Noise and interference, man-made noise, cosmic noise, thermal noise, noise in bio-potential recordings, and noise reduction techniques.

1. By:
Jaspreet Singh
Prashant Kumar
Department of Electrical and
Instrumentaion engineering
SLIET, India
In Bio-electrical Signals

2. With reference to an electrical signal, noise may be defined as any
unwanted form of energy which tends to interfere with transmission,
reception and reproduction of desired signal. It exhibits following
characteristics:
 Arise from variety of sources.
 Random in nature.
 Types are:
 Additive noise- refers to an interference added to the signal during its
transmission over a communication channel[4].
 Multiplicative noise- In signal processing, the term multiplicative noise
refers to an unwanted random signal that gets multiplied into some relevant
signal during capture, transmission, or other processing[4]. An important
example is the speckle noise commonly observed in radar imagery.

3.  The monitored biological signal in most cases is considered an additive
combination of signal and noise. Noise can be from instrumentation
(sensors, amplifiers, filters, etc.), from electromagnetic interference (EMI),
or in general, any signal that is asynchronous and uncorrelated with the
underlying physiology of interest. Therefore different situations warrant
different assumptions for noise characteristics, which will eventually lead
to an appropriate choice of signal-processing method.
 The output voltage of an amplifier used to measure bio-electric events is
not always an accurate representation of the event under examination[1] . It
is composed of desired voltage and number of unwanted voltages.

4.  Noise- The term noise is applied to random fluctuations of
voltage with more or less Gaussian distribution of
instantaneous amplitude [2]. Example: thermal noise
 Interference- The term interference on the other hand will be
used to describe unwanted voltages that are continuous in the
time domain and have a more deterministic repetitive
nature[2]. Example: 50 hertz interference from power lines.

5. Noise may be put into following two categories:-
1) External noise, i.e. noise produced by the signal source [2](also known as source
noise).
This noise signal is caused by thermal agitation currents in the signal source [2].
External noise may be classified into the following three types:
◦ Atmospheric noises
◦ Extraterrestrial noises
◦ Man-made noises or industrial noises.
2) Internal noise-. These are caused by thermal currents in the semiconductor
material resistances. These may be regarded as the noise component contributed
by the amplifier under consideration . Internal noise may be put into the
following four categories[2].
◦ Thermal Noise or White Noise or Johnson Noise
◦ Shot Noise
◦ Popcorn Noise
◦ Flicker Noise or Pink Noise or l/f Noise

6.  Atmospheric noise or static is caused by lighting discharges in
thunderstorms and other natural electrical disturbances
occurring in the atmosphere. These electrical impulses are
random in nature. Hence the energy is spread over the
complete frequency spectrum used for radio communication.
Atmospheric or Ambient Noise

7. There are numerous types of extraterrestrial noise or space
noises depending on their sources. However, these may be put
into following two subgroups.
 Solar noise
 Cosmic noise
These types of noise have strong impact on telecommunication.
So, they have to consider for telemedicine etc.

8. This is the electrical noise emanating from the sun. Under
quite conditions, there is a steady radiation of noise from the
sun. This results because sun is a large body at a very high
temperature (exceeding 6000°C on the surface), and radiates
electrical energy in the form of noise over a very wide
frequency spectrum including the spectrum used for radio
communication. The intensity produced by the sun varies with
time. In fact, the sun has a repeating 11-Year noise cycle.
During the peak of the cycle, the sun produces some amount
of noise that causes tremendous radio signal interference,
making many frequencies unusable for communications.
During other years. the noise is at a minimum level.

9.  Distant stars are also suns and have high temperatures. These
stars, therefore, radiate noise in the same way as our sun. The
noise received from these distant stars is thermal noise (or
black body noise) and is distributing almost uniformly over
the entire sky. We also receive noise from the centre of our
own galaxy (The Milky Way) from other distant galaxies and
from other virtual point sources such as quasars and pulsars

10.  By man-made noise or industrial- noise is meant the electrical noise
produced by such sources as automobiles and aircraft ignition, electrical
motors and switch gears, leakage from high voltage lines, fluorescent
lights, and numerous other heavy electrical machines. Such noises are
produced by the arc discharge taking place during operation of these
machines. Such man-made noise is most intensive in industrial and densely
populated areas. Man-made noise in such areas far exceeds all other
sources of noise in the frequency range extending from about 1 MHz to 600
MHz.
 Example – Electro-surgical Noise-Other equipments attached to subject for
healthcare purpose also alter our bio-signal of interest. So, they act as
source of Noise

