Introduction to Biomedical Instrumentation

Download Report

Transcript Introduction to Biomedical Instrumentation

Introduction to Biomedical
Equipment Technology
Dr. T. Elsarnagawy
1
Text Books & References
Introduction to biomedical equipment
technology; J.J. Carr
Medical Instrumentation; Webster
Electronic devices; Boylestad
Dr. T. Elsarnagawy
2
Syllabus
Introduction to biomedical instrumentation &
measurement
Basic theories of measurement
Pp 26-125
Signals & noise
Electrodes, sensors and transducers
Dr. T. Elsarnagawy
3
What is biomendical engineering
It is a cross-disciplinary field that
incorporates
Engineering
Biology
Chemistry
Medicine
Biomedical instrumentation is used to take
measurements that are used in
Monitoring
Diagnostic means
Therapy
Dr. T. Elsarnagawy
4
Fields of biomedical engineering
Bioinstrumentation
Applies the fundamentals of measurement science to biomedical
instrumentation
Emphasizes the common principles with making measurements in
living cells
Biomaterials
Application of engineering materials in production of medical devices
Biomechanics
Behavior of biological tissues and fluids
Ergonomics (design principles)
Biosignals
The mechanisms of signal production
Fundamental origins in of the variability in the signal
Rehabilitation engineering
Design of equipments for disabled individuals
Dr. T. Elsarnagawy
5
Scientific Notation
The form of a number in scientific notation:
N X 10x {Unit}
N: Numbers
10: Base
x: Exponent
Never forget to write the UNIT ……if it exists
10-x  1/10x
Prefixes:
Nano-, micro-, milli-, centi-, …, kilo-, mega-, giga-, tera10-9 …………………………………………………………………………………..1012
Dr. T. Elsarnagawy
6
Metric Prefixes
Symbol
Name
Multiplication
p
pico
1 x 10-12
n
nano
1 x 10-9
μ
micro
1 x 10-6
m
milli
1 x 10-3
k
kilo
1 x 103
M
Mega
1 x 106
G
Giga
1 x 109
T
Tera
1 x 1012
Dr. T. Elsarnagawy
7
UNITS AND PHYSICAL
CONSTANTS
Dr. T. Elsarnagawy
8
SI Units
The standard unit system for medical,
engineering and scientific practice is taken
from the SI (Systeme Internationale)
CGS or MKS (also called metric
system)
SI depends on multiplying prefixes in the
basic units (see metric prefixes table)
Dr. T. Elsarnagawy
9
Conversion to SI units
Dr. T. Elsarnagawy
10
Conversion from SI units
Dr. T. Elsarnagawy
11
standard physical units
12
Physical constants
Dr. T. Elsarnagawy
13
Definitions
Measurand (Physical quantities):
Position, displacement
Temperature
Force
Pressure,…
Concentrations, chemicals,…,
Sensor:
is a device that detects a change in a physical stimulus
and turns it into a signal which can be measured or
recorded
Signal conditioning:
Amplifying, waveshaping, filtering, rectifying,…
A Transducer
is a device that transfers power from one system to
another in the same or in a different form.
Dr. T. Elsarnagawy
14
Common medical measurands
The measurand is the measured quantity
Dr. T. Elsarnagawy
15
Generalized Instrumentation system
Dr. T. Elsarnagawy
16
Instrumentation System
A Measuring system is required to compare a quantity with a
standard or to provide an output that can be related to the quantity
being measured
The quantity to be measured is detected by an input transducer or
sensor.
The detected quantity may be converted to a mechanical or
electrical form of energy
Output
Input
Sensor
Signal
conditioner
Measurand
Display
Recorder
Dr. T. Elsarnagawy
17
