Abstract
The purpose of this study was to evaluate
the feasibility to develop a new hybrid orthosis
called "cyberthosis" using selectively
closed loop electrical muscle stimulation. The knee joint was taken as the
basis for the study and a knee orthosis including
motor and sensors was specially designed to perform the closed loop functional
electrical stimulation control. A four-channel electrical stimulator was used
as well as a feedforward controller including a pro-portional, integral, derivative algorithm, both developed
at the Ecole Polytechnique Fédérale de
Spinal Cord Injury (SCI) is aggravated by complications: muscular atrophy, articular ankylosis, bedsores, spasticity, poor circulation in the affected limbs, loss of cardiovascular fitness, urinary and intestinal problems or osteoporosis [1][2][3]. To prevent these complications, regular mobilization of the paralyzed limbs from the outset of the lesion has been advocated to ensure the proper working of the different metabolisms and body functions. The aim of this project is to develop a hybrid orthosis for muscular and lower limb movement training activated by electrical muscle stimulation [4]. This equipment must ensure the kinematics and the dynamic similitude to natural movements such as cycling, standing up or walking. It should allow more efficient prevention of complications and improve the possibility of recovering voluntary walking.
In order to get the best kinetic and dynamic similarity of movement for re-education including functional electrical stimulation (FES), a combination of an orthosis made up of motors and sensors, and a closed loop electrical muscle stimulation (CLEMS) real-time control was suggested by the Swiss Foundation for Cybertheses (FSC : Fondation Suisse pour les Cyberthèses). The concept of "Cyberthosis", contraction of the words cybernetic and orthosis, was introduced to name this complex hybrid orthosis.
This paper
describes a feasibility study made on a single joint. A knee orthosis (KO) with one motor and several sensors, combined
with a CLEMS system to stimulate and control
selectively and simultaneously several muscles, was specially designed for this
study. The results of simple leg extensions with SCI
individuals should give the first features needed to develop complete lower
limb cyberthosis with three degrees of freedom.
2 Methods
2.1 Subjects
After approval of the protocol by the ethics committee, the closed loop electrical muscles stimulation knee orthosis (CLEMS-KO) prototype was tested on six healthy subjects. SCI patients were screened during a two-month period (February and March 2003).
Seven patients with one or more of the following were excluded: fixed contractures or contraindication in mobilization of the knee, spasticity ³ 2 according to the modified Ashworth scale upon flexion or extension of the knee [5], autonomic dysreflexia [6], known intolerance to electrical stimulation, pressure sore under treatment or seating time < 2 hours, normal voluntary muscle power of the quadriceps according to ASIA [7].
Eight eligible spinal cord injured patients (6 males and 2 females, 7 inpatients and one outpatient), 34 to 56 years old (mean 45.25), gave their written consent. Subjects were classified according to the NISCI-92 nomenclature [7]. All had incomplete spinal cord lesions. Patellar tendon reflex was scored following Meythaler et al. [8]. The characteristics of the patients are resumed in Table 1.
Table
1 : Subjects. IN. : Initials, NL : Neurological
Level, IS : Impairment Scale, TSI : Time Since Injury [month], QP : Quadriceps
Power, PR : Patellar Reflex, DO : Drop-out, R : right, L : left.
2.2 Electrical Stimulator
We used a programmable four-channel electrical stimulator (Figure 1) specially developed at the Ecole Polytechnique Fédérale de Lausanne (EPFL), with which the current waveform as well as real-time amplitude modulation, pulse width and frequency can be selected. After trials on healthy individuals, rectangular, biphasic pulses with an amplitude within 0 to 100mA, pulse width of 300µs and frequency of 50Hz were chosen. Three pairs of self-adhesive electrodes were used to stimulate separately rectus femoris, vastus lateralis and vastus medialis [9].
|
|
Figure 1: |
2.3 Knee Orthosis
The KO consists of an electrical DC motor, an angular position sensor and a force sensor (Figure 2). This device allows isokinetic and isotonic extension with CLEMS. The motor provides the force resistance, i.e. a joint torque up to 50Nm, and a speed regulation up to 110°/s during exercise.
