Link to Figures Show Figures
	Article Text Figures References
	Send Link Send HTML Article

RESEARCH FORUM-- Methodology- Measurements, Part II: Instrumentation and Apparatus

Robert M. Havey
Thomas M. Gavin, CO
Avinash G. Patwardhan, PHD
Kevin P Meade, PHD

ABSTRACT

Laboratory instrumentation of orthoses and prostheses can be used to objectively assess the functional differences between various componentry designs and to measure the effect of an orthosis or prosthesis on the outcome of treatment. This article describes the electromechanical transducers and instrumentation systems that may be used to accomplish these tasks. The transducers are grouped in the following categories: stress and strain, linear and angular displacement, acceleration, force, pressure, temperature and humidity. Examples of transducer applications, instrumentation selection and integration are described.

Introduction

Recent advances in orthotics and prosthetics (O&P) have led to changes in fit and function of orthoses and prostheses. The desired effects of these advances are improved gait patterns, better limb and torso support, and greater deformity correction with lighter, more cosmetic orthoses and prostheses.

However, some of the newer, more expensive techniques do not significantly improve device fit and function or patients' quality of life. Clinical intuition, which has served orthotists and prosthetists well in the past, is inadequate to evaluate and differentiate significant advances from insignificant ones. Furthermore, the medical community as a whole is not likely to accept intuition as a basis to change its O&P patient management philosophies.

Laboratory experimentation and randomized, controlled clinical trials are the only accurate and reliable methods of evaluating and comparing fit and function of different orthoses and prostheses. Scientific method also will help determine compliance, patient satisfaction and outcome. Instrumentation systems are needed to measure functional differences between various designs of orthoses and prostheses and to document wearing compliance and outcome.

Instrumentation refers to the use of a device to control, regulate or measure an event or system. Instrumenting an orthosis or prosthesis can be as simple as holding a goniometer to measure a joint angle or as sophisticated as using an embedded control system to operate a myoelectric hand. In general, measurement systems consist of four main components:

• the system to be measured,
• transducers or sensors to assimilate information,
• signal conditioners to amplify or filter the transducer output, and
• the output device (1).

This article will cover electronic and mechanical measurement instrumentation systems by type and category and give examples of their use for research in O&P.

Measurement Decision Making

When arranging a laboratory or clinical experiment, some basic decisions must be made about collecting data. Once a research problem or question is clear and a concise hypothesis is formulated, the next step is to answer questions such as, "What do I want to measure?" and "How do I want to measure the event?"

Once the decision about what event to measure is made, some basic questions must be addressed before deciding how to measure:

• How accurately do I need to measure the event?
• Do I want a permanent record of the event (e.g., graph, printout, etc)?
• Does my research group have the technical competence to complete the experiment?
• Is the event static or dynamic?
• What are my cost limitations?

To answer these questions the researcher should have a basic knowledge of instrumentation components to be able to select the most appropriate transducers. The researcher also needs an event recorder and must be able to work with any necessary software for data acquisition. The next section will give a brief overview of available transducers.

Transducers

Transducers are the heart of instrumentation systems. A transducer is a device that converts (transduces) information from one form to another. Both the goniometer and embedded control system are transducers that are used to measure an event.

In the case of measuring a joint angle, a goniometer is used to convert a physical angle into a number that can be read. The myoelectric example may use a small computer and several types of transducers, including components such as EMG electrodes for input of muscle activity, force transducers for grip-force feedback and temperature transducers for thermal feedback. Transducers can be as simple as a ruler and either passive or active, or they can be vastly complex.

The output of transducers can be literal or figurative. With literal output, no conversion need be done to get the information in an understandable form, such as displacement read on a ruler or time on a chronometer. A figurative output may he detected in one form and then converted to another. A displacement transducer, for example, may produce a voltage signal that is proportional to the change in displacement. The voltage must then be scaled appropriately to yield a signal with displacement units such as meters or inches.

An active transducer is supplied power from an external source; this power is modified or limited by the transducer in response to the input stimulus prior to its output. A passive transducer creates its own output energy from the input stimulus. Thermocouples as well as piezoelectric-based devices are passive transducers because their output signals are generated by the device in response to the input stimulus. The appropriate input or excitation signal for an active transducer is supplied by a signal conditioner. Both active and passive transducers require signal conditioning before any signal can be recorded. Signal conditioning includes amplification, filtering and conversion techniques that produce a signal whose voltage varies proportionally with the transducer input.

The object of instrumentation is to acquire data about an event or system in the simplest, most accurate and efficient way. Complex and expensive devices are unnecessary to measure an event that may be measured efficiently and accurately with simple mechanical devices such as a ruler, goniometer, bathroom scale or common thermometer. Electronic transducers should be used to measure dynamic, long-term, high-resolution, complex systems or events that cannot be accurately measured using simple mechanical devices.

Transducer Applications for Orthoses and Prostheses

A strain gage can be used to measure torque or load on a transtibial pylon and strain in a custom-molded socket.

Force transducers can be used to measure force exerted by a pad on the patient; ground reaction force in a lower limb prosthesis; grip force in a hand prosthesis; seat forces; load bearing in a prosthesis; and wearing compliance.

Pressure transducers can be used to measure pad pressure on a patient, seat pressures and socket pressures.

Temperature transducers can be used to measure extreme environmental temperature conditions.

Humidity transducers can be used to measure extreme humidity conditions.

Angle/level transducers can be used to measure absolute angle for posture/inclination, joint angle and range of motion (flexion, extension, pronation, supination, abduction, adduction).

Displacement transducers can be used to measure deformation or deflection of a component as well as creep, expansion or contraction of a component exposed to extreme temperatures.

Acceleration transducers can be used to measure acceleration of prosthetic limbs to design components with the desired acceleration and damping (e.g., shock absorption, dissipation in prosthetic foot).

Criteria for Transducer Selection

Range is the first consideration when choosing a transducer. The measuring device must be capable of measuring the full-scale range of the event. Sufficient margin should remain at the upper end of the transducer's range so the measured event will not exceed the range of the transducer and cause damage.

Transducers with very small or very large ranges can be obtained. Typically, range and sensitivity are related in such a way that as the range of a transducer increases, the sensitivity decreases. For this reason a transducer should be chosen such that more than 10 percent of its range is used in a measurement.

For example, in an experiment designed to measure the force exerted on a pylon of a lower-limb prosthesis during walking, the forces received by the transducer would be between one and two times a subject's body weight. Therefore, a force transducer with a capacity of 500 lbs would be a good choice. If the average body weight is 175 lbs then twice body weight is 350 lbs, which is 70 percent of 500 lbs. A poor choice would be a 5,000-lb capacity transducer in which only 7 percent of the range would be used. Overrange protection is commonly built into transducers to prevent damage.

Frequency response of a transducer is the range of frequencies over which that transducer can give an accurate response. The transducer should be able to measure the full range of frequencies expected in the experiment. If the frequency range of a system is unknown, a device with a large dynamic range should be used. If a DC or static input is expected. it would be unwise to choose a device capable of measuring a device using only dynamic inputs. If high frequencies are expected, a transducer with increased stiffness. such as a piezoelectric transducer, will be required.

Mass or size of a transducer can adversely affect a system being measured. The transducer should not affect the measured system by adding significant mass or inertia.