Medical Measurement Chain
surface electrode
pressure transducer
photocoupler
temperature sensor
pressure gauge
strain gauge
:
Sensor
A/D
Converter
Process
Circuit
LCD
Oscilloscope
Clinical Instrument
ECG
Instrument
EMG
Instrument
...... Blood Pressure
Dr. T. Elsarnagawy
Instrument
18
Generalized Instrumentation System
Dashed lines are optional for some application
Dr. T. Elsarnagawy
19
“Averages”
in Biomedical Engineering
Dr. T. Elsarnagawy
20
Types of Averages
Definition
Most typical value or most expected value in a
collection of numerical data
Different kinds of average
Mean (arithmetic mean):
The sum of all values xn divided by the number n of
different values
Ex.:
mean average???
sum=125, n=28  Xmean=125/28=4.46
Dr. T. Elsarnagawy
21
Types of Averages
Median:
The middle value in
a data set
Mode:
The most frequently
occurring value in a
data set
If data is perfectly
symmetrical  ??
Which average is the best to use for which
kind of data??
If data is symmetrical  use mean average
If data is highly asymm. (outliers)  median
If you need an answer to a question  mode
Ex.: most common cause of death, or most popular TV
show on Friday,…
Other types of averages:
Geometric average  biological studies
Harmonic mean (H.M.)  when data are expressed in
ratios (miles/hr, riyal/dozen,…)
Geometric average
Ex.: if you have 48$, spend
half of your available money
each day for 5 days.
Arithm. Mean=
(48+24+..)/5=18.6
Geometric average
To find the Geometric average
To straighten the curve  semilog paper
Harmonic mean (H.M.)
Is used when data is expressed in ratios
(miles/hrs, riyals/dozen,…)
The expression of H.M.
Harmonic mean: example
Integrated Average
This average is applied
often in RC circuits
The area under the curve
of a time dependent
function divided by the
segment of the range
over which the average is
taken
V
V1
The output of the circuit ~
time average of the input
signal
Dr. T. Elsarnagawy
T
0
t1
t2
t
28
Root-mean-square “rms”
 Used in electrical circuits and other technologies
e.g.when comparing AC sine wave
current with DC current the AC should be
expressed in an equivalent value which
is the rms.
 Definition of rms:
Vrms: is the rms value
T: is the time interval t1 to t2
V(t): is the time-varying voltage function
 Special case: the rms value of a sine
wave voltage is Vp/√2 or 0.707 Vp (Vp is
the peak voltage)
Dr. T. Elsarnagawy
29
Root sum sqaure “rss”
Logarithmic Representation of signal Levels
“Decible Notation dB”
For voltage or current measurements
Original unit was “bel”
The prefix “deci” means
one tenth
Hence, the “decible” is
one tenth of a “bel”
dB expresses
logarithmically the ratio
between two signal levels
For power measurements
(ex.: Vo/Vi = Gain)
dB is dimensionless
Review table 3-8 page 37 in IBET
Dr. T. Elsarnagawy
31
Common dB scales in electronics
dBm, dBmV
dBm: 0 dBm refers to an input power of 1mW
dissipated in 50Ω resistive load
What is the signal level 9mW as expressed in dBm?
dBm = 10 log (9mV/1mW) = 9.54 dBm
Express a signal level of 800 μV in dBm
Use P=V2/R
=0.00000064V/50Ω
=0.0000128mW
 dBm = 10log(P/1mW)= -48.9
Review dBmV and examples page 38,39 in IBET
Dr. T. Elsarnagawy
32
The basic equations to calculate
decibels (Logarithm)
Iin
Io
Vin
Vo
Pin
Po
Dr. T. Elsarnagawy
33
Calculation of the overall strength of a system and
calculating the system gain
V1
0.2
A1 

