Medical Measurement Constraints The amplitude and frequency ranges for each parameter are the major factors that affect the design of all instrument components Proper measurand-sensor interface cannot be obtained Medical variables are seldom deterministic External energy must be minimized to avoid any damage Nearly all biomedical measurements depend on some form of energy being applied to the living tissue or to the sensor, for example: X-ray, ultrasonic imaging and electromagnetic or Doppler ultrasonic Blood flow-meters depend on externally applied energy interacting with living tissue Safe level of applied energy is an important consideration Equipment must be reliable
Additional Medical Measurement Constraints Reliable Simple to operate Withstand physical abuse and exposure to corrosive chemicals Electrical safety (minimize electric shock hazard)
Classification of Biomedical Instruments Quantity being sensed: Pressure, flow, temperature, potential, etc. Advantage: easy comparison of different methods for measuring any quantity Principle of transduction: Resistive, capacitive, inductive, ultrasonic or electrochemical Advantages: a. different applications of each principle can be used to strengthen understanding of each concept b. newer applications readily apparent Measurement technique for each physiological system: Cardiovascular, pulmonary, nervous, endocrine Advantage: isolates all important measurements for specialists Disadvantage: considerable overlap of quantities sensed and the principles of transduction used
Classification of Biomedical Instruments... Clinical medicine specialties: Pediatrics, Obstetrics, Cardiology, Radiology, etc. Advantage: valuable for medical personnel interested in specialized instruments
Interfering and Modifying Inputs Desired input: the measurand that the instrument is designed to isolate and measure Interfering inputs: quantities that inadvertently affect the instrument as a consequence of the principles used to acquire & process the desired inputs Modifying inputs: undesired quantities that indirectly affect the output by altering the performance of the instrument itself Modifying inputs can affect processing of either desired or interfering inputs Some undesirable quantities can act as both a modifying input and an interfering input
Example A simplified ECG recording system provides a good example: In this recording system: The desired input is: V ecg – electrocardiographic voltage between 2 electrodes (RA & LA) The interfering inputs are: 50 Hz or 60 Hz (power-line) noise voltage induced in the shaded loop by ac magnetic fields Also the difference between the currents running through each of the electrodes to the patient and to the ground causes a voltage on Z body
Example... In ECG, the example of a modifying input is the orientation of the patient cables. If the plane of the cable is parallel to the ac magnetic field, magnetically introduced interference is zero. If the plane of the cables is perpendicular to the ac magnetic field, magnetically introduced interference is maximal. Time–dependent changes in electrode impedance Electrode motion
Elimination of Interfering and Modifying Inputs To reduce or eliminate the effects of most interfering and modifying inputs we have two alternatives: 1. Alter the design of essential instrument components (preferred, but hard to achieve) 2. Add new components to offset the undesired inputs
Compensation Methods 4 Types 1. Inherent sensitivity All components only respond to desired inputs: make G i and G m,d = 0 In the ECG recording problem, Figure 1.3 of Webster: twist the electrode wires: this makes G i →0 (remember from Physics: v in ~B (magnetic field density). A (area subtended by the mag. Field). N (number of wire turns) By twisting the wires we make A→0! Therefore v int = 0! Minimize electrode motion → this makes G m,d →0 2. Negative feedback – when the modifying input can not be removed or avoided, we should – minimize the effects of the transfer function of the instrument system on the output.
Compensation... Filtering: For example a notch filter can remove the 60Hz component of the inputs signal. A filter is a device or program that separates data or signals based on a specific criteria. Filtering is an extensive topic. Opposing inputs: Make an additional input that opposes the unwanted input. Example: We can use thermistors (temperature dependent resistors) to counteract the unavoidable temperature-dependent changes in the characteristics of the active elements of the circuits.
Sensor characteristics Static characteristics The properties of the system after all transient effects have settled to their final or steady state Accuracy, Discrimination, Precision, Errors, Drift, Sensitivity, Linearity, Hysteresis (backslash) Dynamic characteristics The properties of the system transient response to an input Zero order systems First order systems Second order systems
Accuracy & discrimination Accuracy is the capacity of a measuring instrument to give RESULTS close to the TRUE VALUE of the measured quantity Accuracy is related to the bias of a set of measurements (IN)Accuracy is measured by the absolute and relative errors More about errors in a later Discrimination is the minimal change of the input necessary to produce a detectable change at the output Discrimination is also known as RESOLUTION When the increment is from zero, it is called THRESHOLD
Precision The capacity of a measuring instrument to give the same reading when repetitively measuring the same quantity under the same prescribed conditions Precision implies agreement between successive readings, NOT closeness to the true value Precision is related to the variance of a set of measurements Precision is a necessary but not sufficient condition for accuracy Two terms closely related to precision Repeatability The precision of a set of measurements taken over a short time interval Reproducibility The precision of a set of measurements BUT taken over a long time interval or Performed by different operators or with different instruments or in different laboratories
Example Shooting darts Discrimination The size of the hole produced by a dart Which shooter is more accurate? Which shooter is more precise? Shooter A Shooter B
Accuracy and Errors Systematic errors Result from a variety of factors Interfering or modifying variables (i.e., temperature) Drift (i.e., changes in chemical structure or mechanical stresses) The measurement process changes the measurand (i.e., loading errors) The transmission process changes the signal (i.e., attenuation) Human observers (i.e., parallax errors) Systematic errors can be corrected with COMPENSATION methods (i.e. feedback, filtering) Random errors Also called NOISE: a signal that carries no information True random errors (white noise) follow a Gaussian distribution Sources of randomness: Repeatability of the measurand itself (i.e., height of a rough surface) Environmental noise (i.e., background noise picked by a microphone) Transmission noise (i.e., 60Hz hum) Signal to noise ratio (SNR) should be >>1 With knowledge of the signal characteristics it may be possible to interpret a signal with a low SNR (i.e., understanding speech in a loud environment)
More static characteristics Input range The maximum and minimum value of the physical variable that can be measured (i.e., - 40F/100F in a thermometer) Output range can be defined similarly Sensitivity The slope of the calibration curve y=f(x) An ideal sensor will have a large and constant sensitivity Sensitivity-related errors: saturation and “dead-bands” Linearity The closeness of the calibration curve to a specified straight line (i.e., theoretical behavior, least-squares fit) Monotonicity A monotonic curve is one in which the dependent variable always increases or decreases as the independent variable increases Hysteresis The difference between two output values that correspond to the same input depending on the trajectory followed by the sensor (i.e., magnetization in ferromagnetic materials) Backslash: hysteresis caused by looseness in a mechanical joint