Sensitivity is defined as the ratio of the change in transducer output to a change in the input to be measured (1). Transducer sensitivity should be maximized for a measurement system. To maximize sensitivity it is usually necessary to make trade-offs between range, weight and frequency response. Typically, the higher the sensitivity, the greater the signal-to-noise ratio, but higher sensitivity can result in a smaller useful range and lower resonant frequency. It may be necessary to consider transverse sensitivity if a multicomponent transducer is to be used or if the input to a transducer contains multiaxis components.

Accuracy refers to how close the output of the transducer is to the actual value and is affected by several factors, including linearity of the transducer and hysteresis.

Linearity is the closeness of a calibration curve or output of a transducer to a straight line. Many transducers are specified as linear, but in reality most deviate from a straight line particularly at either end of their useful range.

Hysteresis is the difference in the output of a transducer between the increasing and decreasing components of the transducer input. A transducer may measure an event differently when the input is increasing compared to when the input is decreasing.

Environmental conditions such as humidity, temperature, oil/grease, dirt and magnetic fields should be considered when choosing a transducer. All transducers are designed to be used in a specified temperature and humidity range. Most transducers are not waterproof, but some can be made so. (Most need to be operated in a small humidity range.) Dirt, oil and grease can affect transducers by interfering with the sensing element and cause changes in frequency response, thermal response or loading of the sensing element. Transducers should be resistant to the environmental conditions to which they will be exposed.

Cost should not be the most important consideration when choosing a transducer. Many characteristics of transducers are defined by cost, and it is important to choose the most cost-effective transducer that will be accurate and reliable enough for a specific experiment. Buying a transducer for a specific application or choosing a transducer for the resolution, range or speed needed for that application is a cost-effective approach.

Other factors that should be considered when choosing a transducer are ease of use and versatility. Ideally, the transducer should not be excessively difficult to use. This often can be the case when a device is selected that is designed to be versatile. The simplest transducers to operate are those designed to do one specific task (2).

Description of Transducers

Stress and strain are very closely related, and it is possible to calculate one if the other is known. Strain (epsilon) is defined as the fractional change in length per unit length (delta-L/L) or the change in length divided by the original length. Strain can be tensile (positive) or compressive (negative) and is a dimensionless quantity since length units cancel one another Strain is the deformation of a body and results from stress.

Stress (sigma) is defined as the force per unit area u= Force/Area) applied to an object (3).

Occasionally, it is not possible to directly measure stress in a system. In such an instance, the strain of that system measured and is then used to calculate the stress. If a body is stressed and the strain along a particular direction is measured, the stress can be calculated if the mechanical properties of the material are known. A complete analysis of stress acting on the body may require strain measurement in more than one direction (4).

The most common way to measure strain is to use a commercially available resistive strain gage. This device has undergone a considerable evolution since its conception. It previously was constructed of thin wire bonded to a dimensionally stable substrate in a zigzag pattern. Advances in the design of the device have resulted in a photo-etched foil strain gage. This new construction technique has improved the reliability and sensitivity of the device while reducing its size and cost.

These photo-etched strain gages consist of a very thin foil sheet (commonly Constantan alloy or nickel-chromium alloy) bonded to a dimensionally stable substrate such as polyimide film. The foil is then etched to produce a single conductor which zig-zags across the substrate (see Figure 1 ). The gage is then attached to the component to be tested usually by bonding or gluing the strain gage to the component with cyanoacrylate or an epoxy adhesive. As the component is stressed, it undergoes some dimensional change or deformation. Since the strain gage is bonded to the surface of the component, it also will undergo the same deformation as the component.

As the thin foil wire changes dimensions, the resistance of the wire changes. This resistance change is very small and needs to be converted to voltage and amplified before it can be measured. To convert the small changes in resistance of a strain gage into a voltage, a Wheatstone bridge is used. A Wheatstone bridge is a circuit consisting of four resistors and a power supply. One or all of these resistors can be strain gages. With voltage supplied to the circuit, a small change in resistance of any one of the four resistors will unbalance the bridge and cause a voltage change at the output points. This voltage can then be amplified and read by a volt meter and can later be converted into microstrain using available equations (4). More information on Wheatstone bridge circuits is available from strain gage manufacturers. Although strain gages can measure very small dimensional changes, they also are very sensitive to temperature changes. A small change in temperature can result in a large resistance change and thus introduce error in the measurement of strain. For this reason it is important to use temperature compensation techniques when using these devices.

The sensitivity of strain gages, called the gage factor, is the ratio of the fractional change in resistance to the fractional change in length along the axis of the gage (3). Foil strain gages typically have gage factors near two. If the component to be measured is very stiff and a much higher gage factor is needed, it may be necessary to use semiconductor or piezoresistive (PR) gages. The gage factor of semiconductor strain gages typically is around 125. Although these devices are much more sensitive than are foil strain gages, they have many less-desirable characteristics. Semiconductor gages are fragile, tend to be nonlinear, have a limited temperature range and have a much more elaborate temperature compensation (2).

Foil gages are available in a wide variety of sizes and configurations, including single gages and two-, three- and four-element rosettes (see Figure 1 ). A rosette contains several gages laid out on a single substrate. These rosettes are used to measure strain in more than one direction or to determine the principal direction of strain. Single-element foil gages are available in gage length from <0.3 mm to >150 mm (4).

Strain gages are used in many types of transducers, including displacement, force, accelerometers and pressure transducers.

Strain gages can be useful in measuring many properties of orthoses and prostheses. Some examples are the use of strain gages to measure torque or load in the pylon on a transtibial prosthesis and to measure the stiffness of Jewett or CASH orthoses, ankle joints on ankle-foot orthoses, knee joints and side bars.

Measurement of Relative Motion (Displacement)

Displacement transducers are used to measure displacement or position of one body relative to another. In general, displacement transducers can be grouped into two categories, contacting and noncontacting. Contacting transducers rely on mechanical contact between the transducer and the object whose displacement is being measured. This contact can be made by an attachment or by spring-loading the transducer-sensing element to maintain contact with the object. Noncontacting transducers use inductive, electromagnetic or capacitive coupling to measure relatively small displacements (millimeters) and ultrasonic or electromagnetic waves such as light-or microwaves to measure large displacements (meters) (2). Figure 2 displays several types of displacement transducers.

Contacting Displacement Transducers

One of the simplest types of contacting displacement transducers is the potentiometric transducer The wire-wound potentiometric transducer generally consists of a wire wound around a nonconducting core. This wire has a finite resistance and typically is made of platinum or nickel alloy.

In a linear potentiometer a sliding contact is introduced to the wire, which is capable of sliding axially along the core while remaining in contact with the wire. The ends of the wire are then connected to the positive and negative terminals of a power source. The wiper contact creates a voltage divider, and the voltage read is proportional to the distance of the wiper from the positive terminal of the device.

Potentiometric transducers are available for measuring linear as well as angular displacement. With a wire-wound potentiometer the resolution of the transducer is given by the number of wire turns per unit length. This type of construction causes wire-wound potentiometers to be expensive despite the many undesirable applications resulting from their low resolution. Improvements to the potentiometer have produced continuous-resolution elements. These elements are made of conductive plastic, carbon film, metal film or a ceramic-metal mix ("cement"). In addition to having better resolution, these devices have better thermal stability and are more cost-effective than the wire-wound type (1).

