FES-Cycling - Measurements and Individual Adaptation
of Stimulation Patterns on a Test Bed and a
M. Gföhler1, T. Angeli1, P. Lugner2,
C. Hofer4, M. Reichel3,
W. Mayr3, M. Bijak3
1Institute of Machine
Elements, Vienna University of Technology, Vienna
2Department of Mechanics,
Vienna University of Technology, Vienna
3Department of Biomedical
Engineering and Physics, Vienna University Medical School, Vienna
4Department of Physical
Medicine, Wilhelminenspital, Vienna
SUMMARY
In
this study static and dynamic measurements with paraplegic test persons on a
freely adjustable test bed were performed to determine the individual
FES-cycling performance and to individually optimize the stimulation patterns.
Hydrogel surface electrodes were used to activate quadriceps, hamstrings, and
gluteus maximus and to elicit the peronaeus reflex. The optimized stimulation
patterns were then applied to cycling on a moving tricycle.
STATE OF THE ART
An
important motivation in studying
Angeli
/1/ has shown that the drive power output of neurologically intact subjects can
be raised by using an optimized pedal path, which is realized by a 4-bar
linkage pedal drive, instead of the common circular pedal path. In a simulation
study this pedal path was also optimized for
MATERIAL AND METHODS
The
measurements with paraplegic test persons were carried out on a freely
adjustable test bed (Fig. 1) /3/. The paraplegic test person is seated on a
specially adapted wheelchair. It was decided to use only one leg for the
measurements to make sure that the results are not influenced by spasms or
other forces generated by the other leg. The inclinations of seat and backrest
are adjustable in 10° steps. The chair is horizontally moveable along two guide
rails, and its position can be fixed in 2.5 cm steps by an alignment pin. For
getting onto the test bed the chair is moved to the horizontal end position E
to provide enough space between chair and crank bearing. The crank bearing is
mounted on an electrical vertical lift. The crank axis is pivoted at the crank
bearing, and the force measuring crank is mounted on the measuring side. A
chain connects the chain wheel to the torque measuring shaft which is coupled
with a gear motor by a torsional stiff coupling. Because the vertical movement
of the crank axis the distance between crank axis and torque measuring shaft is
changed, but the length of the connecting chain is constant. Therefore the
measuring shaft and motor are mounted on a pivoting lever which changes its
inclination according to the vertical position of the crank axis. The lift with
crank axis, along with the lever-carrying motor and measuring shaft, are
mounted on a ground plate. This plate is, like the seat, moveable along two
horizontal guide rails, and its horizontal position can be fixed by an
alignment pin in 5 cm steps. The pedal is either mounted directly to the crank
in PA, as usual moving on a circular path, or to the coupler of a 4-bar linkage
in P. This linkage consists of wing, coupler and crank and makes the pedal move
along a non-circular pedal path. The wing bearing W is mounted on the same
vertical lift as the crank axis C, the bars are connected by pin joints in B
and PA. The foot is fixed to the pedal by Velcro fastenings, which means that
both tensile and compressive forces are transferred. For paraplegic subjects it
is necessary to fix the ankle joint rigidly because they are not able to
stabilize the joint. This is done by an orthosis which also supports
side-to-side stability of the leg. The right foot is placed on the ground plate
of the test bed in a resting position.
|
Figure 1: Schematic of the test bed. K
and H indicate knee and hip. The arrows show how the elements of the test bed
can be adjusted. |
|
|
RESULTS
a) b)
Figure
2: Results of isometric force
measurements at 20 points along the optimized pedal path, stimulation of
quadriceps of test person ZA. (a) Active loads (directly caused by active
muscle forces) at point D of the crank (passive loads have been eliminated).
(b) Resulting pedal force vectors in the parasagittal pedaling plane and region
where positive drive torque is applied.
For
the static measurements 20 equiangular points were defined along the pedal
path. In every point the muscle was stimulated for 0.75 sec, the pedal forces
were measured by a force measuring crank. Orientation and magnitude of the
active forces (directly caused by active muscle forces) applied to the crank in
the parasagittal pedaling plane were calculated for a number of geometrical
positions of the rider and variations of the stimulation parameters. Out of
this data it could be seen in which part of the pedal path positive drive
torque was applied. Figure 2 shows results of static measurements.
In
the dynamic measurements (Fig. 3) the muscle was stimulated in its concentric
range, starting with an estimated interval derived out of the static
measurements. Then the start and end points of the stimulation interval were
varied to find out in which interval maximum positive drive torque could be
applied. Tests with all muscles stimulated together were performed to show how
much the muscles influenced each other (Fig. 3d).
a)
b)
c)
d)
Figure
3: Results of dynamic measurements with test
person ZA during one full rotation at 25 rpm: (a) - (c) show the resulting
active drive torque from stimulation of (a) quadriceps, (b) gluteus maximus and
(c) hamstrings (with optimized stimulation interval (stimulation voltage Ustim)).
(d) comparison between the summation of the results for the single muscles and
the measured results of the stimulation with all muscles during one full
rotation of the crank
Finally a set of optimized parameters for stimulating all muscles together
while pedaling was derived for each individual test person and tested on a
specially developed mobile tricycle for paraplegics. Figure 4 shows a
paraplegic test person on the moving tricycle /4/ and results of measurements
while cycling average crank angular velocity 33 rpm.
Figure 4: Paraplegic test person on the moving tricycle and
results of measurements during steady-state cycling at crank angular velocity
33rpm. The results are compared to results of a simulation study /5/.
DISCUSSION
Leg
muscles are activated by surface stimulation and the measurements show which
forces and torques are applied to the crank. Thus, stimulation patterns and
geometrical position of the rider may be optimized for the development of an
optimized cycling movement, influenced by individual parameters, for paraplegic
subjects. It has been shown that a moving tricycle can be powered by
REFERENCES
/1/
Angeli T. (1996): ‘Propulsion units of bicycles - optimization of capacity’, PhD thesis, Vienna University of
Technology,
/2/
Angeli T., Gföhler M., Eberharter T., Lugner P., Rinder L., and Kern H. (2001): ‘Optimization of the pedal
path for cycling powered by lower extremity muscles activated by Functional Electrical
Stimulation’, in Middleton J., Jones M.L., Shrive N.G., Pande G.N. (Ed):
‘Computer Methods in Biomechanics and Biomedical Engineering-3’, pp.263-268
(Gordon and Breach Science Publishers)
/3/
Gföhler M., Angeli T., Eberharter T., Lugner P., Mayr W., and Hofer C. (2001):
‘Test bed with force measuring crank for static and dynamic investigations on
cycling by means of functional electrical stimulation’, IEEE Trans. neural Systems and Rehabilitation Engineering TRE, 9,
pp. 169-180.
/4/
Angeli T., Gföhler M., Eberharter T. and Rinder L. (1999): ‘Tricycle for
paraplegics using functional electrostimulation’, Med. & Biol. Eng. & Comp., 37, Supp. 2, pp. 326-327.
/5/
Gföhler M., Angeli T. and Lugner P., (2001):
‘Optimal control of cycling by means of functional electrical
stimulation - a dynamic simulation study’, VIIIth International Symposion on
Computer Simulation in Biomechanics,
ACKNOWLEDGEMENT
This
work was sponsored by the Austrian Science Foundation - FWF and Otto Bock
AUTHOR'S ADDRESS
Margit Gföhler Getreidemarkt
9/306
Institute of Machine Elements A-1060
Vienna
Vienna University of Technology margit.gfoehler@tuwien.ac.at