 20
Vin 0.01
V2 0.1
Atten 

 0.5
V1 0.2
AV  A1  Atten  A2  A3  600
V3 1.5
A2 

 15
V2 0.1
AdB  20 log 600  55.6
Vo
6
A3 

4
V3 1.5
Dr. T. Elsarnagawy
34
Converting between dB and Gain notation
For dB = 20 log (Vo/Vin)
if it is needed to convert from
dB to output-input ratio i.e.
Vo/Vin
 Vo = Vin 10dB/20
or Vo = Vin EXP(dB/20)
Vin
1 mV
Av=20dB
Vo
?
 Ex: calculate the output
voltage Vo if the input voltage
Vin=1mV and an amplifier of
+20 dB is used:
 Vo=(0.001V) 10(20/20)
=(0.001) (10) = 0.01V
Dr. T. Elsarnagawy
35
Special decibel scales: dBm
dBm: used in reference frequency
measurements (RF)
0 dBm is defined as 1 mW of RF signal
dissipated in 50-Ω resistive load
dBm = 10 log (P/1 mW)
 EX: What is the signal level 9 mW as
expressed in dBm?
dBm = 10 log (P/1 mW)
dBm = 10 log (9 mW/1 mW) = 9.54 dBm
Dr. T. Elsarnagawy
36
Data Classes
Qualitive
Nonnumerical or categorical (includes the
presence or nonpresence of some factor,
good or bad, defective or not defective,
gender …)
Not inherently with numbers
Can be given a numerical flavor (1 or 0, yes
or no)
Sometimes we assign some kind of scale
Data Classes
Quantitive
Naturally result in some number to represent a factor
(amount of money, length, temperature, voltage,
pressure, weight …)
Interval: referenced to a selected standard zero point (ex.:
calendar is referred to date of birth of Christ or Hijra,
temperature C is referred to the freezing or boiling point of
water) note: centigrade: centi=100 (0-100 divisions from the arbitrarily set 0C to
100C)
Ratio: fixed to a natural zero point, such as weights,
pressure, temperature (Kelvin) referred to the absolute zero
(0 K) at which molecular motion ceases (-273.16C)
Variation and error
Variations (or random variation) are caused by
certain errors in the measurement process.
Caused by type of meter used
Caused by variation in the process being measured
Random variation causes data obtained to
disperse
how to represent this dispersion?
Histogram, normal distribution 
Variation and error: Histograms &
Normal distribution (Gaussian curve)
Data represented in fig.a
 Histogram
Data represented in fig.b
 normal distribution
(Gaussian)
Set of data:
Variance & Standard deviation
The normal distribution gives a measure
of data dispersion
Dispersion of data is summed up as
variance and standard deviation of the
data
Variance:
Standard deviation:
In case of small data sets
X
: the mean
Accuracy of a measurement is indicated by
the size of ΔX
Xi: true value
X0: central value of successive measurements
ΔX: Error
As ΔX  0 then X0  Xi
ΔX
Y
Xi
X0
Dr. T. Elsarnagawy
X
42
Basics of measurements
Before we begin our look at biomedical
instrumentation, we need to study
some general characteristics of
instrumentation
System Characteristics
Specific ch/cs
General ch/cs
Specific Characteristics for a system
Specs for specific biomedical instrumentation as
determined by the committee ………… ex: ECG
ECG specifications
Some specific Characteristics
For example
Dynamic range:
Given is the input
dynamic range -5mV
to +5mV
If input signal exceeds
the dynamic range so
it will cause an error
The amplified signal is
then called to be
saturated
Some specific Characteristics
DC offset
Is the amount the
signal may be moved
from its baseline and
still be amplified
properly by the system
Without DC offset
With DC offset
Some specific Characteristics
Slew rate
Maximum rate at which
the system can observe
a changing voltage per
unit time
If the input signal exceed
the given slew rate the
output will be distorted
Frequency response
The range of frequencies
of the measurand the
system can handle
General characteristics
These are
characteristics all
systems share
Linearity
Analog or digital
system
Significant factors in
measurements
Measured/
Calibration curve
O/p
Max deviation
Idealized curve
(linear fit)
I/p
Dr. T. Elsarnagawy
51
Accuracy
Closeness to the true value of measurand
1.
Dr. T. Elsarnagawy
52
2.
Precision
a measure of the
degree of
agreement within a
group of
measurements –
repeatability of a
system- (however
no guarantee of
accuracy)
Results have Low
scatter  excellent
precision
Dr. T. Elsarnagawy
53
3. Sensitivity
Relation between
change in output for a
given change in input
(scale factor,
magnification).
The relation may be
linear or nonlinear
Dr. T. Elsarnagawy
54
How to calculate Sensitivity (S)
O/p
linear
O/p
Non-linear
I/p
I/p
S:sensitivity=ΔO/p/ΔI/p
Inverse Sens.=1/S
Dr. T. Elsarnagawy
55
Linearity:
An instr. is said to be linear when incremental
changes in input and output are constant
over the specified range
(i.e. Output in lin. prop to the input)
4.
Dr. T. Elsarnagawy
56
Resolution
Smallest i/p increment change that gives some change
in the o/p
5.
Example:
Voltmeter scale with 100 divisions FS=200V, 1/10 of