Although potentiometric transducers are the simplest, the most common transducers for measuring small-scale displacements rely on induction. Induction transducers convert displacement into a change in the self-inductance of a coil. The basic design of an induction transducer consists of two or more independent coils wound around a hollow insulator. A ferromagnetic core is allowed to freely pass through the coils. If an alternating current (AC) is supplied to one coil, a current will be created in the remaining coils. The phase and amplitude of the induced currents will depend on the position of the ferromagnetic core relative to the coils (5).

Induction transducers can be grouped into contacting and noncontacting types. Contacting transducers are most common. There are several types of contacting induction transducers, all of similar theory and construction. The most common of these is the linear variable differential transformer (LVDT). The LVDT is a sensitive device with typical resolution of less than 1.0 micrometer and linearity of less than 0.5 percent. LVDTs come in several different configurations with respect to the core. The core of an LVDT can be either captive or free. A captive core cannot be completely removed from the body or coil of the transducer due to mechanical stops. A free core is completely free and can be removed entirely from the body. A captive core is generally referred to as guided, which means some type of bearing or sleeve guides the core through the body. Captive designs also can use a spring to exert a constant axial force on the core. A spring-loaded core is useful in maintaining contact between the core and the component being measured without a rigid mechanical link. LVDTs also have the ability of being sealed so they can be used in hostile environments or under water (6).

Contacting displacement transducers can be used to mea sure the deflection of a mechanical assembly during a cycleto-failure or load-to-failure test such as testing the durability of a new prosthetic knee or orthotic ankle joint. They also may be used to measure gapping between two moving components such as a wrist unit on a myoelectric arm. Transducers that measure angular displacement or rotation are available in many types. Three of the most common angular displacement sensors are rotary potentiometers, rotary variable-differential transformers (RVDTs) and optical encoders.

Rotary potentiometers are the least expensive and simplest to use. These devices are similar in design to linear potentiometers and have angular measuring ranges from a few degrees to more than 3,600 degrees (10 turns). To improve resolution, mechanical amplification of the input can increase the angular motion of the transducer shaft.

An RVDT is similar in design to the LVDT but is used to measure angular displacement instead of linear displacement. RVDTs have a limited linear range of +/-60 degrees of rotation around the zero position. They are capable of full rotation but are nonlinear in the remainder of their range. These devices have excellent resolution and linearities better than 0.25 percent at +/-30 degrees.

Both of the angular displacement transducers described have analog outputs. Optical encoders also can measure angular displacement although their output is digital; they are available in incremental or absolute configurations. These devices use a plastic or glass disk attached to the rotating shaft of the transducer. This disk is etched with an alternating pattern of opaque and transparent segments that radiate from the center of the disk. A small light source is placed on one side of the disk and a light sensor on the other. As the disk is rotated the opaque and transparent lines alternately block light and illuminate the light sensor, and a string of ones and zeros are output from the light sensor corresponding to the alternating opaque/transparent lines on the disk. On an incremental encoder the ones and zeros are input into an up/down counter and indicate the relative rotational position of the encoder's shaft.

Absolute encoders use several tracks on the disk and multiple light sensors to determine the absolute position of the shaft for each revolution. Unlike an incremental encoder, the absolute encoder is capable of giving the absolute rotational position of the shaft and will do so even if power is interrupted to the device. The higher the number of transparent/opaque lines, the higher the resolution of the device. The line resolution of an encoder is expressed as pulses per revolution (ppr), and common resolutions range from 24 to 10,000 ppr (2).

Transducers used to measure the degree of tilt are commonly referred to as clinometers or tilt transducers. These devices use gravity to measure the angle of tilt from the horizontal. They can be designed many different ways, including attaching a pendulum to one of the previously mentioned angular displacement transducers. One common tilt transducer employs a variable capacitance design and is capable of resolutions of 0.001 degree with a range of +/-60 degrees (7).

Angle transducers can measure the angle of an ankle joint in an ankle-foot orthosis during different phases of gait. A tilt transducer may be used in a study to measure range of torso motion allowed by a TLSO.

Non contacting Displace men t Sensors

Noncontacting induction transducers, or proximity transducers, are useful because they impart no load or friction on the system being measured. Instead of using a conductive-core inductor design, these transducers use an air-core design and a conductive target or plate. They use an oscillator and tuned coil mounted near the tip of the transducer. When a conductive surface is placed near the transducer. eddy currents are induced in the conductive material, effectively dampening the oscillator signal. The signal change is amplified and is proportional to the proximity of the conductive surface relative to the sensing coil (6).

These devices are very accurate and have many advantages due to their noncontacting nature. However, they do require a conductive target whose size can get quite large with transducers capable of measuring displacements larger than 2.5 mm. This type of transducer also should be calibrated with the specific target used in the measurements especially if the target is curved or irregular in shape.

To measure large-scale displacements (>20 cm) it often is necessary to use ultrasonic or electromagnetic waves as the measuring medium. Although ultrasonic displacement transducers are common in consumer products such as autofocus options on some cameras and are quite useful in some applications, they generally are less versatile than systems that rely on light. This article will mainly focus on systems that use visible-light and infrared electromagnetic waves (2).

The Watsmart System is capable of measuring the three-dimensional position and displacement of discrete targets. This system uses optoelectronic cameras to track infrared-emitting targets through a precalibrated space. Three-dimensional positioning can be determined with two cameras. The cameras use lateral effect photodiodes as the sensing elements. These photodiodes are sensitive only to infrared light and generate a voltage on their surface in response to this light.

Each camera is capable of determining the two-dimensional location of the infrared light source within the camera's coordinate system. By using three-dimensional reconstruction, the two-dimensional information from each camera can be converted to three-dimensional coordinates by direct linear transformation (DLT) techniques.

While the Watsmart System is an excellent tool for measuring three-dimensional motion, it does have limitations. Foremost is the system's sensitivity to external and reflected infrared light, which requires the system's isolation from external sources of infrared light, including incandescent lighting and natural sunlight. All surfaces that can reflect infrared light from the targets to the cameras must be eliminated or covered with a nonreflective material. In addition, the Watsmart System must be calibrated for each setup by using a target array supplied by the manufacturer.

Another limitation, which also could be considered a strength, is the nature of the targets. The system uses infrared emitting targets (diodes) that are turned on and off one-by-one to allow the cameras to see only that point of infrared light. This can be a problem because of all of the wires needed to power each diode, especially if the system to be measured is in motion. This hard-wired approach, however, can be an advantage when it comes to data processing. Since the position of each target is measured independently, software can be written to automate the data processing. This automation can save valuable time by processing the data while the test system is still set up. If the data are found to be inadequate or corrupt. more data can immediately be collected. The Watsmart is capable of viewing 64 separate targets at accuracies of less than 1.0 mm.

This system has been superseded by a more advanced system that is easier to use and much more accurate (8). The SelSpot System, made by a separate company. is still manufactured and uses the same technology'.

A new system called the Optotrak System is similar to the Watsmart System in that it still uses optoelectronic-type cameras and infrared targets and can be used for the same applications. However, the technical design of the system is considerably different. Instead of using two or more two-dimensional photodiode sensors as does the Watsmart, the Optotrak System uses three or more one-dimensional charge-coupled device (CCD) arrays. The Optotrak System is available in two configurations: the three-dimensional Bar Optotrak and the two-dimensional Camera Optotrak, each capable of tracking as many as 256 targets.