2.4 Control
The force sensor measures the strength given by the muscles. If the strength corresponds to the torque command value for the exercise, the motor allows the movement with isokinetic speed. If not, the pulse amplitude is modulated by the PC controller to achieve a just-sufficient muscular contraction.
The torque due to body weight is identified before exercise. At the same time, the three muscles are identified in terms of strength-stimulation and strength-angle characteristics. These parameters are used by the feedforward controller including a proportional, integral, derivative (PID) algorithm.
|
|
Figure
2: |
3 Results
3.1
The graph in Figure 3 shows the typical results we obtained. The measured
torque follows the torque command inducing the isokinetic movement. All trial
values were 10 °/s, within 1-10 Nm, extension from 100° to 40°, with 0°
corresponding to the straight leg.
Figure 3:
In some cases, spasms interfered with our control. When this occurred the spasm's torque peak was detected. The controller was then able to stop the process and/or to release the stress in time.
3.2
Clinical
observation
All patients had at least one session. There were
three drop-outs during the study: VM had an associated plexus lesion and did
not respond to
The sessions lasted 20-90 minutes. The duration was closely related
to individual performance in transfer from the wheelchair to comfortable
installation on the module. None of the seven patients complained of pain
during the test period nor developed pelvic or leg pressure sores, fractures,
or had increased spasticity. As often with
|
|
Figure 4: |
4 Discussion and Conclusions
The results with the CLEMS-KO were encouraging from the point of view of safety and tolerance. This study has shown the ability to perform knee extensions by CLEMS applied to three muscles of the quadriceps selectively.
Following these good results, a first "MotionMaker™" prototype was developed and tested on healthy individuals. By mid-March 2004, it was ready for bilateral leg passive movement and closed loop muscle stimulation as described earlier. A 48 year-old male patient presenting complete T11 paraplegia with impairment scale A was invited to participate in the clinical trial.
4.1 Current and future work
The MotionMaker™ is a stationary programmable test and training system (Figure 5 and Figure 6). It is composed of two orthosis with three degrees of freedom including motors and sensors, a multi-channel electrostimulator and a control unit managing the electrical stimulation with real-time regulation. This allows SCI patients to have full leg movements with the desired characteristics of position, speed and torque.
Figure
5: MotionMaker™
To stimulate two legs in a complete movement, we need 14
The MotionMaker™ allows the exercise of movements like press-leg, cycling (with adjustable crank length and position), sit to stand, walking, and others with a whole kinematic chain. Each muscle can be identified in terms of strength-stimulation and strength-angle in order to use it optimally and reduce fatigue.
Figure
6: MotionMaker™ Prototype with healthy individual
References
[1]
Le. Survival from spinal cord
injury
J Chronic Dis. 35(6):487-92, 1982.
[2]
Garland. Osteoporosis
after spinal cord injury. J.Orthop Res. 10: 371-378. 1992.
[3]
Phillips. Effect of spinal cord injury on the heart
and cardiovascular fitness. Curr Probl Cardiol. 23:641–716. 1998.
[4]
Popovic. Control of movement for the physically disabled. Springer. 277-281,
2000.
[5]
Bohannon. Interrater reliability of a modified Ashworth scale of muscle spasticity. Physical Therapy. 67 (2), 206-207, 1987.
[6]
Gutmann. Spinal cord injuries.
Comprehensive management and research. Blackwell,
[7]
[8]
Meythaler. Am J Phys Med Rehabil. 71:
321-327, 1992.
[9]
Vanderthommen. Contribution à l'étude fondamentale et clinique de l'électrostimulation neuromusculaire.
Université de Liège, mémoire de doctorat. 1993-94.
Acknowledgements
We would like to express our sincere thanks to the SCI patients who volunteered to take part in the study. We are grateful to the FSC, the EPFL and to the Swiss Lottery (Loterie Romande) for their financial support.