1

CHAPTER 1

INTRODUCTION

Biosignals represent the physiological changes occurring inside the

human body, which convey useful information about the functions of the
physiological system. These signals are the bioelectric potentials associated
with nerve conduction, brain activity, heart beat, muscle activity and so on.
Commonly used bioelectric signals are ECG, EEG, and EMG. These signals
are generally contaminated with many artifacts or noise. The main focus of
this research is to cancel the interferences in these biosignals using Artificial
Intelligence (AI) techniques. This chapter explains the significance of
biosignals, motivation for cancellation of interference in the biosignals and
also gives the organization of the report.

1.1 CHARACTERISTICS OF BIOSIGNALS

Signals produced by biological system are always continuous. Real

biosignals almost always have some unpredictable noise or change in
parameters and therefore they are called random signals. Since the statistical
properties or frequency spectra of the biosignals vary with time, biosignals are
called non-stationary signals. Biosignals are often very small in amplitude.
These are often masked by other biosignals from different biological
phenomena.
 2

1.2 BIOPOTENTIAL MEASUREMENTS

Biopotential measurements are made using different kinds of

specialized electrodes. The function of these electrodes is to couple the ionic
potentials generated inside the body with an electronic instrument.
Biopotential electrodes are classified either as noninvasive (skin surface) or as
invasive (microelectrodes or wire electrodes). Biopotential electrode for ECG
recording is made of silver/silver chloride (Ag/AgCl). Cup electrodes and
subdermal needle electrodes are used for recording EEG signals. Circular
discs, about 1 cm in diameter, that are made of silver or platinum are used for
surface EMG recording. The shape and size of the recorded EMG signals
depends on the electrical property of these electrodes and the recording
location.

1.3 SIGNIFICANCE OF ELECTROCARDIOGRAM

An ECG is a recording of the heart’s electrical activity. Each

heartbeat is caused by a section of the heart generating an electrical signal,
which then conducts through specialized pathways to all parts of the heart.
These electrical signals also get transmitted through the chest to the skin.
Nowadays, cardiac failure and cardiac diseases are becoming one of the main
causes of death. Therefore, it is necessary for physicians and researchers to
have a satisfactory way to determine the cardiac condition of the patient.
ECG is one of the most direct ways to study the cardiac condition because it
records the electrical current generated over the cardiac cycle. Any cardiac
abnormalities will be clearly seen through investigation of ECG signals. This
provides doctors with the most updated cardiac condition of the patient so that
proper treatment can be given. The physician can assess the patient’s
sickness by understanding abnormalities of certain sections of the signal.
 3

The main applications of the ECG to cardiological diagnosis

include the following:

 coronary artery disease in which blood flow to the heart and

body is restricted due to hardened arteries

 heart attack

 arrhythmias which causes abnormal heart rhythms

 atrial fibrillation

 cardiac hypertrophy which shows enlarged or thickened

heart muscle

 cardiomyopathy in which the heart muscle functions

improperly due to a variety of conditions

 pericarditis which is the inflammation of the pericardium

 long Q-T syndrome is a disorder of the heart’s electrical

system, which could lead to fainting or sudden cardiac death

 myocarditis is the inflammation of the heart muscle due to

viral or bacterial infection

 certain congenital heart defects.

In the recording of biomedical signals, a significant noise

component is introduced by the electrode. Power line interference, muscle
contraction, and respiration are few of the main sources of noise observed in
the ECG signals. The magnitude of these noises is considerably higher than
the equivalent thermal noise from the electrode impedance. Hence, the
removal of noise from ECG signal is a very important step in the process of
ECG examination by physicians.
 4

1.4 SIGNIFICANCE OF FETAL ELECTROCARDIOGRAM

FECG is useful to get reliable information on fetal status, the

detection of abnormalities and monitorization task during labor, to enable the
adoption of measures for assuring fetal well-being, to detect whether the fetus
is alive or dead, and to determine twin pregnancies. The diagnostic tests of
fetal well-being can be categorized as invasive and noninvasive. During
delivery, accurate recordings can be made by placing an electrode on the fetal
scalp. However, as long as the membranes protecting the child are not broken,
one should look for noninvasive techniques. Electrode output from the
abdomen using noninvasive techniques is hampered by MECG, EMG due to
uterus contraction and respiration, power line interference and thermal noise
from electronic equipments. The main source of interference in FECG is
MECG, the amplitude of which is much higher than that of the FECG. One of
the aims of this work is the cancellation of MECG from FECG signal.