The three-dimensional Bar Optotrak is designed for application with fixed volumes. This fixed volume can range from an area 1 m x 1 m at a camera distance of 2 m up to an area of 2 m x 2 m at a camera distance of 4 m. The accuracies locating a marker in this volume can be as low as 0.1 mm in horizontal and vertical directions and 0.15 mm in depth. The three-dimensional Bar system uses three one-dimensional CCD arrays mounted in a single unit. This configuration is precalibrated at the factory; so no user calibration is necessary.

The two-dimensional Camera Optotrak also is capable of measuring three-dimensional motion, but its configuration is similar to that of the Watsmart System. This system uses two or more cameras, each containing two one-dimensional CCD arrays, allowing each camera to see in two dimensions. With multiple cameras capable of two-dimensional digitization, three-dimensional motion can be calculated as with the Watsmart System.

The two-dimensional Camera Optotrak System is capable of measuring volumes as large as 4 m x 4 m x 4 m with accuracies of about 0.4 mm. Since this system uses more than one camera unit, it is subject to calibration prior to use. This calibration is rather simple and much less time-consuming than the calibration for the Watsmart System.

Other systems for measuring large displacement use video-type cameras for data collection. These video-based systems use CCD cameras similar to those used in the Optotrak System. Video systems generally use passive or reflective markers placed on the object (subject) to be measured. These markers are then illuminated simultaneously, and a video frame is recorded.

These video-based systems have several advantages over active target systems such as the Watsmart and Optotrak systems. The main advantage of the video-based systems is they need no wire connections to the targets. Therefore, the object (subject) being measured can contain very complex motions or moving parts. Data processing can be more involved with these types of systems since all targets are illuminated simultaneously. In this situation it is necessary for the operator to identify the markers in the first frame of data. The software is capable of tracking the targets in the data after they are identified. Video-based systems are capable of viewing a greater quantity of targets than an active system. However, the sampling rates for video-based systems are typically lower than for systems with active markers.

The Watsmart, Optotrak and video-based systems can all be used in a comparative study of different spinal orthoses to evaluate their ability to limit gross trunk motion. These three-dimensional motion analysis systems also can be used for gait analysis studies to compare the effects of different prosthetic feet or different socket designs on gait.

Measurement of Acceleration

Acceleration is the change in velocity over a specific time period and is generally expressed in SI units of m/sec/ sec(m/sec²). The acceleration produced by the force of gravity is g and is equal to 9.81 in/sec² (32.2 ft/sec²).

A device to measure acceleration is called an accelerometer. All acceleration transducers work on the principle of a seismic mass. This mass is restrained by a spring, and when an acceleration is given to the transducer case, the mass will move with respect to the case and exert a force on the spring. The acceleration can be calculated by measuring the displacement of the mass or by measuring the force exerted by the mass on the spring. There are a variety of methods for measuring the displacement or force of the mass. Acceleration transducers have been made using unbonded, bonded, metallic foil, and semiconducting (piezoresistive) strain gages, capacitive elements and piezoelectric elements, among others (2).

Loads on hip and knee joints during gait in patients wearing lower-limb prostheses are influenced by mass and acceleration of the prosthetic limb. Accelerometers can be used to measure the acceleration, deceleration or shock imparted to a prosthesis or orthosis during walking and running.

Capacitive transducers work on the principle that the gap between two plates or electrodes will change under acceleration if one of the electrodes is fixed to the case of the transducer and the other, supporting a mass, is allowed to move or deflect. As the gap between the electrodes changes, the capacitance changes. This change can be either positive or negative depending on the acceleration. These devices typically have a large mass, a low frequency response (DC to 1,000 Hz) and a small range of acceleration (up to 50 g).

Capacitive accelerometers do have a fairly large shock limit of 10,000 g and have very high thermal stability (-50 to 120 degrees Celsius) and sensitivity. Capacitive transducers typically are larger and more expensive than PE and PR devices because the signal conditioning must be integrated into the transducer to prevent capacitive effects of connecting cables (2).

In a piezoelectric (PE) transducer, the acceleration acts on a mass that exerts a force on a ceramic or quartz crystal. The force caused by the acceleration produces a charge in the crystal. The quartz or ceramic plates generally are preloaded with a static load so the transducer can measure positive as well as negative accelerations. Piezoelectric transducers have a low output signal and a high output impedance. Due to the high stiffness of the crystals, piezoelectric transducers have a very wide useful frequency range, typically from .05 to 10,000 Hz.

Since PB devices cannot measure static or DC signals, their use should be restricted to systems in which acceleration is not constant. PB accelerometers are available in a wide range of capacities from +/-5 g to +/-10,000 g. The temperature range for PB devices is typically from - 100 degrees to 200 degrees Celsius (2).

In strain gage accelerometers the strain or deflection of the spring mass system is converted into a change in resistance. The strain gages used can be either metal foil or semiconductor types. As mentioned earlier, semiconductor gages are much more sensitive than foil gages, and the semiconductor-based transducers can have a much stiffer spring-mass system. The increased stiffness of this system will produce a device that has a much higher frequency response as well as a higher output signal. Foil gage accelerometers are uncommon so this article will focus only on semiconductor or piezoresistive devices.

Semiconductive strain gage devices, also referred to as piezoresistive devices, are the simplest in concept and require the simplest signal conditioning. These devices are made of silicone strain gage elements bonded to cantilever beams and electrically connected in a Wheatstone bridge configuration. With this configuration the devices need only an excitation voltage (commonly 10 volts) to output a voltage proportional to the acceleration input. The output voltage is then amplified using a standard instrumentation amplifier to a level appropriate for the recording device.

Piezoresistive devices are commonly available in ranges from +/-25 to +/-6,000 g and frequency responses of DC to 5 kl-lz with shock limits as high as 20,000 g. The temperature range for PR devices is much smaller than that of other accelerometer types. Typically the useful range for these devices is from -25 to 100 degrees Celsius (2).

Semiconductor strain gage and piezoelectric devices are good choices for low mass devices. Accelerometers are available in single- and multiaxis configurations. If measurements are needed on more than one axis of acceleration, a multiaxis device, which is generally lighter and smaller than a group of single-axis devices, can be used.

Measurement of Force

Force transducers are used for measuring force as well as measuring the weight or mass of an object. Force can be measured by a variety of means; two of the most common ways include balancing an unknown mass against a known mass and measuring the elastic strain produced in an element by an unknown force. The first of these is demonstrated by devices known as scales. These devices measure the force caused by the mass of an object and its acceleration due to gravity. The most familiar of these devices is the spring or balance scale. This type of scale can be found in the home or at the doctor's office. They typically are accurate and simple to use.

The second type of force transducer mentioned is a device that uses some method of measuring the deflection of a mechanical member due to an unknown force. This deflection can be measured in a variety of ways using almost any type of displacement transducer, but the most common way is by using a bonded strain gage. The strain gage used can be either foil or semiconductive depending on the size of the device. Strain gage-based force transducers are the most common type of force transducer and are commonly referred to as load cells." Load cells are constructed in many different ways to measure several different forces from fractions to thousands of Newtons.

