable Gate Array) as the unique
controller for both Selective and Permanent Stimulations. With a multiplexer,
the FPGA and the stimuli generator are powered either from the inductive power
or the embedded battery power. This is required for the Permanent Stimulation
that must run autonomously when the inductive power is removed. It is, however,
difficult to guaranty a continuous transition of power, therefore, the risk of
losing power and memorised data makes this solution inconvenient.
Fig. 4 - Architecture I
The second
architecture (Fig.5) solves this power discontinuity risk by powering the FPGA
permanently by the embedded battery. Power multiplexing is applied to the
stimuli generator only; that way Permanent Stimulation could run on the battery
power. Unfortunately, the energetic performance of the (less recent technology)
FPGA was not suitable for this neurostimulator since low power consumption is
of crucial importance for the battery life.
Fig. 5 - Architecture II
In the third
architecture (Fig.6), the FPGA is replaced by a microcontroller (PIC) that has
the advantage of offering low power consumption modes as well as a non-volatile
memory. However, to achieve the same performance as the FPGA in selective
stimulation, the clock frequency of the PIC has to be increased. Thus, the low
power advantage will not be as expected. In addition, decoding data and
extracting the clock from the signal emitted by the external controller is
complex to implement with the PIC.
Fig. 6 - Architecture III
The final
architecture of the SPN (Fig.7) combines both types of controllers to benefit
from the advantages of each one. The FPGA is used for the Selective Stimulation
with the inductive power and the PIC is used for the Permanent Stimulation with
the battery power. Control signals and power multiplexing is applied to the
stimuli generator as it is shared by both controllers.
Fig. 7 - Architecture IV
One of the main concerns
discovered throughout chronic animal experiments has been the complete
interruption of all neurostimulator functions when the battery runs down. It
was desirable that the selective stimulation stays functional. This is due to
the fact that the PIC has the control of the power multiplexer of the stimuli
generator. Once the battery runs down, the stimuli generator is no longer powered.
Moreover,
previous animal experiments proved that the impedance of the
cuff-electrode/nerve interface may be as high as 2KΩ. Hence, the stimuli
generator, and specifically the current source, needs a voltage supply of 5V at
least to provide a selective stimulation current as high as 2mA.
Fig. 8 - Architecture V
The chosen
approach (Fig.8) to solve these issues is to use separate stimuli generators
for each controller. This way, the FPGA stimuli generator may be powered at 5V
(or more depending on the available inductive power) whereas the PIC stimuli
generator will stay battery powered at 3.3V. Selection of one of them is
achieved using Single Pole Double Throw (SPDT) Reed Switches mounted as a
multiplexer. The Reed Switches are activated externally using a magnet. The MNS
is based on this new architecture and made with commercially available
electronic devices on a printed circuit board (PCB). Even though it requires
more components, the small scale factor of recent technologies has made it
possible.
5. CONCLUSION
A new implantable
neurostimulator prototype has been elaborated with an architecture review of
previous designs. At the time it was written, several prototypes were being
assembled and tested before starting in-vivo animal experiments.
References
[1] Blaivas JG. The neurophysiology of micturition; A clinical study of 550 patients. J Urol 1982;127:958–63.
[2] Bradley W.E., Rockswold G.L., Timm G.W., Scott
F.B.: Neurology of micturition. J Urol 1976, 115, 481-486.
[3] Fam BA, Rossier AB and Blunt K, Experience in the urologic management in 120 early spinal cord injury patients. J Urol 1978, 119(4):485-7.
[4] Rijkhoff, N.J.M., Wikstra, H., Van Kerrebroeck, P.E.V. and Debruyne, F.M.J., Urinary bladder control by electrical stimulation: review of electrical stimulation techniques in spinal cord injury, Neurourol. Urodynam, 1997, 16:39-53.
[5] Robin,
S., Sawan, M., Abdel-Gawad, M., Abdel-Baky, and Elhilali, M.M, Implantable
stimulation system dedicated for neural selective stimulation, Med. and Bio.
[6] Schneider, E., Abdel-Karim, A.M., SAWAN, M. and
Elhilali, M.M. New stimulation strategy to improve the
bladder function in paraplegics: Chronic experiments in dogs. 23rd IEEE Int.
Conf. of
[7] Ba, A., Schneider, E., Abdel-Karim, A.M.,
SAWAN, M. and Elhilali, M.M. Implantable dual stimulator to recuperate the
bladder functions: Chronic experiments in dogs, IFESS,
[8] Sawan,
M. and Elhilali, M.M., “Electronic Stimulator Implant for Modulating and
Synchronizing Bladder and Sphincter Function”,
[9] Schneider, E. "Conception et évaluation
d'un système de stimulation électrique neurale dédié à la réhabilitation des
fonctions vésicales." 2001 MScA Thesis may be
obtained from École
Acknowledgements
The authors would like to acknowledge the financial support from the Canadian Institutes for Health Research (CIHR) and that of Victhom Human Bionics.