1.5 SIGNIFICANCE OF ELECTROENCEPHALOGRAM

Bioelectric potentials generated by the neuronal activity of the brain

are called the EEG. Due to its capability to reflect both the normal and
abnormal electrical activity of the brain, EEG has been found to be a very
powerful tool in the field of neurology and clinical neurophysiology. EEG can
determine the relative strengths and positions of electrical activity in different
brain regions. The important clinical applications of the EEG are to

 monitor alertness, coma and brain death

 locate areas of damage following head injury, stroke, tumor,

etc

 test afferent pathways monitor cognitive engagement

 5

 produce biofeedback situations, alpha, etc

 control anesthesia depth

 investigate epilepsy and locate seizure origin

 test epilepsy drug effects

 assist in experimental cortical excision of epileptic focus

 monitor human and animal brain development

 test drugs for convulsive effects, and

 investigate sleep disorder and physiology.

Signals in the EEG that are of non-cerebral origin are called

artifacts. The EEG is nearly always contaminated by such artifacts. This is
one of the reasons why considerable expertise is required to interpret EEGs
clinically. The most common EEG artifact sources are eye blink, eye
movement, power line interference, muscle activity, and ECG. In this work,
an attempt has been made to remove some important artifacts from the EEG
signal.

1.6 SIGNIFICANCE OF ELECTROMYOGRAM

The bioelectric potentials associated with muscle activity are called

the EMG. When EMG is acquired from electrodes mounted directly on the
skin, the signal is a composite of all the muscle fiber action potentials
occurring in the muscle(s) underlying the skin. These action potentials occur
at somewhat random intervals and hence at any one moment, the EMG signal
may be either positive or negative voltage. Individual muscle fiber action
potentials are sometimes acquired using wire or needle electrodes placed
directly in the muscle. Electromyography is a test that assesses the health of
 6

the muscles and the nerves controlling the muscles. It measures muscle
response to nervous stimulation.

EMG is used clinically for the diagnosis of neurological and

neuromuscular problems. It is used diagnostically by gait laboratories and by
clinicians trained in the use of biofeedback or ergonomic assessment. EMG is
also used in many types of research laboratories, including those involved in
biomechanics, motor control, neuromuscular physiology, movement
disorders, postural control, physical therapy, and many others. EMG is most
often used when people have symptoms of weakness and examination shows
impaired muscle strength. It can help to differentiate primary muscle
conditions from muscle weakness caused by neurologic disorders. As
contraction is increased, more and more muscle fibers produce action
potentials until a disorderly group of action potentials of varying rates and
amplitudes (complete recruitment and interference pattern) appears with full
contraction. The signal is susceptible to numerous artifacts like power line
interference, baseline drift, white noise, ECG, EEG, stimulation artifacts. The
artifacts cancellation in EMG using AI techniques is presented in this work.

1.7 PROBLEM FORMULATION

Biosignals need to be monitored carefully by doctors, as they serve

as symptoms for diagnosis. In order to provide an accurate diagnosis, the
measured biosignals must be noise free. Generally, noninvasive method of
acquiring biosignals is preferred in medical applications. The biosignals
obtained using this method always consist of noise or artifacts. The
cancellation of noise in biosignals can be performed using conventional fixed
filters. These filters also remove an important part of the signal if the
frequency of the noise is very close to that of the required signal. Since the
 7

noise present in the biosignal changes from time to time, AIC method is used
in this research work.

