DETERMINING USAGE OF A JUVENILE MYOELECTRIC PROSTHETIC ARM
Anita Bagley, Ph.D.; Michelle James, M.D.; Shriners Hospitals for Children, Northern California, Sacramento, CA Brian P. Self, Ph.D.; Jon S. H. Schoenberg, Ph.D.; Rachel Coleman, B.S.; Brian Denaro, B.S.; U.S. Air Force Academy, Colorado Springs, CO
Unilateral congenital below-elbow deficiency is the most common level of transverse failure of formation; its incidence is estimated at 1:20,000 live births (Birch-Jensen). It is the level most amenable to prosthetic use.
There are a range of prescription choices for children with unilateral below-elbow limb deficiency, including no prosthesis. The three general types of upper extremity prostheses available are: active body-powered, active myoelectric, and passive (sports or cosmetic) devices. Body-powered prostheses cost approximately $7,000.00, myoelectric prostheses cost approximately $15,000.00, and passive prostheses cost approximately $3,000.00.
Prescription choice involves a number of factors including age of the child, length of the residual limb, functional need of the child, reliability of return of the child to the clinic for appointments, and cost. There are no age-related standards for prescription of below-elbow prostheses; training protocols for prosthetic use and methods to assess function vary widely. Each type of prosthesis has its advantages and disadvantages. The body-powered prosthesis is lighter, faster, sturdier and easier to repair than the myoelectric device. The myoelectric prosthesis does not need a harness, and has a stronger pinch for the younger child. The passive prosthesis has no mobility but is light-weight and can have greater cosmetic appeal.
Research on upper extremity prosthetic use is limited. A recent survey (Davidson) of an adult population revealed that 56% of the 70 respondents wore their prosthesis once in a while or never. And the association between the individual's wear pattern and functional ability was low. Lerman et al. used a validated musculoskeletal functional health questionnaire to study 64 parents of both prosthesis users and non-users. There were no differences between the two groups in upper extremity function, ability to transfer, sports or happiness. Prosthesis users reported more pain.
For children, there is weak evidence to suggest that the best prosthesis, in terms of speed of performance and satisfaction, may be the body-powered type (Edelstein and Berger; Kruger and Fishman). However, both children and parents are frequently dissatisfied with upper extremity prostheses because functional gain may be limited and sensory feedback is blocked. Many children abandon their prosthesis when they are old enough to choose not to wear it (Scotland and Galway; Kruger and Fishman; Crandall and Tom-have).
Decisions to continue or change prosthetic prescription rely on parental or child report of use and desired functional capabilities. The actual use of the device may be judged by glove discoloration and history of repair. This subjective information may be augmented with occupational therapy notes that record ranges of motion and the ability to perform specific tasks. Validated tests (such as the Jep-son-Taylor Hand Function Test and the University of New Brunswick Test of Prosthetics Function) may be used to record standardized scores of performance. However, even these measures may be not reflect true function as behavior may be influenced by the clinical setting and the child's desire to perform well for the clinician.
It would be useful to have a direct measure of the amount of time the child wears his/her prosthesis in the "real world" and the number of times the device is used for prehension. Such information may be used to improve prosthetic prescription, to define specific training for use of the prescribed device, and to provide more cost-effective use of prosthetic funds. The purpose of this work is to develop a device to monitor prosthesis wear and usage patterns.
In order for the device to meet the demands of functioning in a real world situation, a number of system specifications were defined. First, the system needed to integrate unobtrusively into the prosthetic arm. The device needed to fit in the prosthetic cavity and be as light-weight as possible. The monitoring device needed to check the status of the prosthesis every minute to record whether the arm was being worn or not. If the prosthesis was noted to be "on", the device needed to record the number of activations during each 15 minute wear period. The device needed to function on an independent battery that did not need to be recharged or replaced for at least three months. The device needed to be able to communicate with a personal computer so data could be downloaded.
The prototype device was designed to operate in a myoelectric arm.
To determine whether the prosthesis was being worn, a force sensing resistor (FSR) was placed on the inner surface of the socket in contact with the residual limb Figure 1 .
This resistor is commercially made (Interlink Electronics) and consists of thin (approximately one mm) metallic conductors printed on flexible plastic. The sensor is round in shape and 2 cm in diameter. When the residual limb is in contact with the FSR, resistance decreases from 10MO to less than 100O. The FSR is connected to an operational amplifier, which converts the resistance change to a voltage change. This change in voltage is recorded and defined as a wear period.
