Artif Life Robotics (2000) 4:212-219

@ ISAROB 2000

Yoichi Shimada 9 Shigeru Ando 9 Satoaki Chida

Functional electrical stimulation

Received and accepted: April 5, 2001

A b s t r a c t Percutaneous intramuscular electrodes and a por-

table multichannel system were used to restore the function of paralyzed upper and lower extremities in spinal cord injuries and hemiplegic patients. The hybrid functional electrical stimulation (FES) using percutaneous intramuscular electrodes provides practical ambulation for paraplegics. FES was more effective than the flexor hinge splint in increasing the grasping power (GP) of tetraplegic patients, and a stronger and stable G P which was not affected by wrist position make FES practical for improving the activities of daily living. A tilt sensor, which was put on the thigh, could be used to measure a gait cycle. This study suggests that putting a tilt sensor on a hemiplegic patient's thigh improves the FES procedure and may help to restore gait in these patients. We conclude that percutaneous intramuscular FES is useful for spinal cord injuries and hemiplegic patients. Key words FES - Percutaneous intramuscular electrode

Introduction
Restoring independence in performing daily functions is the main goal in treating paralytic patients. Recent advances in computer technology have made it possible to control paralyzed muscles by electrical stimulation. We have used functional electrical stimulation (FES) to restore the paralyzed muscles in the upper and lower extremities since 1990.14 There are a lot of potential benefits to paraplegic patients if they can achieve FES standing. Standing may help to prevent joint contracture by eliminating the chronic

sitting posture, and it may reduce osteoporosis because it increases weight-bearing. The upright posture may improve the position of the internal organs, and aid bowel and bladder function. It may also reduce the chance of pressure sores, and it could aid circulation. Increased functional abilities while standing may also enhance personal self-esteem. FES enables patients with a severe spinal cord injury to reconstruct grasp movements such as the palmar and lateral grasps of the upper extremities, and therefore it is useful for improving the activities of daily living (ADL). However, there are few reports on how grasping power (GP) can be measured by FES, as well as on the relation between G P and the practical uses of FES. FES has been used to improve the gait of stroke patients. The most c o m m o n method of control uses foot switches, as originally proposed by Liberson e t al., 4 but this requires wires or telemetry to connect the switch to the stimulator. With FES, we have been preventing a foot-drop gait, which is caused by stroke, by using footswitches under the heel, as in the Akita heel sensor system (AHSS) invented in 1996. s A H S S is useful for foot-drop-gait patients. Some potential improvements include the ease of installing the system on the patient's heels, durability, and appearance. In 1996, Dai et al. 6 reported that with FES the tilt sensors need to be put on the calf in order to prevent a foot-drop gait. A miniature foot-drop stimulator was designed with a magnetoresistive tilt sensor built in; no external sensor cables were required. Theory has it that if the tilt sensor is put on the thigh, it will still be possible to detect steps in the gait. Patients can also put the sensors on more easily. Here we descibe the clinical applications of FES in paraplegia, and evaluate the relation between G P and the practical uses of FES for a tetraplegic patient. We also evaluate the detection of a normal and a foot-drop-gait cycle using a tilt sensor placed on the anterior portion of the thigh.

Y. Shimada ([]) - S. Ando. S. Chida Rehabilitation Division, Akita University Hospital, 1-1-1 Hondo, Akita 010-8543, Japan Tel. +81-18-884-6147;Fax +81-18-884-6373 e-mail: yshimada@med.akita-u.ac.jp

213
Table 1, Subjects

Case 1 2 3 4 5 6 7 Average

Age 19 25 24 33 49 33 19 29

Sex Male Male Male Male Male Male Male

Disease SCI SCI SCI SCI SCI SCI SCI

Paresis Complete Complete Complete Complete Incomplete Incomplete Incomplete

Level of injury T6 T5 T8 T6 T10 T12 T12

Time since injury 3m 1 y 10 m l y 4m 7m ly 2y 1m 4m Iy Im

Follow-up 5y 10m 6 y 11 m 4y 8m 8m 2y 9m 3y 4m 8m 3y 7m

Number of electrodes 30 34 34 30 18 22 10 25

Hybrid orthosis AFO AFO AFO, AKJ, Walkabout Walkabout Pure RGO, Walkabout Walkabout

SCI, spinal cord injury; AFO, ankle-foot orthosis; AKJ, Akita knee joint; RGO, reciprocal gait orthosis m, months; y, years