11. Internal Noise
Thermal Noise
Conductors contain a large number of 'free" electrons and "ions"
strongly bound by molecular forces. The ions vibrate randomly
about their normal (average) positions, however, this vibration being
a function of the temperature. Continuous collisions between the
electrons and the vibrating ions take place. Thus there is a
continuous transfer of energy between the ions and electrons. This is
the source of resistance in a conductor. The movement of free
electrons constitutes a current which is purely random in nature and
over a long time averages zero. There is a random motion of the
electrons which give rise to noise voltage called thermal noise.

12.  Thus noise generated in any resistance due to random motion
of electrons is called thermal noise or white or Johnson noise.
 The analysis of thermal noise is based on the Kinetic theory. It
shows that the temperature of particles is a way of expressing
its internal kinetic energy. Thus "Temperature" of a body can
be said to be equivalent to the statistical rms value of the
velocity of motion of the particles in the body. At -273°C (or
zero degree Kelvin) the kinetic energy of the particles of a
body becomes zero .Thus we can relate the noise power
generated by a resistor to be proportional to its absolute
temperature. Noise power is also proportional to the
bandwidth over which it is measured.

13. From the above discussion we can
write down:
Pn ∝ TB
Pn = KTB ------ (1)
Where
Pn = Maximum noise power output
of a resistor.
K = Boltzmann’s constant
= 1.38 x10-23 joules / Kelvin.
T = Absolute temperature.
B = Bandwidth over which noise is
measured.
From equation (1), an equivalent
circuit can be drawn as shown in
below figure

14.  Problem
R.F. amplifier is saving an input resistor of 8Kr and
works in the frequency range of 12 to 15.5 MHz
Calculate the rms noise voltage at the input to this
amplifier at an ambient temperature of 17oC?
Solution:

15.  The most common type of noise is referred to as shot noise
which is produced by the random arrival of electrons or holes
at the output element, at the plate in a tube, or at the collector
or drain in a transistor. Shot noise is a consequence of DC
current flowing in any conductor[2] .Shot noise is also
produced by the random movement of electrons or holes
across a PN junction. Even through current flow is established
by external bias voltages, there will still be some random
movement of electrons or holes due to discontinuities in the
device. An example of such a discontinuity is the contact
between the copper lead and the semiconductor materials. The
interface between the two creates a discontinuity that causes
random movement of the current carriers.

16. • The electrons are emitted at Random times, Ʈk where -
∞ < k < ∞ and assume this random emission have been
gone for a long time.
• Thus the Total Current flowing through the Photodiode
may be modeled as the sum of these Current Pulses.
• This process X(t) is Stationary and is called SHOT
NOISE

17.  Several other forms of noise are present in linear integrated
circuits (ICs) and other semiconductor amplifiers to some
extent . For example , because current flow at the quantum
level is not smooth and predictable , an intermittent burst
phenomenon sometimes occurs. This noise , called Popcorn
noise ,consists of pulses of many milliseconds duration. This
noise is also known as burst noise[2].
 It consists of sudden step-like transitions between two or more
discrete voltage or current levels, as high as several hundred
micro volts at random and unpredictable times. Each shift in
offset voltage or current often lasts from several milliseconds
to seconds, and sounds like popcorn popping if hooked up to
an audio speaker.

18.  Flicker noise or modulation noise is the one appearing in
transistors operating at low audio frequencies. This noise is
also called as Pink noise or l/f noise. The latter name applies
because flicker noise is predominantly a low-frequency
(<1000 Hz) phenomenon. Flicker noise is proportional to the
emitter current and junction temperature[2].
 However, this noise is inversely proportional to the frequency.
Hence it may be neglected at frequencies above about 500 Hz
and it, Therefore, possess no serious problem.
 Its origin is believed to be attributable to contaminants and
defects in the crystal structure in semiconductors, and in the
oxide coating on the cathode of vacuum tube devices[2].

