Biomedical Instrumentation

BME-301

Lecture -3

Dr. Muhammad Umair Ahmad Khan

[email protected]

11-09-2024
 Problems encountered in measuring
biopotentials of the human body

Biopotentials are electrical potentials that arise from biochemical processes

in the human body. Commonly measured biopotentials include the
electroencephalogram (EEG) for brain activity, the electrocardiogram (ECG or
EKG) for heart activity, the electromyogram (EMG) for muscle activity, and
the electrooculogram (EOG) for eye movement. Despite the advancement in
technologies to measure these potentials, several challenges persist:
1.Artifacts and Noise:
1. Motion Artifacts: Movements can introduce significant noise in the
recorded signals. For instance, adjusting the position during EEG or
ECG can produce artifacts.
2. Electrical Interference: External electrical devices or power lines can
interfere with the recordings.
3. Biological Artifacts: Activities such as swallowing, blinking, or sweating
can interfere with measurements.
 Problems encountered in measuring
biopotentials of the human body
2 Skin-Electrode Interface:
1.Impedance: The electrical contact between the skin and the
electrode can vary, leading to varying resistance which can
distort the signals.
2.Skin Preparation: The skin often needs to be cleaned or
slightly abraded to reduce impedance, which can be
uncomfortable.
3.Electrode Gel Degradation: Over time, the gel or solution
used to ensure good contact can dry out or degrade,
affecting the signal quality.
 Problems encountered in
measuring biopotentials of the
human body

3. Reference Electrode Placement: The choice and placement of

reference electrodes can influence the measured potentials,
and sometimes it's challenging to decide the optimal
placement.
4.Spatial Resolution: Electrodes measure the potential from a
large number of cells. Distinguishing activity from closely located
sources can be challenging, especially in EEG.
5.Temporal Resolution: While biopotentials provide excellent
temporal resolution, the delay introduced by some equipment
or signal processing techniques can sometimes be an issue,
especially for real-time applications.
 Problems encountered in
measuring biopotentials of the
human body

6.Subject Comfort: Especially for long-term monitoring, the

comfort of the subject is a concern. Adhesive electrodes, wires,
and the associated equipment can be cumbersome.8.Electrode
7.Drift: Over extended periods, the characteristics of electrodes
can change, leading to a drift in the measured signal.
8.Signal Processing Challenges: Techniques like filtering,
amplification, and analog-to-digital conversion introduce their
own sets of problems, such as phase shifts or aliasing.
 Problems encountered in
measuring biopotentials of the
human body

9. Inter-individual Variability: The morphology and amplitude of

biopotentials can vary widely between individuals, making it
challenging to establish standard benchmarks or thresholds.
10.Invasiveness: For some measurements, like intracranial EEG, the
electrodes need to be implanted, introducing risks of infection, tissue
damage, or other complications.
11.Cost: High-quality equipment and disposable supplies, like certain
types of electrodes, can be expensive.
12.Calibration and Maintenance: Regular calibration and
maintenance of equipment are required to ensure accurate
measurements.
13.Data Storage and Management: Continuous long-term recording
can generate large amounts of data, presenting challenges in storage,
management, and analysis.
 Statistical Control

Errors in measurements can be broadly

categorized into two types: systematic
errors and random errors. Each type of
error has distinct characteristics and
sources, and understanding the difference
between them is crucial for ensuring the
accuracy and precision of measurements.
 Statistical Control

1.Systematic Errors:
1. Definition: Systematic errors are consistent, repeatable errors that are
introduced by a particular factor or factors that skew all measurements
in a specific direction (either too high or too low).
2. Causes:
1.Instrumental bias (e.g., a scale that always reads 0.5 kg more than
the actual weight).
2.Flawed experimental design or procedure.
3.Human bias in reading instruments or taking measurements.
4.Environmental conditions that aren’t accounted for (e.g., a
thermometer that isn’t properly shielded from radiant heat sources).
5.Incorrect calibration of equipment.
 Statistical Control
3. Characteristics:
1. They are consistent and have the same magnitude and direction.
2. They can be identified and corrected.
3. If not detected, they can seriously jeopardize the validity of the
experimental results.
4. Correction: Systematic errors can often be minimized by calibration of
the measurement instrument, using alternative measurement
techniques, or by compensating for the known sources of errors.
 Statistical Control