Fig. 2. Percutaneous intramuscular electrode

Fig. 1. Portable stimulator

Materials and methods

FES for paraplegia The subjects were four patients with complete paraplegia and three patients with incomplete paraparesis. None of the patients could stand and walk by themselves, and they have all used a wheelchair in daily living. The average age of the patients was 29 years (range 19-49 years). The subjects suffered from spinal cord injury at the level of T5-12. The average time since injury was 1 year 1 m o n t h (range 3 months to 2 years and 1 month), and the average follow-up time was 3 years and 7 months (range 8 months to 6 years 11 months) (Table 1). The FES system developed by the Nippon Electric C o m p a n y was used in this study. 7-9 This system consists of a stimulation data creating system (SDC) and a 30-channel stimulator. The SDC was used to compose and store the stimulation parameters that set the threshold voltages for each muscle, and controlled the pulse shape and individual

pulse sequences. The portable stimulator measured 8.9cm 14.5 cm 3.1 cm, weighed 330 g, and was usually worn on a belt (Fig. 1). The indwelling electrode 8 was made of helically w o u n d Teflon-coated 19-strand stainless steel (Nippon Seisen Co.) (Fig. 2). The tip of the electrode was deinsulated. The electrodes were implanted percutaneously into the motor point of the muscles. The pulse amplitude was m o d u l a t e d from 0 to - 1 5 V . Rectangular pulse trains were used with a 200-~ pulse width and a pulse interval of 50ms. A customized p a t t e r n of stimulation was composed for each subject based on observed muscle function. Prior to implantation, the sensitivity and location of the motor point were determined using surface-electrode stimulation. I m p l a n t a t i o n of the electrodes was performed u n d e r sterile conditions (Fig. 3). W e chose a c o m m o n bodyentry point for the electrodes, which was at both anterior proximal thighs (Fig. 4). A hybrid orthosis, an a n k l e - f o o t orthosis ( A F O ) , an Akita knee joint (AKJ), 1~ a Walkabout, and a reciprocal gait orthosis ( R G O ) 11 were provided for stabilization in six patients. Electrical exercise commenced 2 weeks after implantation to allow time for the tissues to heal. This reduced the

214 Table 2. MMT before reconstruction R Shoulder Flexion Abduction Elbow Flexion Extension Forearm Pronation Supination Wrist Flexion Extension Finger Flexion Extension MMT, muscle manual testing her GP, 31 normal female students, aged 19-24 years (X = 21), were used as controls. The patient was injured in a traffic accident in June 1995. She was admitted to our hospital for the purpose of reconstructing upper-extremity function by FES in N o v e m b e r 1996. Before reconstruction, M M T of her upper extremities showed only T in elbow extension, and F in forearm pronation. Wrist extension was relatively better at G. Wrist flexion and finger flexion and extension were both Z (Table 2). There was little joint contracture and spasticity. Because her palmar grasp was of no practical use because of the effect of dynamic tenodesis, devices were necessary for eating, teeth-brushing, grooming, and writing. Percutaneous intramuscular electrodes were implanted on December 6, 1997, 18 months after the injury. In order to reconstruct finger extension motion, the implantation regions were the deep branch of the radial nerve, the extensor digitorum, the extensor indicis, the extensor pollicis longus, the abductor pollicis brevis, and the opponens pollicis. In order to reconstruct finger flexion motion, the implantation regions were the flexor digitorum superficialis, the flexor digitorum profundus, the flexor pollicis longus, the adductor pollicis, and the first dorsal interosseous. Therapeutic electrical stimulation was started 13 days after the implantation operation. Palmar and lateral grasps were reconstructed in the right upper extremity by means of a portable multichannel FES system ( F E S M A T E CE1230, N E C Medical Systems, Tokyo, Japan) 8 weeks after the operation (Fig. 5). Following FES reconstruction, the patient was able to perform many actions using the palmar and l~teral grasps. FES enabled her to drink canned juice, eat with a spoon, write with a pen, brush her hair, and turn a tap on and off without using devices. The G P of palmar grasps were measured in the right upper extremities of both the control group and the patient using A J A M A R hydraulic hand dynamometer (PC5030J1, Preston, USA); lateral grasp strength was measured using a hand finger dynamometer set (SPR-6500, S A K A I Medical Co., Japan). In the patient, palmar grasps were performed by the dynamic tenodesis effect, the flexor hinge G+ G+ N T F NZ G Z Z L G+ G+ N T F NZ G Z Z

Fig. 3. Insertion of percutaneous intramuscular electrode

Fig. 4. Entry points of electrodes at anterior thighs

problem associated with electrode movement. Stimulation of the muscles was conducted for 5 min, three times daily at the beginning. After 5 weeks, the stimulation was applied for 30min, five times daily. After exercising like this for more than 3 months, standing and walking training started. FES for tetraplegia A 20-year-old female patient with C6 complete tetraplegia (Frenkel A, Zancolli 2B-2) was evaluated. In order to assess

