The Development And The Control Of A Powered Prosthesis For Children

D. C. Simpson, B. Sc., Ph.D. G. D. G. D. Sunderland, M. Sc.

This article originally appeared in Health Bulletin, Vol. XXII: No. 4, pp. 67-69, October 1964, issued by the Chief Medical Officer of the Scottish Home and Health Department. It is reprinted here by permission of the authors and publishers.

The experience of less than one year's work might be considered an inadequate basis for a discussion of the factors involved in the design of powered prostheses. Nevertheless, as we were forced by an urgent clinical need to design and produce our own system, it may be helpful to others to consider the underlying principles which have guided our work.

Because of the urgency of the problem, our first efforts were concentrated on the production of a working system with the minimum of delay. At the same time we tried to ensure that such a system would not retard or prevent later progress along lines which we envisaged as being logical, but which might require one or two years further research work to bring into a practical form.

To understand our approach, it is necessary to enumerate and discuss the many possible methods of powering and controlling a prosthesis as well as our concepts of the function of the prosthesis and its relationship to the child.

We consider that the first and most important function of a prosthesis is to give independence to the handicapped person. For the child, this can be interpreted as allowing him to eat, to use the lavatory, and later to receive education for a suitable career. Not only must the prosthesis enable the child to carry out these activities, but it must also enable him to begin the activity unaided, i.e. to pick up a spoon, to adjust his clothing, to select a pen. In achieving these essential functions, the experience and the sphere of activity of the child is enlarged.

Special skills can be built into a powered prosthesis, but nothing must be built in for one purpose which will conflict with its use for any other basic functions. It should be possible for the prosthesis to carry out its different tasks without modification, and special spoons, forks, pencils, should not be required.

The prosthesis must also be cosmetically acceptable. The case of the patient who rejects a functionally satisfactory hook in favour of a useless cosmetic hand when outside the hospital is a familiar one. By cosmetically acceptable we mean that not only should the prosthesis look natural, but that it should move in a natural way and should not be noisy. Noise in operation and puppet-like movement in a natural looking limb can give a grotesque effect.

It is sometimes argued that the patient's equipment should consist of a number of special purpose units, which would be separate from him, and that a more efficient system would be provided, for example, if instead of using a prosthesis, he were to feed himself by a machine placed on the table in front of him. This would undoubtably be the easy way out of many problems, but it neglects the human wish not only to do the same things as other people, but to do them in the same way. It would be unfortunate if the differences between the normal and the handicapped were emphasised by the equipment, on the grounds of expediency. This method would also limit the mobility of the child and would not give real independence.

Because the handicapped person wishes to appear, and behave like the normal, it does not follow that he will regard his prosthesis as a limb rather than an aid, and indeed it may be that the child with congenital dysmelia, and therefore no previous direct experience of the use and control of a natural limb, may view it in a much more impersonal way than the amputee. In the present state of prosthetic development there is no satisfactory tactile feedback from the device to the child and, without any built-in positional awareness, all his guidance and control must be based on his visual observations. This is a very similar situation to that experienced by the normal car driver who has no direct experience of being 12 feet long by 5 feet wide, but, by his visual observation can project and extend himself to these limits, and can adapt himself to controlling the mechanism after a period of training. The driver, as a person, ends at his control points, his hands and feet, although he extends himself in the machine -- the child, too, ends at his control points, although he extends himself in his prosthesis. If this is a correct analogy, it is important, because there is little evidence of left, or right, side domination in car control. A driver can change from left hand drive to right hand drive, from left hand gears to right hand gears, from conventional three pedal to two pedal controls with only a short period for re-orientation. It is, therefore, possible that mistakes made with small children in the choice of the site for the prosthesis and for its controls, may not be serious. It is important, however, to establish in the child the concept of the shoulder, or digit, as a control point.

With the conventional prosthesis the patient not only controls his prosthesis, but provides the power to operate it. In the powered case the patient is only required to control the power, which is provided from an external source. This control can be exercised by using limb or body movements to operate a lever,1 by using changes in muscle shape to operate the valve,2 or by using the amplified electromyographic potentials from selected muscles as the controlling signal.4 In physical form the controlling devices can be micro-switches, magnetically operated reed proximity switches, or variable resistances, all used directly, or with associated circuity for electrical control. For direct gas control miniature pneumatic valves can be operated by appropriate levers and linkage. With the correct transducers, an electrical control system possibly using logic units can, of course, be used to control gas or hydraulic power.

The modes of control of a prosthesis which can be obtained using these systems can be divided into three types, as follows:-

1. On-Off control. The power supplied to the prosthetic device can only be on or off, without any intermediate position. Fine control can be achieved

(a) by judging the 'on' time precisely or

(b) by arranging that the control is rapidly cycled on-off, on-off, etc., and by varying the ratio of the period when the power is on to the period when the power is off in each cycle, i.e. 'mark-space' ratio, to obtain the correct rate of movement.

2. Proportional control where the rate of movement of, or force exerted by, the prosthesis is proportional to the force applied to, or the displacement of, the control lever, or to a voltage which is itself directly related to an effort made by the patient.

3. Position-servo control where the position which the prosthesis will take up is directly related to the position into which the control lever has been moved, or to the amplitude of an applied voltage.

At present compressed CO2 and electricity are the only two convenient sources of energy for powered prostheses, and for a given weight of container, similar quantities of energy in either form can be stored. Both systems have advantages and in the general case there is no obvious superiority at either.