1.Random Errors:
1. Definition: Random errors are statistical fluctuations (in either
direction) in the measured data due to the precision limitations of
the measurement device.
2. Causes:
1.Unpredictable fluctuations in readings.
2.Inherent limitations in the measuring instrument.
3.External environmental effects like sudden temperature changes.
4.Human errors that occur randomly, like momentary lapses in
attention.
 Statistical Control

3.Characteristics:
1.They have no consistent pattern (i.e., sometimes too high,
sometimes too low).
2.They tend to average out over multiple measurements.
3.Their effects can be reduced, but not entirely eliminated, by taking
multiple measurements and then averaging the results.
4.Correction: Random errors cannot be eliminated but can be reduced by
improving experimental design and technique. Statistical methods, such
as standard deviation and error propagation, can be used to estimate the
uncertainty introduced by random errors.
 Statistical Control
o Any systematic errors or bias can be
removed by calibration and correction
factors, but random variations pose a
more difficult problem.
o Random errors makes output
unreproducible.
o If the cause of variability cannot be
eliminated, statistical analysis must be
done.
o Make multiple measurements and
averaging the results can improve the
estimate of the true value.
 Statistical Control
o Systematic errors Result from a variety of factors
n Using the instrument at atemperature other than the
calibrated temperature
n Aging of the components will change the sensor response
and hence the calibration
n The measurement process changes the measurand (i.e.,
loading errors)
n Signal path of the measurement process
n Human observers (i.e., parallax errors)

o Systematic errors can be corrected with

COMPENSATION methods (i.e., feedback,
filtering)
o Random errors are the noise (ideally Gaussian)
left after systematic errors are removed. Sources
of randomness:
n Repeatability of the measurand itself (i.e., height of a rough
surface)
n Environmental noise (i.e., background noise picked by a
microphone)
n Transmission noise (i.e., 60Hz hum) before amplifier.
n SNR of amplifier
 Calibration
o The relationship between the
physical measurement variable input
and the signal variable (output) for a
specific sensor is known as the
calibration of the sensor.
o Typically, a sensor (or an entire
instrument system) is calibrated by
providing a known physical input to
the system and recording the output
as shown in Figure.
o The sensor has a linear response for
values of the physical input less than
X0.
o The sensitivity of the device is
determined by the slope of the
calibration curve.
 Calibration
o In this example, for values of
the physical input greater
than X0, the calibration
curve becomes less
sensitive until it reaches a
limiting value of the output
signal, referred as
saturation.
o In some cases, the sensor will
not respond to very small
values of the physical input
variable.
o The difference between the
smallest and the largest
physical inputs that can
reliably be measured by an
instrument determines the
dynamic range of the
device.
 Calibration
Calibration in medical instrumentation refers to the process of setting or
correcting the performance of a medical instrument or device to ensure
that its readings or outputs are accurate, consistent, and in accordance
with a defined standard. This is crucial to ensure the reliability of
diagnostic and therapeutic procedures, and patient safety.

Calibration often involves comparing the readings of the device in

question to those of a reference or standard device, known for its
accuracy. If discrepancies exist, adjustments are made to the device under
test to align its performance with the standard.

Blood Pressure Monitor:

One common medical instrument is the electronic blood pressure (BP)
monitor. Calibration of a BP monitor ensures that the readings it provides
(systolic and diastolic pressures) are accurate.
 Calibration
Calibration Process:

A reference manometer, known to provide accurate pressure

readings, is connected in parallel with the electronic BP monitor.
A test is initiated, and the pressure is increased incrementally.
At each increment, readings from the BP monitor and the reference
manometer are recorded and compared.
If the readings from the BP monitor consistently differ from the
reference manometer by an unacceptable margin, the BP monitor
may require recalibration or servicing.
Once calibrated, the monitor should read blood pressures
accurately, ensuring that patients and clinicians receive reliable data
for diagnosis and monitoring.
Zero Drift
o Zero drift in the context of medical instrumentation refers to
the unwanted and consistent deviation or change of an
instrument's output or reading over time when the
actual input (or the quantity being measured) remains
at zero or at a fixed value. It's important to note that drift
is usually not due to the immediate application of a stimulus
but rather due to factors like changes in ambient conditions,
component aging, or other long-term instabilities.
o Causes
n Manufacturing misalignment
n Variations in ambient temperature
n Hysteresis
n Vibration
o Offsets are easy to correct with a single-point
calibration.
 Zero Drift
o Pulse Oximeter:
o A pulse oximeter is a medical device used to measure the oxygen
saturation level in a patient's blood non-invasively.