215 Fig. 5. A Prehension of drink can with a palmar grasp, and B prehension of a spoon with a lateral grasp

Hg. 6. Measurement position of A palmar grasp, and B lateral grasp

splint, and with FES. Lateral grasp was performed with FES only. Both the control group and our patient were made to sit right back in a wheelchair for the measurements. In addition, the right upper extremity was hung down beside the hand rim naturally, with the elbow joint extended, to measure palmar grasps (Fig. 6A). One-third of the distal forearm was placed lightly on the desk to measure the lateral grasp (Fig. 6B). The grip size of the hand dynamometer was 2 inches. For the control group, three G P measurements were taken for each person, and the mean of the largest value in each set of three was calculated. In the patient, GP values were measured in the same manner three times during 1 week, and a mean value was calculated. FES for hemiplegia The control subjects were ten healthy males, whose average age was 26 (23-29) years. One stroke patient (56 years old) was also measured.

The tilt sensor used in this study (Model UA-1) (Fig. 7) was from Midori American Corporation. It was put on the anterior middle portion of the thigh (Fig. 8). We measured the tilt of the thigh during gait using the tilt sensor and an automatic coordination system simultaneously. Forcesensing resistors were placed under the heel and toe to provide reference points for the step cycle. Using a percentage of the gait cycle, we normalized the angular change of the thigh during one complete motion. We also paid attention to the point of inflection in the angular change of the thigh during a gait cycle. The point is close to toe-off, and it activates the switch to come on and stimulate in the swing phase (Fig. 9). All points of inflection in the angular change of the thigh were compared while the subjects walked normally, with a gait speed of 0.5m/s; and a step length of 50cm. The subjects were measured in ten trials.

216

Results
FES for paraplegia Although all patients could stand and walk with a hybrid FES using an A F O (Fig. 10), the clinical results for the m a x i m u m duration of continuous standing and the maxim u m distance of continuous walking were shown with the hybrid orthosis being used in daily life. The m a x i m u m duration of standing with electrical stimulation only was 60rain in case 5, with an A F O it was 25rain in cases 1 and 3, with a Walkabout it was 150min in case 6, with a R G O it was 60min in case 6, and with an A K J it was 30min in case 3. The m a x i m u m distance walked with electrical stimu-

lation only was 120m in case 5, with an A F O it was 3 0 m in cases 1 and 3, with a Walkabout it was 300m in case 6, with a R G O it was 7 0 m in case 6, and with an A K J it was 40m in case 3. The maximal gait speed was 0.5m/s with FES only, 0.1m/s with an A F O , 0.33m/s with a Walka-bout, 0.22m/s with a R G O , and 0.1 m/s with an A K J (Table 3). Only one electrode broke (0.6%). This was in case 1 at a site deep in the iliopsoas muscle. The loss of an adequate contraction force due to m o v e m e n t of the electrodes required ten electrodes to be reimplanted (5.6%). On ten occasions (5.6%) there was a superficial infection around

standing p h a s e 50 40 ~ \ 30 20 10 0 -10 ,\tilt sensor

swing phase

- - automatic coordination system/It////

% gait c y c l e point o f inflection

Fig. 7. Tilt sensor (UA-1)

Fig. 9. Point of inflection in the gait cycle

Fig. 8. Tilt sensor on the thigh

217

Fig. 11. Grasping power of palmar and lateral grasps in the control group, in the patient with a dynamic tenodesis effect, with a flexor hinge splint, and with FES. Light shading, palmar grasp; dark shading, lateral grasp

Table 4. Comparison of point of inflection in percent of gait cycle (mean _+ SD) Healthy males Fig. 10. Standing with complete paraplegia by functional electrical stimulation (FES) Tilt sensor UA-1 Automatic coordination system 60.1 _+ 6.5% 65.0 _+ 5.8% Stroke patient 94.7 _+ 0.4% 79.8 _+ 1.8%