Strain gage-based force transducers typically are de signed to measure compressive as well as tensile forces, but they can be designed to measure one or the other as needed (2). Forces that occur along more than one axis must be measured in certain instances; to do so, a multiaxis (or multicomponent) load cell must be used. In multiaxis load cells an element that measures force in one direction should have no influence or interference with elements that measure forces in other directions. In reality this interference or crosstalk is generally of the order of 1-2 percent of full-scale load.

Load cells have a rated capacity or full-scale load. This is the amount of load the device can transduce without significant error. These devices also have a safe overload range, which is the amount of load the device can receive before damage or permanent deformation occurs to the sensing member. This overload range is typically 150 percent of the full-scale load.

Load cells are available in many different sizes (see Figure 3 ), ranging from units that will fit on your fingernail to large devices used in industry that weigh hundreds of kilograms. The operating temperature range for strain gage-based load cells is typically between -50 and 120 degrees Celsius, but this range may be significantly smaller for some cells (specifically semiconductor gage-based cells).

Strain gage load cells rely on the deflection or deformation of a mechanical member to measure force. This deformation occurs in a finite amount of time and is directly related to the stiffness of the cell. The time for one of these devices to respond to a step input can range from milliseconds to seconds. If a quickly changing force (dynamic force) such as impact is to be measured, a much faster measurement device is needed.

To measure dynamic forces a piezoelectric force transducer can be used. These devices measure only rapidly changing forces and cannot measure static or constant forces. Dynamic force transducers use quartz crystals as the sensing and transduction element, thereby eliminating the elastic member found in strain gage load cells. (The relatively large displacement of the sensing element in strain gage load cells is responsible for their poor frequency response characteristics.) Quartz crystals have an inherently high stiffness and are ideally suited for high-frequency measurements.

Force transducers can be integrated into an orthosis or prosthesis to measure ankle/wrist forces, moment and load. They also may be used in gait analysis to measure ground reaction forces and loads exerted by a strap or pad in an orthosis.

A Force-Sensing Resistor? (FSR?) is a thin polymer device that exhibits decreasing electrical resistance in response to increasing force applied perpendicular to the surface of the device.

FSRs are available in many shapes and sizes and can be custom ordered in single-element or multiple-transducer arrays in almost any shape (see Figure 4 ). The FSR has a three-part construction as shown in Figure 5 . The top layer is a piece of flexible polymer bearing a semiconductive/resistive material shown in black. The bottom layer also is a piece of flexible polymer, but it has a set of metallic interdigiting electrodes deposited on it. These two electrodes with their interdigiting fingers are arranged such that the fingers come in close proximity to each other but never touch. The middle layer of the device is a thin adhesive strip that holds the device together. The adhesive also acts as a spacer to prevent the resistive material from contacting the electrodes in the unloaded state.

As force is applied to the FSR, the semiconductive material contacts the interdigiting electrodes, and current flows from one electrode to the other through the semiconductive material. As force increases, contact becomes more intimate, and more current is able to flow, thereby lowering the resistance of the device.

The force versus resistance relationship of an FSR is almost logarithmic. With no load on the FSR there is no contact between the electrodes and the resistive material, and the FSR is seen as an open circuit. In the low force range resistance decreases very rapidly as compared to resistance change at higher loads. It is this relationship that makes the FSR well-suited for use as a switch. The force range of a typical FSR is 0.1 to 200 Newtons. However, using different types of polymer film and different formulations of the black semiconductive material, a wide variety of force ranges and sensitivities can be created. The thickness of a typical FSR is less than 0.5 mm and can be made almost any shape or size, from millimeters to meters. FSRs have a fairly large temperature range of -30 to 170 degrees Celsius and can be designed for use in situations where the relative humidity is 100 percent.

Due to their mechanical nature, FSRs have a very slow rise time (typically 1-2 m/sec) and therefore are not well-suited for high-frequency applications. One of the major limitations of FSRs is their creep response. Creep is the deformation or strain of a mechanical member over time when exposed to a constant force. In FSRs this creep manifests itself as a decrease in resistance over time. If the FSR is used to measure the absolute value of a force, this creep can be a source of error; however, if the FSR is being used simply as a switch, this creep can be valuable inasmuch as the creep will force the FSR further into the triggered zone. The creep for a typical FSR is of the order of I percent/kg-day'.

Applications for FSRs include the measurement of contact forces between an orthosis and soft tissue. During gait analysis FSRs attached to the foot can determine heelstrike and toe-off, and they can be used in a compliance monitor to determine if an orthosis or prosthesis is being worn properly and for the prescribed amount of time. However, when using an FSR to measure interface forces in an orthosis or prosthesis, the researcher must be aware that the FSR cannot measure shear forces and should only be used to measure pure compressive loads.

Measurement of Pressure

To measure pressure an electrical signal must be generated in response to a pressure input. Typically, pressure is measured by allowing the pressure to deflect or strain a member of the transducer. This strain can then be measured in a variety of ways, using capacitive, PR and PE as well as other techniques to measure displacement.

Pressure is defined as the force exerted over a surface per unit area of surface. Pressure is measured in the same units as mechanical stress N/in² (Pascals, Pa).

When measuring pressure, the relative difference between two pressures usually is measured. When selecting a pressure transducer it is necessary to decide which of four references is appropriate for the application. These references are absolute pressure, which is pressure referenced to zero pressure or a vacuum; differential pressure, which is the difference in pressure between two user-supplied points; gauge pressure, which is the measurement of pressure relative to ambient or atmospheric pressure; and sealed gauge pressure, which is the measurement of pressure referenced to a sealed chamber at standard atmospheric pressure.

Pressure can be measured several ways. The most familiar pressure transducer to most researchers is the aneroid barometer, which is domestically used to measure atmospheric pressure. This article will focus on sophisticated pressure transducers capable of measuring static and dynamic pressures with high accuracy.

Pressure transducers rely on the bending or strain produced in an element of the transducer. One simple way is to use a bellows or flexible enclosure. A displacement transducer such as an LVDT can then be used to measure the change in length or size of the enclosure. With this device the deflection will be proportional to the pressure. Although these types of transducers work quite well over a large range of pressures, due to their mass they have a very low frequency response.

Pressure typically is measured with a diaphragm pressure transducer. These devices started as thin, flexible, round diaphragms with wire or foil strain gages mounted to them. The diaphragms used in most pressure transducers today are monocrystal silicon though some devices still use a sheet-metal diaphragm. These diaphragms cover a small cavity, which can be sealed or left unsealed. The strain in the diaphragm can be measured using foil strain gage, piezoelectric, piezoresistive, capacitive and other techniques (9).

Piezoresistive transducers use a diaphragm machined from monocrystal silicone. This diaphragm is placed over a cavity that can either be unsealed or sealed to measure gauge or absolute pressure. These devices are made by depositing piezoresistive elements or semiconducting strain gages on the silicone diaphragm. When the diaphragm is exposed to a pressure the diaphragm is strained, and the PR elements can detect this strain. PR devices can have full-scale pressure ranges from 0.013 to 140 MPa (2-20,000 psi) and resonant frequencies from 70 to >1,000 kHz. The temperature range for these devices typically is from -50 to 120 degrees Celsius.