o Imagine a scenario where a pulse oximeter, when placed on a person's

finger, should ideally show an oxygen saturation level of 98%. However,
due to zero drift, even when not placed on a finger (and thus not
receiving any valid input), the device shows a reading of 2%. This
means when it's actually used on a patient with 98% saturation, the
reading might show 100% due to this drift. This can lead to a
misinterpretation of a patient's oxygen levels.

o Such inaccuracies, especially in critical medical instrumentation, can

lead to inappropriate medical decisions. Therefore, recognizing,
correcting, and minimizing drift is crucial in the field of medical
instrumentation. Regular calibration and maintenance of instruments
help in addressing these issues.
One Point Calibration
o One-point calibration is a method used to adjust
an instrument so that its output corresponds
accurately to a single known input value. It's a
simple calibration method, and its accuracy
assumes that the instrument's response is linear
across its entire range.
o For example, thermocouples used at very high
temperatures exhibit an 'aging' effect. This can
be detected by performing periodic one point
calibrations, and comparing the resulting offset
with the previous calibration.
o To perform a one point calibration:
n 1. Take a measurement with your sensor.
n 2. Compare that measurement with your
reference standard.
n 3. Subtract the sensor reading from the
reference reading to get the offset.
n 4. In your code, add the offset to every sensor
reading to obtain the calibrated value.
One Point Calibration
o Example: Imagine that you have a
competition robot that needs to position
itself exactly 6" from a goal in preparation
for scoring. You have an ultrasonic
rangefinder for your distance sensor.
Since you only require maximum accuracy
at one distance, a one point calibration is
a simple and effictive solution.
n Using a measuring tape as your reference
standard, position the robot exactly 6" from
the goal.
n If you take a reading with your sensor and
it says 6.3", then you have a -0.3" offset.
n Now edit your code to subtract 0.3" from
every reading. Since this is known to be a
linear sensor it will likely be pretty accurate
over most of its range. But you know with
n great confidence that it will be spot-on at
the critical distance of 6".
 Sensistivity Drift
o Sensistivity drift which is changes in the slope of the calibration
curve occurs due to interfering and/or modifying input.
o Sensitivity drift causes error that is proportional to the
magnitude of input
o The slope of the calibartion curve can either increase or
decrease.
o Causes
n Manufacturing tolerance
n Variations in power supply
n Nonlinearities
n Change in ambient temperature and pressure
o Example: variations of ECG amplifier gain due to the fluctuation
of dc power supply voltage or temperature
o The Two-point calibration process can correct
differences in slope.
 Sensistivity Drift
Sensitivity drift refers to the change in the sensitivity of a measuring instrument
over time. Sensitivity, in this context, is the ratio of the change in the output or
response of an instrument to a change in the input or quantity being measured.
When the sensitivity changes (increases or decreases) without any intentional
adjustment, we refer to this as sensitivity drift.

Factors that can contribute to sensitivity drift include component aging, changes
in ambient conditions (e.g., temperature, humidity), mechanical wear, and long-
term instabilities in electronic components.

Implications:
Sensitivity drift can lead to inaccurate measurements. Even if the instrument is
correctly calibrated at a single point, the changed sensitivity can cause it to
respond differently to other input levels than it would have when it was initially
calibrated.
 Sensistivity Drift
ECG (Electrocardiogram) Machine:
An ECG machine measures the electrical activity of the heart over time. Let's say
the machine is designed such that a 1-mV change in heart electrical activity leads
to a 1-cm movement of the ECG pen on paper (or equivalent display change in
digital devices).

Over time, due to sensitivity drift, a 1-mV change in heart electrical activity might
now result in a 0.8-cm movement (reduced sensitivity) or a 1.2-cm movement
(increased sensitivity) of the ECG pen. This would lead to misrepresentations of
the heart's electrical activity, which could potentially lead to misdiagnoses.