Table 3. Results

Case
1

Hybrid orthosis AFO AFO AFO AKJ Walkabout Walkabout Pure RGO Walkabout Walkabout

Continuous standing (rain) 25 20 25 30 80 60 60 60 150 60

Continuous walking (m) 30 25 30 40 100 60 120 70 300 60

Gait speed (m/s) 0.1 0.1 0.1 0. l 0.25 0.2 0.5 0.22 0.33 0.2

1 . 1 + / - 0 . 2 k g , i.e., approximately 13% of that of the control group (Fig. 11). F E S for hemiplagia The correlations between the signal of the tilt sensor on the thigh and the foot-contact events were r e c o r d e d simultaneously. T h e change of angle using a tilt sensor was similar to that using an automatic coordination system. The point of inflection in healthy subjects was 60.1 --+ 6.5% of the gait cycle, and this point in a patient was 94.7 + 0.4% of the gait cycle in angular motion of the thigh using the tilt sensor. The point in healthy subjects using an automatic coordination system was 65.0 -+ 5.8% of the gait cycle, and the p o i n t in a patient was 79.8 1.8% of the gait cycle (Table 4). In n o r m a l gait, the slope of the relation between the timing of toe-off in a gait cycle and the p o i n t of inflection in a tilt sensor was 0.57 (P = 0.007). T h e r e was a correlation with the timing of toe-off in a gait cycle.

2 3 4 5 6 7

the site of electrode insertion. These infections were treated by topical sterilization of the skin and oral antibiotics. No electrodes had to be removed, and no visceral complications occurred in this series. F E S for tetraplegia The G P of p a l m a r and lateral grasps were 3 3 . 2 + / - 3 . 1 kg and 8 . 3 + / - 1 . 0 k g , respectively, in the control group. In the patient, grasping p o w e r was not m e a s u r a b l e because of the dynamic tenodesis effect. P a l m a r grasp strength with the flexor hinge splint was 2 . 2 + / - 0 . 3 k g , and with FES, it was 5 . 3 + / - 1 . 5 k g , i.e., a p p r o x i m a t e l y 16% of that of the control group, and 2.4 times greater than that with the flexor hinge splint. L a t e r a l grasp strength with F E S was

Discussion
F E S for p a r a p l e g i a T h e r e are t h r e e stimulation methods: surface electrodes, p e r c u t a n e o u s electrodes, and i m p l a n t e d electrodes. Surface electrodes require daily p l a c e m e n t and removal. In addition, some patients feel discomfort when excessive

218 stimulation is given by surface electrodes. It is difficult to stimulate the deep muscles, such as the iliopsoas muscle, by surface electrodes. Implanted electrodes need another operation when the electrodes break or move. In the past, percutaneous electrodes have failed at a high rate. Marsolais and Kobetic 12 reported that 35% failed within 4 months, and only 30% continued to function 1 year later, with 20% surviving for 2 years, giving an average failure rate of 2.5 electrodes per month that required replacement in a 48-electrode system. In our series, the rate of breakage and movement of electrodes was only 6.5% for an average 2.2 years. Handa et al., 8 using the same electrodes, reported that the rate of breakage was 1.3% in 457 electrodes. The helically coiled structure of the electrode has demonstrated resistance to breakage even when the electrode was implanted into bulky leg muscles. 7-9 These results suggested that the ultrafine intramuscular electrode was practical and available for long-term use in paraplegic patients. In restoring locomotion in paraplegics, it is necessary to control the hip, knee, and ankle joints, and to maintain muscle activity to sustain an upright posture during standing and walking. Muscle fatigue is a major problem for restoration by FES. Ichie .3 described several attempts to resolve this problem: (1) by controlling all the muscles which are related to locomotion and antigravity posture, TM (2) by introducing a withdrawal reflex for making locomotive motion, (3) by decreasing the loads on the antigravity muscles with an orthosis, and (4) by introducing a closed-loop control system with sensors. 14We believe that reducing the amount of stimulation of the muscles can be achieved using a hybrid orthosis and closed-loop control. The closed-loop control system requires joint position sensors that provide feedback information to the host computer to prevent falling, regulate the amount of stimulation, and determine the timing of stimulations. We have recently developed a new stimulator for a closed-loop control system with our percutaneous electrodes. 3 We have used several types of hybrid orthosis. The hybrid FES with an AFO was superior to the others in cosmesis, donning-doffing, standing up, and using a wheelchair. The Walkabout and R G O were superior in stability, reducing muscle fatigue, energy consumption, and maintenance of standing. The AKJ was intermediate in these hybrid FES systems. We conclude that hybrid FES using percutaneous intramuscular electrodes provides practical ambulation for paraplegics.