In this method, an estimate of the interference in the composite

signal is obtained by mapping the known measurable noise signal with the
composite signal which has both signal and unknown interference. The
estimated interference is subtracted from the composite signal in order to
retrieve the required signal. The efficiency of an interference cancellation
method depends very much on the accuracy of the estimated interference.
Linear adaptive filters have been used for successful cancellation of periodic
interferences in stationary signals. Not much work has been reported in the
literature regarding cancellation of nonlinear, non-stationary interferences in
biosignals. Moreover, the existing methods do not provide the quantitative
analysis to validate their results. Hence, an attempt has been made to use AI
techniques namely BPN, CCN, ANFIS, and ANFIS-FCM to accurately
estimate the unknown interferences and to recover the required signal without
any degradation. Also some experiments are carried out to cancel
interferences in simulated signals using conventional methods like LMS filter,
Adaline network, and ANFIS.

1.8 OBJECTIVES OF THE WORK

The main objectives of the work are

 to cancel the random noise and power line interference from

the simulated signals like sine wave, synthetic ECG signal
and real ECG signal using LMS filter, Adaline network, and
ANFIS
 8

 to cancel the MECG from the composite abdominal ECG in

order to recover the FECG using BPN, CCN, ANFIS, and
ANFIS-FCM

 to cancel various physiological interferences like EMG

EOG-Right, and EOG-Left from EEG signal using BPN,
CCN, ANFIS, and ANFIS-FCM

 to cancel power line interference, EEG, and ECG

interferences from EMG signal using BPN, CCN, ANFIS,
and ANFIS-FCM, and

 to perform quantitative analysis of the results obtained using

BPN, CCN, ANFIS, and ANFIS-FCM in terms of
performance measures like Mean Square value of the
estimated signal, SNR, and convergence time.

The first part of the thesis is to investigate the removal of random

noise from the simulated signals like sine wave, synthetic ECG signal, and
real ECG signal with simulated noise, using traditional linear filters namely
LMS, and Adaline network. The results obtained using these techniques are
not satisfactory. Hence a nonlinear technique called ANFIS is employed for
interference cancellation in the above simulated signals. The second part of
this research work deals with interference cancellation in real biosignals using
AI techniques. A comparative study of these techniques is made in terms of
Mean Square value of the estimated signal, SNR, and convergence time.
 9

The block diagram of the proposed work is given in Figure 1.1.

Data Base

Adaptive Interference Cancellation (AIC) in

Simulated Signals Real Signals

Required signals
after AIC

Synthetic EEG
Real ECG
Sine ECG Signal Signal FECG
Wave EMG
Techniques used

1. BPN
1. LMS Filter 2. CCN
2. Adaline Network 3. ANFIS
3. ANFIS 4. ANFIS-FCM

Comparison between techniques

Conclusion

Figure 1.1 Block diagram of proposed work

1.9 ORGANIZATION OF THE THESIS

The rest of the thesis is organized as follows.

A comprehensive review of literature on FECG extraction, and

cancellation of interference from FECG, EEG and EMG are presented in
 10

Chapter 2. This chapter also includes literature survey on the applications of

AI techniques such as Back Propagation Network, Cascade Correlation
Network, Adaptive Neuro Fuzzy Inference System and Fuzzy C-Means
clustering.

Chapter 3 describes the concept of interference cancellation

proposed in this work. The architecture and operations of the BPN, CCN, and
ANFIS are presented. The use of clustering algorithm is also discussed in this
chapter.

Chapter 4 provides detailed information on interference

cancellation in various simulated signals using LMS filter, Adaline network,
and ANFIS.

Chapter 5 discusses the use of proposed AI techniques for the

cancellation of interference (particularly MECG) in FECG.

The application of four AI techniques for the interference

cancellation in EEG is discussed in Chapter 6.

Chapter 7 presents the use of the AI techniques for the cancellation

of interference in EMG.

The results of interference cancellation are presented in the

respective Chapters.

The final chapter gives the summary of work done, contribution of

the work, limitations and scope for further work.