With compressed CO2, cylinders to contain the gas, and regulators to control the working pressure are easily obtained. Miniature lever operated valves to control the gas flow can also be obtained commercially. The transducers to operate the prosthesis can be pistons which will give thrust or pull, or bellows, and can be light. They have the great advantage that, by varying the cross-sectional area of the piston or bellows, the force available from a given gas pressure can be matched to the task without gearing. The stroke length of the actuator can also be easily adjusted. They have the further advantage that no energy is expended in holding a position. The system has the disadvantage of compliance and it is not possible to reduce this effect without reduction of the overall efficiency.

In Britain, it is not possible to recharge the storage cylinders in the home, and a recharging service must be arranged.

For the same power consumption the electrical transducer is, in general, heavier than its gas operated counterpart. The most efficient conversion of electrical energy to linear motion is probably made by a suitably geared high speed, low torque motor. Noise, wear and the inertia of the motor are the main disadvantages of this type of system. The great advantage of electricity lies in the ease of control.

A more convenient, but possibly less efficient, method of using electrical energy might be to use it to drive a small hydraulic pump and use the hydraulic system to actuate the prosthetic mechanism.5

On the basis of these assumptions and facts, we have produced a target specification for a child's upper limb prosthesis. It will be modified as work progresses .

The prosthesis should incorporate the following movements, functions and properties.

  1. . Prehension, which should be provided by a powered hand rather than any form of hook. It should have at least two modes of operation to allow for grasping, and for picking up small objects.
  2. Elbow flexion, with linked wrist movement.
  3. Pronation and supination.
  4. A shoulder movement, probably rotation of the arm.
  5. Compressed CO2 as a source of power. At present it has some advantages over electricity for small prostheses, because the transducers are smaller, quieter and suffer less from wear.
  6. The type of control should, in general, be position-servo. This will take full advantage of the prominent scapula which some of the thalidomide children possess, but could also exploit control by the e.m.g. potential in a new way by making the position taken up dependent on the tension of the muscle.
  7. The valves to control the gas flow must be electrically operated. This will not only enable servo control of the limb for some functions, but will also permit the use of magnetically operated proximity switches in cases where the child has a very weak digit.
  8. With so many separate functions it must be possible to perform several of them simultaneously rather than sequentially, in order to provide a natural movement.This implies that it should be possible to operate separate controls either in a combined form or simultaneously. Where there are only a limited number of control points it may be necessary to use controls on both sides of the body to operate a prosthesis on one side. This is not illogical if the 'car driving' analogy is correct.

The intention at present is to provide a fully equipped powered prosthesis and to balance it with a passive arm. In some cases the number of control points is so few, that the provision of active elements on both sides would reduce the capabilities of the prosthetic equipment.

We are also attempting to achieve all the necessary controls from the shoulders, digits, or the phocomelia, rather than by introducing methods which might have to be abandoned later or which necessitate unsightly or gross movement.

Our efforts at present have only gone a short distance along the road to this goal. Two controls are probably as many as a child of two or three years of age can manage and we have, therefore, produced a two function prosthesis, which has enabled children of this age to feed themselves. This has been done by combining elbow flexion with wrist movement so that the holding device maintains the same angle to the body throughout flexion. This is perhaps the only combined movement which can be built into a prosthesis which does not limit its use for other activities, and it has enabled the children to appreciate the benefit of a prosthesis at an early age. The natural type of movement which it performs is produced by the combination of the two movements. Twelve children have now been fitted and the first prosthesis has been described elsewhere,6 but, as the children are developing, it is now being extended by the addition of a pronation and supination movement(Fig. 1 ) superimposed on the same basic mechanism. The arm will be covered by a plastic material.

The advantages of the gas as a source of power for this type of prosthesis is demonstrated by the grouping of the actuators in the upper arm, giving a sound dynamic design and stability of the centre of gravity during use. The piston and cylinder operating the pronation movement is of the pull type, and a similar unit will be incorporated in the same region, to operate a powered hand when this is fitted. Until this is ready the hook shown in the illustration has been fitted. The two parts of the hook represent the thumb and first finger of a hand, and give a firm three point grip to spoon, pencils, etc.

It is also our intention to replace the present proportional control of elbow flexion by position-servo control, when our electrically controlled gas valve has been fully developed, as a further step towards our target of the fully equipped, five function, prosthesis.

D. C. Simpson, B. Sc., Ph.D. and G. D. Sunderland, M. Sc. are associated with the Powered Prosthetic Unit Orthopaedic Department Edinburgh University

1. Marquardt, E. Aktive Prosthesenversorgung eines armlosen Kleinkindes im 2. Lebensjahr. Jahrbuch der Fursorge fur Korperbehinderte 1962. 
2. Marquardt, E. Muskelsteuerung von pneumatischen Unter-und-Oberarmprosthesen. Archiv, fur orthopadische und Unfall Chirurgie 49, pp. 419-426, 1957. 
3. Battye, C. K., Nightingale, A. and Whillis, J. The use of myo-electric currents in the operation of prostheses. J. Bone Jt. Surg. 37B, pp. 506-510, 1955. 
4. Bottomley, A., Kinnear-Wilson, A. B., Nightingale, A. Muscle substitutes and myo-electric control. Journal Brit. I.R.E. pp. 439-448, 1963. 
5. Hauberg, G. Private communication. 
6. Lamb, D. W., Simpson, D. C., Schutt, W.H., Speir, N. T., Sunderland, G. D. and Baker, G. To be published.