Regular calibration, maintenance, and checking against known standards can help
in identifying and correcting sensitivity drift, ensuring that medical instruments
remain accurate and reliable.
Two Point Calibration
o A Two Point calibration essentially rescales the
output and is capable of correcting both slope and
offset errors. Two point calibration can be used in
cases where the sensor output is known to be
reasonably linear over the measurement range.
o To perform a two point calibration:
n 1. Take two measurement with your sensor: One near
the low end of the measurement range and one near the
high end of the measurement range. Record these
readings as "RawLow" and "RawHigh"
n 2.Repeat these measurements with your reference
instrument. Record these readings as "ReferenceLow"
and "ReferenceHigh"
n 3. Calculate "RawRange" as RawHigh – RawLow.
n 4.Calculate "ReferenceRange" as ReferenceHigh –
ReferenceLow
n 5.In your code, calculate the "CorrectedValue" using the
formula below:
n CorrectedValue = (((RawValue – RawLow) *
ReferenceRange) / RawRange) + ReferenceLow
Two Point Calibration
Two-point calibration is a method used to adjust an instrument so
that its output corresponds accurately to two known input values.
This calibration technique assumes that the instrument's response
between these two points is linear, allowing for a more reliable
calibration across a range of measurements compared to a one-
point calibration.

In a two-point calibration, the instrument is exposed to two

standards or reference values. After noting the instrument's
response at these points, a linear relationship (line of best fit) is
established between the input and output values.
 Linearity
o A system is linear if following
requirements are satisfied.
o Linearity eliminates the
need to do any complex
curve- fitting and
simplifies the calibration
process.
o High accuracy does not
necessarily implies
linearity. A sensor can
be highly accurate but
also have a totally non-
linear output function,
and sensors with very
linear outputs can be
highly inaccurate.
 Linearity
o In practice, no instrument has a
perfect linear response (Fig. a), so
a measure of deviation from
linearity is needed.
Fig. a

Linearity refers to the ability of

a system or instrument to
produce outputs that are
directly proportional to the
input values. If an instrument's
response is linear, then a
change in the input will result
in a consistent proportional
change in the output across the
entire measurement range.
Fig. b
Biosensor
A biosensor is an analytical device that combines
a biological component with a physicochemical
detector component to detect specific
biochemical substances. The biological
component can be an enzyme, antibody,
microorganism, or nucleic acids, among others.
The detector component, on the other hand, can
be optical, piezoelectric, electrochemical, or
thermometric, depending on the specific
application of the biosensor.
The big question