Fig. 12. High practical prehension: brushing the hair. Stable prehension was maintained even during wrist joint flexion

for the pinch grasp. FES reconstruction made it possible for our patient to perform many actions using palmar and lateral grasps in daily life. Brushing her hair and turning a tap on and off were the most practical skills that FES allowed her to do. Both of these tasks require a strong and stable GP not affected by wrist position. Although the grasping power in our patient is less than that of the normal group, it is better than the standard established by Peckham. As Fig. 12 shows, stable prehension was maintained even during wrist joint flexion. It is likely that such advantages make FES practical for ADL, and better than the flexor hinge splint. FES for hemiplegia The point of inflection measured by the tilt sensor was detected earlier in a gait cycle than the point of inflection measured by an automatic coordination system. However, it correlated with the timing of toe-off in the gait cycle, and was assumed to be available to signal the timing of the stimulation of the peroneal nerve which was synchronized with the swing phase of the gait. By timing the gait cycle in the stroke patient's gait, the point of inflection measured by the tilt sensor was detected later than the point of inflection measured by the automatic coordination system. The standard deviation in the point of inflection measured by the tilt sensor was as small as +0.4, and there was reproducibility of this point. We think that we can stimulate the peroneal nerve synchronized with the swing phase of the gait of stroke patients by using a setting which is specific to each stroke patient.

FES for tetraplegia Hatta et al. is explains that because finger muscle forces and motor hand skills show different aspects of upper extremity function, we should measure both force and skill when evaluating upper extremity functions. Accordingly, we should also measure the finger muscle strength of grasping and pinching when evaluating upper extremity function reconstructed by means of FES. Peckham et al. 16 has suggested that generally I kg is the minimal acceptable strength

219

References
1. Shimada Y, Sato K, Abe E, et al. (1996) Clinical experience of functional electrical stimulation in complete paraplegia. Spinal Cord 34:615-619 2. Shimada Y, Sato K, Kagaya H, et al. (1996) Clinical use of percutaneous intramuscular electrodes for functional electrical stimulation. Arch Phys Med Rehabil 77:1014 1018 3. ShimadaY, SatooK, Kagaya H, et al. (1995) Closed-loop controlfor functional electrical stimulation with percutaneous electrodes in paraplegics. In: Proceedings of the 5th Vienna International Workshop on FES, August 17-19 1995, Vienna, Austria. Austrian Society for Artificial Organs, Biomaterials and Medical Replacement Devices, pp 259-262 4. Liberson WT, Holmquest HJ, Scott D, et al. (1961) Functional electrotherapy: stimulation of the peroneal nerve synchronized with the swing phase of the gait of hemiplegic patients. Arch Phys Med Rehabil 42:101-105 5. Konishi N, Shimada Y, Sato K, et al. (1996) Correction of equinus gait in hemiplegic patients using the Akita heel sensor system. J Jpn Med Soc Paraplegia 9:290-29i 6. Dai R, Stein RB, Andrews B J, et al. (1996) Application of tilt sensors in functional electrical stimulation. IEEE Trans Rehabil Eng 4:63-72 7. Handa Y, Ohkubo K, Hoshimiya N (1989) A portable nmltichannel FES system for restoration of motor function of the paralyzed extremities. Automedica 11:221-231

8. Handa Y, Hoshimiya N, lguchi Y, et al. (1989) Development of a percutaneous intramuscular electrode for a multichannel FES system. I E E E Trans Biomed Eng 36:705-710 9. Hoshimiya N, Handa YA (1989) Master-slave-type multichannel functional electrical stimulation (FES) system for the control of the paralyzed upper extremities. Automedica 11:209-220 10. Kagaya H, Shimada Y, Sato K, et al. (1996) An electrical knee-lock system for functional electrical stimulation. Arch Phys Med Rehabil 77:870-873 11. Solomonow M, Baratta R, Beaudette P, et al. (1989) Gait performance of paraplegics ambulating with the reciprocating gait orthosis powered by electrical muscle stimulation. Proceedings of the l l t h International Conference of IEEE/EMBS, 1013 12. Marsolais EB, Kobetic R (1987) Functional electrical stimulation for walking in paraplegia. J Bone Joint Snrg 69-A:728-733 13. Ichie M (1990) Restoration of locomotion by FES. Proceedings of a Rehabilitation International Seminar pp 1-12 14. Andrews B J, Baxendale RH, Barnett R, et al. (1988) Hybrid FES orthosis incorporating closed and sensory feedback. J Biomed Eng 10:189-195 15. Hatta M, et al. (1993) Grip, pinch strength and motor hand skill in tile elderly: relationship to age, sex and anthropometric factors. Sogo Riha 21:489-492 16. Peckham PH, et al. (1997) Restoration of hand function in the quadriplegie through electrical stimulation. In: Hambrecht FT, Resuiek JB (eds) Functional electrical stimulation: application in neural prostheses. Marcel Dekker, New York, pp 83-95