19.  Signal to Noise Ratio
Noise is usually expressed as a power because the received
signal is also expressed in terms of power. By Knowing the
signal to noise powers the signal to noise ratio can be
computed. Rather than express the signal to noise ratio as
simply a number, you will usually see it expressed in terms of
decibels.

20.  Noise Figure
Noise Figure F is designed as the ratio of the signal-to-
noise power at the input to the signal to noise power at
the output. The device under consideration can be the
entire receiver or a single amplifier stage. The noise
figure F also called the noise factor can be computed
with the expression
F = Signal to Noise power Input/Signal to noise power
output
You can express the noise figure as a number, more
often you will see it expressed in decibels.

21.  Motion Artifacts- movements can cause changes that are
created when an electrode is applied to the skin. Normally
when patient is relaxed and high quality electrodes are used ,
the recording is not distorted by motion artifact[1].
 A source of artifact only measured in moving cables (during
exercise) is tri-bioelectric noise[1]. Movement of electrode
cables inside a magnetic field may cause an inductive signal in
the wires.(example- in the strong magnetic field of a MRI-
scanner)

22.  Electric and Magnetic Field Interference
(a) Capacitive Coupling- The patient , the electrodes and the
electrode cable are capacitive coupled to nearby electric field ,
for instance the power lines. Interference appears in the non-
ideal situation in which the impedance of electrode or cable
pair differs[1]. A difference in cable impedance may for
instance be caused by the usage of cables of different length
which may lead to an interference potential of 200 micro-volt
peak to peak.
(b) Inductive Coupling- Because the measurement setup forms
a closed loop , a magnetic field can cause an inductive current
to flow in the loop[1] .
(c) Other causes are loose contact ,dirty electrodes and
disconnected electrodes resulting in strong disturbing signal.

23.  Baseline Wander- Variation in electrode skin impedance and
activities like patient movement and breathe cause baseline
wander thereby shifting[3]. Baseline wander disturbances is
dominant in exercise ECG and in ambulatory and Holter
monitoring . The range of frequency in which bandwidth is
dominant is less then 1 Hz .
 In wandering baseline, the iso-electric line change positions.
 Crosstalk- Phenomenon in which a signal transmitted in one
circuit or channel of a transmission system creates undesired
interference onto a signal in another channel[3]. Example –
muscle crosstalk( a phenomenon in which signal recorded over
one muscle is infact generated by a neighboring muscle and
conducted to the recording electrodes[1].

24.  Muscle Tremor- Irregular trembling of the ECG trace,
without wondering of the baseline occurs when patient is not
relaxed or is cold[3]. It is generally found in the case of older
patients. Muscle tremor signals are especially bothersome on
limb leads when a patient moves or the muscles are stretched.
Therefore, for long term monitoring, the electrodes are applied
on the chest and not on limbs.
 For normal routine ECG recording , the patient must be
advised to get warm and to relax so that muscle tremor from
shivering or tension is eliminated[3].

25.  Ocular Artifacts-The human eye generates an electrical
dipole caused by a positive cornea and negative retina. Eye
movements and blinks change the dipole causing an electrical
signal known as an EOG whose shape depends on factors such
as direction of eye movements[1]. A fraction of EOG spreads
across the scalp and it is superimposed on the EEG.
 Cardiac Artifacts-Because of its relatively high electrical
energy , it may have severe effects on EEG, especially upon
the non-cephalic reference recordings of EEG[1]. The QRS
complex appears in the EEG signal like regular spikes

26. 1) Faraday cage – A Faraday cage is a good way to reduce the
overall noise in a complete circuit[1]. The Faraday cage can be
thought of as an enclosure that separates the complete circuit
from outside power lines and any other signal that may alter
the true signal. A Faraday cage will usually block out most
electromagnetic and electrostatic noise.
2) Capacitive coupling – A current through two resistors, or any
other type of conductor, close to each other in a circuit can
create unwanted capacitive coupling[1]. If this happens an AC
signal from one part of the circuit can be accidentally picked
up in another part. The two resistors (conductors) act like a
capacitor thus transferring AC signals. There may be other
reasons for which capacitive coupling is wanted but then it
would not be thought of as electronic noise.