WHY HEALTHCARE IS SO EXPENSIVE

WHY WE ARE NOT ABLE TO CATCH
DISEASES
DIFFICULTY OF REAL
TIME MOLECULAR
MEASUREMENTS
Biosensor
 Key components of a
biosensor:
Biological Recognition Element: This is the part of the biosensor
responsible for recognizing the specific target analyte. It could be:
Enzymes
Antibodies or antigens
Nucleic acids (like DNA strands for specific sequences)
Tissues or cells
Receptors
Transducer Element: This component translates the interaction between
the analyte and the biological element into a measurable signal. The
transduction method can be:
Electrochemical (measuring changes in current, potential, or impedance)
Optical (measuring changes in light absorbance, fluorescence, or refractive
index)
Thermal (measuring heat changes)
Piezoelectric (measuring changes in resonance frequency)
Key components of a
biosensor:
Signal Processor: This part converts and
amplifies the signal from the transducer into a
more easily read form. This often includes
amplifiers, filters, and analog-to-digital
converters. Some modern biosensors also
include microprocessors for on-device data
analysis.
 Applications of
Biosensors:
Medical Diagnosis: Glucose meters used by diabetics are among
the most well-known biosensors. They use an enzymatic reaction
catalyzed by glucose oxidase to detect glucose in blood samples.
Environmental Monitoring: Biosensors can detect contaminants in
water or air, such as heavy metals or pathogens.
Food Safety: They can be used to detect pathogenic
microorganisms in food samples or measure the concentration of
certain additives.
Research: In scientific research, biosensors can be used to measure
drug interactions, study enzyme kinetics, or detect specific molecules
in complex mixtures.
Defense and Security: Some biosensors are designed to detect
harmful agents or potential bioterrorism threats.
Advantages of
Biosensors:
Speed: Many biosensors offer almost real-time
or rapid detection.
Specificity: The biological component can be
chosen to specifically bind to a particular
molecule of interest.
Portability: Modern biosensors can be
miniaturized and are often portable, suitable
for field tests.
Integration with Electronics: Many
biosensors can be integrated into electronic
devices for easy readouts, data storage, and
communication.
 Based on Parameter Measured:
Physiological Sensors: Measure physiological parameters
like heart rate, blood pressure, or body temperature.
Biochemical Sensors: Measure chemical parameters in the
body, such as blood glucose levels, cholesterol, or specific
biomarkers.
Biopotential Sensors: Measure electrical potentials
generated by specific body parts, such as the heart
(EKG/ECG), muscles (EMG), or brain (EEG).
Molecular Sensors: Designed to detect specific molecules,
often using antibody-antigen interactions or specific binding
sites.
Imaging Sensors: Used in medical imaging to capture visual
data. Examples include the sensors in MRI, CT scans, or
ultrasound devices.
Biological sensing
component
According to biological sensing component, biosensor
may be divided into five classes:
• enzyme sensor,
• microbe sensor
• cell sensor,
• tissue sensor,
• immune sensors.
According to the signal converter of biosensor, biosensor
may be also divided into five classes: bioelectrode sensor,
semiconductor biosensor, optical biosensor, piezoelectric
biosensor and thermal biosensor. According to the
interaction between sensing component and measured
material, biosensor can be divided into two classes: affinity
biosensor and catalytic biosensor.
Biomedical sensor’s
application
In biomedical field, main applications of
biomedical sensor are as follows:
Detecting the information of clinical chemistry.
In the field of medical clinic and basic
research, the biology’s information needs to be
detected to ensure the present state of given
biology. For example, before operating on a
patient, a doctor needs to know the body
temperature and blood pressure. Under this
condition, clinic thermometer and blood sensor
has to be employed to help doctor quickly
detect body temperature and blood pressure of
patient.
Biomedical sensor’s
application
Continuously monitoring some parameters of
biology outside and inside. In biomedical field,
heart frequency has to be monitored
continuously by heart sound sensor for a few
days after operation. In military, some viruses
need to be found by biosensor to hold back the
attacking from enemy.
Biomedical sensor
classification
Many different kinds of sensors can be used in
biomedical application. According to the
sensing principle in biomedical application,
biomedical sensors can be classified into
physical sensors and chemical sensors, seen
in
 Biomedical sensor
classification
It’s possible to categorize all sensors as being physical or chemical. In the case
of physical sensors, quantities such as geometric, mechanical, thermal, and
hydraulic variables are measured.
In biomedical applications these variables can include things such as muscle
displacement, blood pressure, core body temperature, blood flow,
cerebrospinal fluid pressure, and bone growth velocity.
Two types of physical sensors deserve special mention with regard to their
biomedical application: sensors of electrical phenomena in the body, usually
known as electrodes, play a special role as a result of their diagnostic
therapeutic applications.
The most familiar of these are sensors used to pick up the electrocardiogram,
an electrical signal produced by the heart.
The other type of physical sensor that finds many applications in biology and
medicine is optical sensor. These sensors can utilize light to collect
information, and, in the case of fiber optic sensors, light is the signal
transmission medium as well.
 Biomedical sensor
classification
The second major classification of sensing device is
chemical sensors. In this case sensors are concerned with
the chemical quantities such as identifying the presence of
chemical composite, detecting the concentration of
various chemical species, and monitoring the chemical
activities in the body for diagnostic and therapeutic
application. A wide variety of chemical sensors are
classified in many ways. Chemical sensors are used to
detect chemical components being measured and
chemical composition measured in the gas phase.
 Biomedical sensor
classification
Electrochemical sensors are utilized to measure chemical
concentration, or more precisely, activities based on chemical
reactions that interact with electrical systems. Photometric
chemical sensors are optical devices that detect chemical
concentrations based on changes in light transmission,
reflection or color. Other types of physical chemical sensors
such as the mass spectrometer utilize various physical methods
to detect and quantify chemicals associated with biologic
systems.
Biomedical sensor
classification
Historical Perspective: 20th Century
n 1903 ECG – heart n 1957 Ultrasound Imaging
diagnostic (anatomical imaging)
n 1924 EEG – brain waves n 1970 CT Scanner
n 1928 ESU – cauterizing (anatomical imaging)
scalpel n 1975 Inter aortic balloon
n 1928 Iron Lung – pump
respiration assist n 1982 MRI (anatomical
n 1936 Nuclear Medicine imaging)
n 1956 Defibrillation n 1984 Artificial Heart
n 1957 Pacemaker (1960 n 1990s PET - use radio
implantable) isotopes (physiological
imaging)
52