Piezoelectric pressure transducers use quartz as the sensing element. The physical properties of quartz, including its high stiffness, strength and wide temperature range, make it an almost ideal choice for use in a pressure transducer. Quartz-based transducers ideally are suited for measuring dynamic or quasistatic events and cannot measure purely static events. Due to the high stiffness of quartz, quartz transducers can have extremely short rise times commonly in the I microsecond range and resonant frequencies as high as 500 kHz. The temperature range for these devices can be from - 200 to 350 degrees Celsius with pressure ranges in excess of 1,100 MPa (160,000 psi).

Piezoelectric pressure transducers generally are referred to as gauge pressure transducers. This is because PE devices produce an output only when the pressure acting on the diaphragm changes. To measure pressure relative to a constant, such as atmospheric pressure, a device that can measure static or steady-state pressure is needed.

A capacitive pressure transducer can be made by measuring the capacitance change between a plate attached to a diaphragm and a stationary plate. These transducers are not common but do have higher sensitivity to pressure than do PR devices (typically 10 to 100 times). They are much less sensitive to thermal stresses and local diaphragm stresses since capacitive transducers integrate the movement of the entire surface of the diaphragm while PR and PF devices use localized strain measurements. Capacitive transducers commonly have small capacities and generally are more expensive and larger than other devices because they must carry their signal conditioning circuitry on the same chip as the sensor (7).

Pressure transducers may be used to measure seat pressures, interface pressure between orthosis and skin, or pressure in a prosthetic socket to evaluate the quality of fit. They also can be used to measure pressure in cadaveric intervertebral discs during compressive loading and quantify the effect of a TLSO on reducing intradiscal pressure.

Fuji? film is used to measure contact pressure. The film is purchased in rolls and is cut to size as needed. This film is available in five pressure ranges. The lowest pressure film has a usable range of 2 to 6 kgf/cm2 (28 to 85 psi); the highest pressure film has a range of 500 to 1,300 kgf/cm2 (7,100 to 18,500 psi).

This pressure-sensitive film is very useful for measuring the distribution of pressure between two surfaces. Fuji film is useful for measuring static pressures only and cannot measure pressure-versus-time data. Only the maximum pressure is recorded by the film. When using this film it is important to keep both surfaces parallel; Fuji film does not work well on irregular or curved surfaces due to the shear forces encountered. These shear forces can erroneously be recorded as axial forces.

There are two different types of Fuji film. The first uses two sheets of polymer material, and each sheet is coated with a different material. These sheets are referred to as A-film and C-film. The A-film coating consists of very small microcapsules of non-colored, color-forming material. The C-film coating consists of a special color-developing material. When these two films are placed together and exposed to a pressure, the microcapsules on the A-film break open, spilling their color-forming material. This color-forming material is then absorbed by the C-film and is developed by the color-developing material.

The intensity of the resulting color is a function of the amount of pressure applied to the two films. This color ranges from a very light pink for very light pressure to a bright red for full-scale pressure. The second type of Fuji film uses a single sheet of film onto which the C-material is first deposited followed by the A-material. This technique is used only in the highest pressure-range film.

The intensity of the resulting color is then read by one of two ways. The easier method uses a densitometer to measure the intensity of the color on the developed film. This technique is adequate if only spot measurements are needed. If the center of pressure, center of force or total force are needed, a much more sophisticated analysis technique must be employed. This method of analysis uses a camera to digitize the film and a computer with special software to analyze the resulting image.

Fuji film is a very useful tool, but special precautions must be taken when working with it. The film must be used within the temperature range of 5 to 35 degrees Celsius and 20 to 90 percent RH. The coated surfaces of the film should not be exposed to moisture or fluids. The films should be stored properly and used within the given shelf life. These films are not reusable and should be handled with care.

Fuji film can be used to measure joint contact pressure in cadaver specimens, pinch pressure in a prosthesis, axial load on pylons, and pressures on the plantar surface of normal or prosthetic feet.

Measurement of Temperature

Temperature transducers generally fall into one of two categories. They are either resistive devices with semiconductive or metallic-sensing elements or thermoelectric devices. Temperature transducers have been used as activation transducers in some compliance monitor prototypes (8).

Thermoelectric sensors work on a principle known as the Seebeck effect. This phenomenon converts a temperature difference directly into a voltage without an external power source. Thermoelectric transducers are commonly known as thermocouples and consist of two wires or strips of dissimilar metals or semiconductors. These wires are laid parallel to one another and joined at one end. As the common end of the transducer is heated, a voltage can be measured between the two free ends. This voltage is proportional to the temperature difference between the hot and cold ends. The voltage between the free ends will be zero if there is no temperature difference between the two ends and will increase as the temperature difference increases. The output of a typical thermocouple is less than 5 mV per 100 degrees Celsius temperature difference. This output voltage generally must be amplified before it can be measured. The useful temperature range for these devices can be from -200 to 1,200 degrees Celsius with some tungsten alloys useful up to 2,000 degrees Celsius (7,9).

Metallic-resistive temperature transducers are constructed of metallic wire wound around an insulating core. These devices are very reliable and have stable, repeatable responses. The resistance of the wire changes in response to temperature and the resistance-temperature relationship is linear. One disadvantage of these transducers is their wire-wound construction may cause a very long response time, or time constant, typically in the order of seconds.

Newer versions of these transducers are constructed by depositing a matrix of the metal on a flat ceramic substrate. These devices can be trimmed by a laser to produce a desired output response and generally have much shorter time constants due to their smaller size (approximately 3 cm x 1 m x 1 mm). In addition, these devices are useful over a very large temperature range-typically from - 200 to 900 degrees Celsius (10).

Thermistors are semiconducting devices made from mixtures of metal oxides that produce a change in resistance in response to a change in temperature. These devices may have a positive (PTC) or a negative temperature coefficient (NTC), depending on their construction; the majority have an NTC. The resistance-versus-temperature characteristics of thermistors are highly nonlinear; however, this nonlinearity can be significantly improved over small ranges by using special circuit designs or by using a lookup table. These devices are useful for temperatures of - 100 to 300 degrees Celsius. The main advantages of using a thermistor are its low cost and fast response time, typically a few milliseconds; however, some accuracy is lost (5,6,10).

A second type of semiconductive temperature transducer relies on a pn junction. A pn junction is the boundary area between p-type and n-type semiconductors (11). Pn junction devices rely on the forward voltage/temperature coefficient of a diode or transistor, which is typically about -2mV/°C. Pn junction temperature transducers can be used in one of two ways: Either the voltage can be held constant and the current can be monitored across the junction or the current can be held constant and the voltage can be monitored.

The advantage of these devices is their construction allows them to be integrated directly onto integrated circuit chips. These chips also can contain all of the circuitry needed to amplify or condition the signal. The ease of manufacture of these devices makes them very inexpensive, and the fact that all signal conditioning can be built into a single chip makes them very attractive to designers. These devices do. however, have a limited useful temperature range, typically from -50 to 150 degrees Celsius but have a very high linearity over this range of less than 1 percent (6,7,10).

Temperature transducers can be used in environmental studies of orthoses and prostheses and to test the durability of materials.

Measurement of Humidity

Humidity refers to the presence of moisture in a gas. Absolute humidity refers to the mass of water contained in a unit mass of gas. The amount of moisture that can be suspended in a gas is limited, and the maximum humidity attainable is the saturation humidity.