27. 3) Ground loops – When grounding a circuit, it is important to avoid
ground loops. Ground loops occur when there is a voltage drop
between the two ground potentials. Since ground is thought of as
0V, the presence of a voltage is undesirable at any point of a ground
bus. If this is the case, it would not be a true ground. A good way to
fix this is to bring all the ground wires to the same potential in a
ground bus[1].
4) Shielding cables – In general, using shielded cables to protect the
wires from unwanted noise frequencies in a sensitive circuit is good
practice. A shielded wire can be thought of as a small Faraday cage
for a specific wire as it uses a plastic or rubber enclosing the true
wire. Just outside of the rubber/plastic covering is a conductive
metal that intercepts any noise signal. Because the conductive metal
is grounded, the noise signal runs straight to ground before ever
getting to the true wire[1]. It is important to ground the shield at
only one end to avoid a ground loop on the shield.

28. 5) Twisted Pair Wiring- This dramatically reduce EM noise. It
decreases the loop size in which magnetic field can run through to
produce a current between the wires[1]. Even if the wires are
twisted very lightly , there may still be small loops somewhere
between them ,but because they are twisted the magnetic field going
through the smaller loops induces a current flowing in opposite
ways in each wire and thus cancelling them out.
6) Notch Filters-These are essential when eliminating a specific
noise frequency. In most cases the power line within a building run
at 60Hz. Sometimes a sensitive circuit will pick up this 60 Hz noise
through some unwanted antenna[1]. Running the output through a
notch filter tuned at 60Hz will amplify the desired signal without
amplifying the 60Hz noise.

29. 6) Filter-In signal processing, a filter is a device or process that
removes from a signal some unwanted component or feature.
Most commonly used filters are:-
 Electronic filters were originally entirely passive consisting
of resistance, inductance and capacitance. Active technology
makes design easier and opens up new possibilities in filter
specifications.
 Digital filters operate on signals represented in digital form.
The essence of a digital filter is that it directly implements a
mathematical algorithm, corresponding to the desired filter
transfer function, in its programming or microcode.

30.  Electronic filters Passive filters

31.  Electronic filters Active filters
Capacitors impedance depends on frequency (Xc = 1/(2πfC)) and the corner
frequency of an RC filter is fc = 1/(2πRC)

32.  Electronic filters Active filters
The circuit has the capacitor in series with R1. When f << fc the
capacitors reactance is large and R1 + Xc = Xc.
Therefore the gain is 1 + R2/ Xc = 1 when
Xc >> R2. When f >> fc the capacitors
reactance is small and R1 + Xc = R1.
Therefore the gain is 1 + R2/ R1 = 10
when Xc << R1.

33.  A digital filter system usually consists of an analog-to-digital
converter to sample the input signal, followed by a
microprocessor and some peripheral components such as
memory to store data and filter coefficients etc. Finally a
digital-to-analog converter to complete the output stage.
Program Instructions (software) running on the
microprocessor implement the digital filter by performing the
necessary mathematical operations on the numbers received
from the ADC.

34.  The removal or reduction of baseline wander and power line interference from
biomedical signal have been studied since ages and lot of techniques have been
proposed.
 Some techniques include maximally decimated filter, non-linear filter banks,
advanced averaging , wavelet transform, adaptive filtering techniques etc[1].
 In all ECG devices , digital FIR filters are used to filter and select ECG signal
in presence of different interfering signal which can be classified as LPF(to
remove undesired high frequency signal noise), HPF(to remove low frequency
noise signal), Band stop filter(to remove noise signal from power line
frequency of 50Hz,60Hz and also noise signal from muscle with frequency
25,35,45 Hz[1].
 A digital Notch filter design using hamming window removes the effect of
power line interference which achieves 13.4 dB attenuation. Using Adaptive
filter ,attenuation of 34.2 dB is obtained.

35.  [1] E.Huigen, “Noise in Bio-potential Recording using Surface Electrode” .
(University of Amsterdam Section Medical Physics, November 2000). Report no.,
S-915 , 37.
 [2] Joseph J. Carr, John M. Brown, " Introduction to Biomedical Equipment
Technology”. Fourth Edition
 [3] R.S.Khandpur, “Handbook of Biomedical Instrumentation”, Tata McGraw-
Hill,2007
 [4] R.Gallager, "Information theory and reliable communication" , McGraw-Hill
(1968)