scale division can be estimated  determine the
resolution?
Solution:
1 division=200/100=2V
Resolution=1/10 scale division=1/10x2=0.2V
Dr. T. Elsarnagawy
57
Threshold
Minimum input value below which no output can be detected.
6.
7. Hystresis
Tendency for indications on an upward cycle to differ from
the same points on the downward cycle
Causes: Friction, relaxation
Numeric value of Hyster.: % of full scale
Dr. T. Elsarnagawy
58
8.Drift:
Variation in output without change in input
(...Temp. Changes or component instability)
9. Zero Stability
Ability to return to zero when measurand = 0
Dr. T. Elsarnagawy
59
All together
Dr. T. Elsarnagawy
60
10. Dynamic range:
Rdyn=Ymax-Ymin
Given is the input dynamic range 5mV to +5mV
If input signal exceeds the
dynamic range so it will cause an
error in the output
The amplified signal is then called
to be saturated
Dr. T. Elsarnagawy
61
11. DC offset
Is the amount the
signal may be moved
from its baseline and
still be amplified
properly by the system
Without DC offset
With DC offset
Dr. T. Elsarnagawy
62
13. Frequency response
characteristics
The range of frequencies of
the measurand the system
can handle
Wideband
Band-pass
--- Phase distortion
Typical for sensors
Dr. T. Elsarnagawy
63
12. Slew rate
Maximum rate at which the system can
observe a changing voltage per unit time
If the input signal exceed the given slew rate
the output will be distorted
Dr. T. Elsarnagawy
64
‫المعـــايره ‪Calibration‬‬
‫‪Measurements for calibration means‬‬
‫‪ - Creep‬قياسات اإلنحراف عن المنحنى الخطى‬
‫‪ - Temperature‬قياسات الحراره‬
‫‪Hystresis‬‬
‫‪ - Zero drift‬اإلنحراف عن الصفر‬
‫‪ - Zero error‬خطأ الرجوع إلى الصفر (نقطه البدايه)‬
‫‪ - Reproducability‬خطأ تكرار القياسات‬
Calibration procedure
Calibration is used to detect the errors in a sensor
 Correction if possible
Sensor
Transducer
Calibration
measurements
Reference standard
Assignment for next week
Measurement errors
Describe the four general categories of error
Dealing with measurement errors
Signals
Sinusoidal waveform
Dr. T. Elsarnagawy
69
Dr. T. Elsarnagawy
70
Types of signal
a. Static: dc
b. Quasistatic
c. Periodic: sine, square,…
v(t)=v(t+T)
d. Repetitive: quasiperiodic
e. Single event transient signal
f. Repetitive single event
Dr. T. Elsarnagawy
71
Waveform symmetry
Dr. T. Elsarnagawy
72
Signal sampling
Most instrumentation transducers have analog output
At the interface between analog transducers and digital
computers the signal must be digitized
So the signal is sampled at regular intervals
Each sample voltage is then converted into an
equivalent digital value
The next sample cannot be taken until the conversion of
the last sample is to digital form is completed
Dr. T. Elsarnagawy
73
Sampled signals
Dr. T. Elsarnagawy
74
Effect of the sampling rate
12 sample/sec
1 Sample/sec
If fsampling > fsignal  o.k.
Ideally fsampling = 2 fsignal
If fsampling < fsignal  aliasing
Dr. T. Elsarnagawy
75
To reconstruct the original signal after
sampling  pass the sampled waveform
through a low-pass filter that blocks fs
Sampling is used to form
AM, PM,
Some applications don’t accept
fsampling=2fsignal as in ECG  =5fsignal
Dr. T. Elsarnagawy
76
Essential Electronics Formula
Dr. T. Elsarnagawy
77
Essential Electronics Formula
Ohm's Law
The first of these is Ohm's Law, which
states that a voltage of 1V across a
resistance of 1 Ohm will cause a current of
1 Amp to flow. The formula is
R=V/I
(where R = resistance in Ohms, V = Voltage
in Volts, and I = current in Amps)
V=R*I
I=V/R
Dr. T. Elsarnagawy
78
Reactance
The impedance (reactance) of a capacitor,
which varies inversely with frequency (as
frequency is increased, the reactance falls and
vice versa).
XC = 1 / (2 Π f C)
where Xc is capacitive reactance in Ohms, (Π pi) is 3.14159,
f is frequency in Hz, and C is capacitance in Farads.
Inductive reactance, being the reactance of an
inductor. This is proportional to frequency.
XL = 2 Π f L
where XL is inductive reactance in Ohms, and L is inductance
in Henrys
Dr. T. Elsarnagawy
79
Decibels (dB)
dB = 20 log (V1 / V2)
dB = 20 log (I1 / I2)
dB = 10 log (P1 / P2)
Either way, a drop of 3dB represents half
the power and vice versa.
Dr. T. Elsarnagawy
80
Frequency
There are many different calculations for this, depending
on the combination of components.
The -3dB frequency for resistance and capacitance (the most
common in amplifier design) is determined by
fo = 1 / (2 Π R C)
where fo is the -3dB frequency
When resistance and inductance are combined, the
formula is
fo = R / (2 Π L)
Dr. T. Elsarnagawy
81
Power
Power in any form can be calculated by a
number of means:
P=V I
P = V2 / R
P = I2 R
where P is power in watts, V is voltage in Volts,
and I is current in Amps.
Dr. T. Elsarnagawy
82