The saturation humidity of a gas depends greatly on temperature. A very cold gas is capable of containing a small amount of humidity while a warm gas may contain very high amounts of humidity. It often is more important to know the relative humidity (RH) rather than the absolute humidity. Relative humidity is the absolute humidity at a given temperature divided by the saturation humidity at that same temperature and usually is expressed as a percent. In other words, the percent of relative humidity is equal to 100 times the actual or absolute humidity divided by the saturated humidity.

Another measure of humidity is the dew frost point (DFP). The DFP is the temperature at which the water vapor in a gas becomes so saturated it can no longer remain suspended and condenses into water or ice (5).

Many different types of humidity transducers will be explained in this article. A direct measure of the relative humidity can be made by using a resistive, capacitive or liquid sorption transducer. Resistive RH transducers use a water vapor-permeable coating covering electrodes that are laid out on a ceramic substrate. As an alternating current is passed through the transducer its resistance changes according to the amount of water vapor present. Capacitive RH transducers are similar to resistive transducers except capacitive transducers use a dielectric material that absorbs and desorbs water vapor, which changes the capacitance of the device. The sorption type of transducer simply changes volume and mass by absorbing and desorbing water vapor. This volume or mass change can be measured in a variety of ways.

DFP transducers can be used to calculate relative humidity or absolute humidity. DFP transducers generally work in one of two ways. DFP can be measured using a chilled mirror. A light beam is reflected off the mirror and when the mirror reaches the temperature of saturation the water vapor will condense on the mirror, thus preventing the reflection of light. Another way to measure DEP is with a saturated salt transducer that uses a fabric or wick treated with electrically conductive lithium chloride and a coil used to heat the fabric. As water vapor is absorbed by the fabric, it becomes conductive. This current flow heats the coil, which evaporates some of the moisture. This absorption/evaporation process eventually reaches equilibrium, and the temperature measured is the DEP' (5,12).

Humidity transducers are helpful in simulating adverse or extreme environmental conditions to determine what effects those conditions may have on the long-term use of an orthosis or prosthesis and on commonly used adhesives and materials.

Peripherals

Signal Conditioning/Amplification

Most of the transducers discussed in this article need some type of signal conditioning or amplification before their signals can be read. Signal conditioners generally are purchased with the transducers and are configured to give their outputs in convenient units of voltage.

Signal conditioning is no place to skimp on quality. The output of a transducer is only as good as the conditioning equipment that receives it. If the signal conditioning equipment cannot respond to the frequencies and amplitudes of the transducer signal, then the time and money spent on a quality transducer has been wasted.

Each transducer requires a specific type of signal conditioner. The manufacturer of the transducer typically can supply the appropriate signal conditioning equipment and/or the specifications if a third-party device is to be used.

Data Acquisition Systems

Transducers are needed to change information from one form into another. Typically, the transducer output is converted by the signal conditioner into voltage that must be measured or recorded. Depending on the frequency of the event being measured, the data acquisition system can be as simple as a voltage meter and a paper and pencil. For low-frequency or DC signals, this technique may be sufficient. However, if the signal contains high frequencies or has a long duration, more sophisticated data acquisition and recording techniques may be needed.

When choosing a data-recording device it is necessary to know the characteristics of the signals to be recorded. These characteristics are amplitude or voltage range, signal frequency, event duration and required resolution. A decision also must be made about what type of output is needed. The signals can be saved on paper or an oscilloscope screen or in a computer.

Strip Chart

Strip charts use a pen or thermal process to trace a signal voltage onto a strip of graph paper. The paper speed can be changed from millimeters-per-hour for very slow or static signals to meters-per-second for high-frequency or dynamic signals. Strip charts are useful for some low-frequency applications since they're easy to use, and their output is easily understood. For signals with long durations or frequencies greater than 100 Hz, large amounts of paper are needed, which can get expensive and wasteful. Strip charts also are limited in the number of channels available for recording (typically >8). In addition, since the output of a strip chart is on paper, the resolution and types of data analysis that can be performed on the data are limited.

Oscilloscopes

Oscilloscopes are useful for recording transducer signals. However, they too have some limiting characteristics. Oscilloscopes are good for frequency as they display small segments of data that contain components from DC to many megahertz. However, as the record duration increases, the resolution of the recording decreases; thus, these devices are not useful for recording high-frequency information for extended periods of time. Oscilloscopes also are limited in the number of channels available.

Analog-to-Digital Conversion

The recording devices discussed previously limit the frequency, resolution and duration of signals that can be recorded. A more versatile device, such as an analog-to-digital convertor (ADC), is able to accurately acquire signals with frequency components from DC to several hundred thousand hertz when it is coupled with a personal computer.

Analog-to-digital conversion is a technique in which an analog signal is divided into discreet points that can be represented by a string of numbers. ADCs are commonly available in resolutions from 8- to 16-bit. An 8-bit convertor has 28, or 256, discreet voltage levels over a specified voltage range; a 12-bit device has 212 or 4,096; and a 16-bit has 216, or 65,536. This means that if the voltage range for an ADC is 10 volts, the 10 volts will be divided by 4,096 for a 12-bit AID convertor so each discreet step will be 2.44 millivolts (mV). As the number of bits increases, the resolution of the device increases.

An ADC is purchased and installed inside or connected to a personal computer. These devices are available for all of the major computer and software platforms available today, including IBM-compatibles running DOS, Windows and OS/2, and Macintosh systems. ADCs are software-programmable devices. Data acquisition software to control ADCs is available from the manufacturer and is capable of recording, displaying and providing limited data analysis. Software also can be custom-designed for specific applications with routine libraries that are available from the manufacturer.

Examples of Instrumentation Selections for O&P

Compliance Monitor for Orthoses and Prostheses

A compliance monitor is a good example of the importance of choosing a proper transducer and ensuring the recorded stimulus accurately portrays an event.

Spinal orthoses play an important role in the treatment of spinal injuries, low-back pain and spinal deformities. Whether or not a patient complies with the prescribed orthosis wearing hours greatly influences the clinical outcome of orthotic treatment. However, there are no objective data to evaluate whether a correlation exists between the orthosis wearing time and the outcome of treatment. Further, there is no rational basis to determine the minimum number of wearing hours necessary to achieve good outcome for a given condition and type of orthosis.

Currently, no reliable and objective method of measuring orthosis wearing time exists. Current estimates are based on self-reported compliance and estimated wear and tear of the orthosis itself. Availability of an accurate and reliable technique to measure how long a patient wears a prescribed spinal orthosis will allow clinicians to objectively study the relationship between patient compliance and outcome of orthotic treatment and arrive at rational guidelines for prescribing orthosis wearing hours.

Measuring wearing compliance in an orthosis requires the use of a transducer or combination of transducers to determine if the patient is wearing the device. When choosing a transducer, cost, durability, power consumption, size and ease of use must be considered. An ideal transducer for this application would be flat and thin enough to be inserted into the orthosis between the foam and low-density polyethylene (LDPE) shell. It would be able to measure the amount of load (like a load cell); be insensitive to temperature and moisture extremes; be reliable, accurate and lightweight; require low power; and be cost-effective.

Many types of transducers can be used to determine if a person is wearing an orthosis, including temperature transducers, force transducers, pressure transducers and proximity transducers. Of these, the temperature transducer is the least reliable since temperature fluctuations due to atmospheric and environmental changes can cause inaccurate triggering. A proximity transducer can do an adequate job of determining if a patient is wearing the orthosis; however, it cannot determine if the orthosis is being worn properly.