To sense terminal device actuations, a current sensing resistor was added in series to the circuitry for the hand motor. When the terminal device grasps an object, the current to the motor increases and a proportional voltage change is generated across the sensing resistor. This voltage change is recorded as an activation. Up to 255 terminal device activations are recordable in each 15 minute interval over a three month period.
The center of the device is a microcontroller (Vesta Technologies; SBC2000-074 32K) which was chosen for its light weight and low power consumption. The microcontroller consists of a programmable interface controller (PIC) that includes memory sufficient for the three month data collection period, and a clock for accurate time-interval generation and measurement. It is compatible with the FSR and the current sensing resistor; it is easy to program and has a flexible layout. The PIC development kit included Vesta Basic ® software that is used for programming the chip and for allocating memory for recording wear time and grasping events.
The PIC has 34 kilobytes of erasable programmable read-only memory (EEPROM). It has two serial ports and five logical input ports. An important feature is the non-volatile data storage that preserves the integrity of the collected data if power to the system is interrupted. To program the EEPROM, the Vesta software code is loaded from a computer through a serial port.
A design block diagram of the PIC and sensors is shown below.
To initialize the system or to download the collected data, a serial cable is used to link the PIC with a personal computer. A simple terminal emulator program is used for the software interface. When connected to the computer, the PIC is initialized and the program stored in the EEPROM will prompt the clinician to enter the time of day. This information is used to set the PIC's on-board clock and the system is triggered to begin to collect data. Communication from the computer can also be used to reprogram the PIC for different time periods of data collection.
When the system is in operation and the FSR shows no contact or "no wear", the PIC goes into sleep mode to conserve energy (the sleep mode reduces power consumption by over 500%). It "wakes up" every minute to reassess the wear status of the arm. For data collection, each day is divided into 96 fifteen minute periods. Each period is assigned an 8-bit (one byte) memory address.
The collected data is read into an Excel (Microsoft Corp.) spreadsheet and arranged in columns. Each row represents one day. The first column lists the total number of actuations. The second column lists the total time worn in that day. Each subsequent column is a fifteen minute interval. If a column holds the number "0" (zero), that means that the arm was not worn during that interval. If the column holds the number "1" (one), that means the arm was worn during that interval. A number greater than one means that the arm was worn and that the number of activations in that interval equals the number in the column minus one.
Testing of the system is ongoing. Currently the prototype monitoring device weighs 87 grams and is powered by two Energizer AA e2 lithium cells (the batteries constitute 29 of the system's 87 grams). The prototype has been connected to the components of a myoelectric prosthesis and bench testing is being performed to ensure measurement accuracy in time worn and number of activations recorded.
Further issues being explored are increasing the device's memory and battery life, and decreasing its cost and weight. A device to record body-powered terminal device activations is being researched. Finally, it has been proposed to install prototypes into three myoelectric arms and have the children use their devices in their typical way for a three month period to assess the operation of the monitoring device under real world conditions. This wear and use data may be combined with the results of a multi-center study currently underway which assesses the function, quality of life and satisfaction of children (both prosthesis users and non-users) with unilateral congenital below-elbow deficiency. The goal of all of this work is ultimately to improve prosthesis prescription, training and design for this population.
Birch-Jensen A. Congenital deformities of the upper extremities. Andelsbogtrykkeriet and Det Danske Forlag, Denmark, Odense, 1949, pp 15-36.
Crandall RC, and Tomhave W. The role of the passive hand in unilateral below-elbow amputees; a long-term retrospective analysis of 35 individuals given multiple prosthetic options. ACPOC presentation, Banff, Alberta, Canada, April 27, 2000.
Davidson J. A survey of the satisfaction of upper limb amputees with their prostheses, their lifestyles, and their ability. J Hand Therapy Jan-Mar 62-70, 2002.
Edelstein JE, and Berger N. Performance comparison among children fitted with myoelectric and bodypowered hands. Arch Phys Med Rehab 74: 376-380, 1993.
Kruger LM, and Fishman S. Myoelectric and body-powered prostheses. J Ped Ortho 13(1): 68-75, 1993.
Lerman JA, Barnes DA, Sullivan E, and Haynes RJ. The Pediatric Outcomes Data Collection Instrument and unilateral pediatric upper limb deficiency patients. ACPOC presentation, Houston, TX, April 6, 2001.
Scotland TR, and Galway HR. A long-term review of children with congenital and acquired upper limb deficiency. J Bone Joint Surg 65(B): 346-349, 1983.