Only force and pressure transducers can accurately determine if the orthosis is being worn properly. Pressure transducers require a gas or liquid to apply a force on the transduction element of the transducer. This commonly is accomplished with a bladder filled with a noncompressible fluid. This type of sensor system is quite accurate; however, the complexity of the system would make fabrication of the orthosis too complex. Therefore, force transducers are the only devices that can accomplish the task. Load cells can be used for this application, but they are costly and usually too large to conceal in an orthosis/prosthesis.

One transducer that matches all of the criteria is the force-sensitive resistor (FSR). The FSR can be used as a trigger or to measure the amount of load. The creep of the FSR can be a benefit as it prevents spurious triggering when the pressure applied is close to the trigger threshold. Another advantage of the FSR is the simplicity of the signal conditioning required by the device. A second resistor and voltage source are all that are required to get a voltage out of the device proportional to the force applied.

Gait Analysis

A gait analysis study is an excellent example of the integration of several transducer systems to measure an event. Gait analysis can be used to study a normal subject or patient during walking, running or standing. Quantities that can be measured include stride length, step length, cadence, limb velocity, limb acceleration and joint angular displacements.

A hypothetical experiment may use a motion analysis system to study transtibial gait patterns and the effects of different prosthetic feet. In this example, a multicomponent force transducer integrated into a walkway can give information about ground reaction forces. This force plate can be useful in measuring any asymmetry in loading of the sound limb and prosthetic limb and can provide information pertaining to the energy efficiency of particular prostheses.

After the experimental design is complete, the equipment to accomplish the gait analysis can be chosen. Of the previously discussed noncontacting displacement sensors, all have advantages and disadvantages. If an active marker system is used, the subjects must contend with wires that connect the markers to the computer control system. These wires power the infrared targets. The active markers also have a limited angle of view (+/-60 degrees) and can rotate out of view of one or more cameras. An advantage of the active marker system is it is capable of faster sampling rates than is a video-based system, and the analysis of its data is less complex.

The markers used in video-based systems are commonly spherical and generally will remain in view of the cameras for a longer period than will the active markers. Since the markers are not tethered to a control computer, motion of the subject may be less encumbered. For this hypothetical experiment, either system will accomplish the task with little difficulty. The data sampling rate of the video-based system will be sufficient due to the absence of highspeed motion. Since there are no complex motions, there is no significant advantage to the active marker system's ease of data analysis.

For the experimental design discussed here, any of the three motion analysis systems is capable of sufficient resolution. The costs of the systems range from $30,000 to almost $60,000; resolution, ease of use and flexibility increase proportionately with cost. In the end, ease of use and sophistication of data analysis software may play the most important roles in choosing a motion analysis system for an experiment.

A force transducer or force plate is an effective way to measure ground reaction forces in this experiment. These devices are available in models with one to six outputs, each corresponding to the three forces FX, FY and FZ and the three moments MX, MY and MZ. Force platforms also are available in different capacities and sizes.

For this experiment a device approximately eight to 10 inches wide and 16 to 18 inches long would be sufficient to accommodate a subject's foot during walking. Capacity should be chosen such that the maximum load received by the force transducer does not exceed its capacity in any of the six components.

During walking a subject would not exert more than two times his or her body weight on either foot. If the maximum weight of the subjects is 200 lbs, then the maximum force the transducer would receive would be 400 lbs. For this experiment, a 1,000-lb axial capacity transducer would be sufficient. However, if the experiment is altered to include sudden starts or stops, or running or jumping, one of the six components of the transducer may be exceeded. In this case, a larger-capacity transducer is required.

Summary

Experimental instrumentation of orthoses and prostheses implies the use of instrumentation systems integrated into orthoses or prostheses to measure an event and collect de sired data in an efficient, readable format. To produce accurate and reliable results, this must be done in a multidisciplinary environment by investigators trained in the use of instrumentation systems, research methodology, data collection and statistical analysis as well as orthotics and prosthetics.

However, this does not mean O&P professionals should not become familiar with basic instrumentation components that may assist them in their daily patient responsibilities. For example, an inexpensive force transducer like the FSR connected to an ohm meter can yield some valuable information regarding pressure decreases in thoracic pads of scoliosis orthoses over time. This may yield a better understanding of the time frame of viscoelastic creep of scoliotic curves and may suggest that follow-up pad tightening should occur sooner than the standard one-month interval, thus maximizing the outcome of curve correction. Load cells attached at the proximal end of a transtibial pylon and connected to a computer may help a prosthetist better understand the effects of alignment changes on load and moment on the pylon.

As clinicians, orthotists and prosthetists see repeating clinical problems, and they may realize that a routinely performed clinical method has room for improvement. The various transducers and instrumentation systems described in this article can be used as tools to objectively assess a patient's pretreatment status and the effect of an orthosis or prosthesis on the outcome. Such objective information could improve and optimize the treatment parameters.

ROBERT M. HAVEY works in the surgery at Loyola University Chicago, 21605. First Ave., Maywood, IL 60153; (708) 343-7200, ext. 3781. He also works in the Rehabilitation Research and Development Center at the VA Hospital in Hines, Ill.

THOMAS M. GAVIN, CO, is an ABC-certified orthotist at the Orthotic-Prosthetic Center of Bio Concepts Inc. in Burr Ridge, Ill. He also works in the department of orthopaedic surgery at Loyola University Chicago in Maywood, Ill., and in the orthopaedic biomechanics laboratory at the Rehabilitation Research and Development Center of the VA Hospital in Hines, Ill.

AVINASH G. PATWARDHAN, PhD works in the department of orthopaedic surgery at Loyola University Chicago in Maywood, Ill., and at the Rehabilitation Research and Development Center at the VA Hospital in Hines, Ill.

KEVIN P MEADE, PhD works in the department of mechanical, materials and aerospace engineering at the Illinois Institute of Technology in Chicago and in the Rehabilitation Research and Development Center at the VA Hospital in Hines, Ill.

References:

1. Cobbold RSC. Transducers for biomedical measurements: principles and applications. New York: Wiley, 1974.
2. Norton I-IN. Handbook of transducers. Prentice-Hall, 1989.
3. Omega Engineering Inc. The pressure strain and force handbook. P0. Box 2349, Stamford, CT 06906.
4. Murray WM, Miller WR. The bonded electrical resistance strain gage. Oxford University Press, 1992.
5. Sinclair IR. Sensors and transducers. England: Blackwell Scientific Publications, 1988.
6. Cluley JC. Transducers for microprocessor systems. Macmillan Publishers Ltd., 1985.
7. Korites BJ. Microsensors. Kern International Inc., 1987.
8. Smith K, Platts RGS, Sweeting C. Monitoring brace compliance in idiopathic scoliosis. Proceedings of the seventh world congress of ISPO, Chicago, June 28-July 3, 1992.
9. Gopel W, Hesse J, Zemel JN. Sensors: a comprehensive survey. Vol. 7. New York: VCH Publishers Inc., 1989.
10. Usher MJ. Sensors and transducers. Macmillan Publishers Ltd., 1985.
11. Sedra AS, Smith KC. Microelectronic circuits. CBS College Publishing, 1987.
12. Bachiochi J. Vaporwear: revealing your humidity. Computer Applications, April 1995;57:92-6.

Website built by oandp.com