Biosignal and Medical Image Processing

John L. Semmlow, Benjamin Griffel

• Publisher: CRC Press
• Year: 2021
• Print ISBN: 9781466567368
• eBook ISBN: 9781000535792
• Topic: Medicine
• Subtopic: Biomedical Science

Written specifically for biomedical engineers, Biosignal and Medical Image Processing, Third Edition provides a complete set of signal and image processing tools, including diagnostic decision-making tools, and classification methods. Thoroughly revised and updated, it supplies important new material on nonlinear methods for describing and classify

1

Introduction

1.1 Biosignals

Much of the activity in biomedical engineering, be it clinical or research, involves the measurement, processing, analysis, display, and/or generation of biosignals. Signals are variations in energy that carry information, and the search for information from living systems has been a preoccupation of medicine since its beginnings (Figure 1.1). This chapter is about the modification of such signals to make them more useful or more informative. To this end, this chapter presents tools and strategies that use digital signal processing to enhance the information content or interpretation of biosignals.

Figure 1.1 Information about internal states of the body can be obtained through the acquisition and interpretation of biosignals. Expanding such information is an ongoing endeavor of medicine and medical research. It is also the primary motivation of this chapter.

The variable that carries information (the specific energy fluctuation) depends on the type of energy involved. Biological signals are usually encoded into variations of electrical, chemical, or mechanical energy, although, occasionally, variations in thermal energy are of interest. For communication within the body, signals are primarily encoded as variations in electrical or chemical energy. When chemical energy is used, encoding is usually done by varying the concentration of the chemical within a “physiological compartment,” for example, the concentration of a hormone in blood. Bioelectric signals use the flow or concentration of ions, the primary charge carriers within the body, to transmit information. Speech, the primary form of communication between humans, encodes information as variations in pressure. Table 1.1 summarizes the different types of energy that can be used to carry information and the associated variables that encode this information. Table 1.1 also shows the physiological measurements that involve these energy forms.

Table 1.1 Energy Forms and Associated Information-Carrying Variables

Energy	Variables (Specific Fluctuation)	Common Measurements
Chemical	Chemical activity and/or concentration	Blood ion, O2, CO2, pH, hormonal concentrations, and other chemistry
Mechanical	Position Force, torque, or pressure	Muscle movement, cardiovascular pressures, muscle contractility, valve, and other cardiac sounds
Electrical	Voltage (potential energy of charge carriers) Current (charge carrier flow)	EEG, ECG, EMG, EOG, ERG, EGG, and GSR
Thermal	Temperature	Body temperature and thermography

Outside the body, information is commonly transmitted and processed as variations in electrical energy, although mechanical energy was used in the seventeenth and early eighteenth centuries to send messages. The semaphore telegraph used the position of one or more large arms placed on a tower or high point to encode letters of the alphabet. These arm positions could be observed at some distance (on a clear day) and relayed onward if necessary. Information processing can also be accomplished mechanically, as in the early numerical processors constructed by Babbage. Even mechanically based digital components have been attempted using variations in fluid flow. Modern electronics provides numerous techniques for modifying electrical signals at very high speeds. The body also uses electrical energy to carry information when speed is important. Since the body does not have many free electrons, it relies on ions, notably Na⁺, K⁺, and Cl^-, as the primary charge carriers. Outside the body, electrically based signals are so useful that signals carried by other energy forms are usually converted into electrical energy when significant transmission or processing tasks are required. The conversion of physiological energy into an electric signal is an important step, often the first step, in gathering information for clinical or research use. The energy conversion task is done by a device termed a transducer,^* specifically, a biotransducer. The biotransducer is usually the most critical component in systems designed to measure biosignals.

With the exception of this chapter, this book is limited to topics on digital signal and image processing. To the extent possible, each topic is introduced with the minimum amount of information required to use and understand the approach along with enough information to apply the methodology in an intelligent manner. Strengths and weaknesses of the various methods are also explored, particularly through discovery in the problems at the end of the chapter. Hence, the problems at the end of each chapter, most of which utilize the MATLAB^® software package (Waltham, MA), constitute an integral part of the book and a few topics are introduced only in the problems.

A fundamental assumption of this book is that an in-depth mathematical treatment of signal-processing methodology is not essential for effective and appropriate application of these tools. This book is designed to develop skills in the application of signal- and image-processing technology, but it may not provide the skills necessary to develop new techniques and algorithms. References are provided for those who need to move beyond the application of signal- and image-processing tools to the design and development of new methodology. In the subsequent chapters, the major topics include sections on implementation using the MATLAB software package. Fluency with the MATLAB language is assumed and is essential for the use of this book. Where appropriate, a topic area may also include a more in-depth treatment, including some of the underlying mathematics.

1.2 Biosignal Measurement Systems

Biomedical measurement systems are designed to measure and usually record one or more biosignals. A schematic representation of a typical biomedical measurement system is shown in Figure 1.2. The term biomedical measurement is quite general and includes image acquisition and the acquisition of different types of diagnostic information. Information from the biological process of interest must first be converted into an electric signal via the transducer. Some analog signal processing is usually required, often including amplification and lowpass (or bandpass) filtering. Since most signal processing is easier to implement using digital methods, the analog signal is converted into a digital format using an analog-to-digital converter (ADC). Once converted, the signal is often stored or buffered in the memory to facilitate subsequent signal processing. Alternatively, in real-time applications, the incoming data are processed as they come in, often with minimal buffering, and may not be permanently stored. Digital signal-processing algorithms can then be applied to the digitized signal. These signal-processing techniques can take on a wide variety of forms with varying levels of sophistication and they make up the major topic areas of this book. In some applications such as diagnosis, a classification algorithm may be applied to the processed data to determine the state of a disease or the class of a tumor or tissue. A wide variety of classification techniques exist and the most popular techniques are discussed in Chapters 16 and 17. Finally, some sort of output is necessary in any useful system. This usually takes the form of a display, as in imaging systems, but it may be some type of effector mechanism such as in an automated drug delivery system. The basic elements shown in Figure 1.2 are discussed in greater detail in the next section.

Figure 1.2 Schematic representation of a typical biomedical measurement system.

1.3 Transducers

A transducer is a device that converts energy from one form to another. By this definition, a lightbulb or a motor is a transducer. In signal-processing applications, energy is used to carry information; the purpose of energy conversion is to transfer that information, not to transform energy as with a lightbulb or a motor. In measurement systems, all transducers are so-called input transducers: they convert nonelectrical energy into an electronic signal. An exception to this is the electrode, a transducer that converts ionic electrical energy into electronic electrical energy. Usually, the o...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Dedication
• Table of Contents
• Preface
• Acknowledgments
• Authors
• Chapter 1 Introduction
• Chapter 2 Biosignal Measurements, Noise, and Analysis
• Chapter 3 Spectral Analysis: Classical Methods
• Chapter 4 Noise Reduction and Digital Filters
• Chapter 5 Modern Spectral Analysis: The Search for Narrowband Signals
• Chapter 6 Time–Frequency Analysis
• Chapter 7 Wavelet Analysis
• Chapter 8 Optimal and Adaptive Filters
• Chapter 9 Multivariate Analyses: Principal Component Analysis and Independent Component Analysis
• Chapter 10 Chaos and Nonlinear Dynamics
• Chapter 11 Nonlinearity Detection: Information-Based Methods
• Chapter 12 Fundamentals of Image Processing: The MATLAB Image Processing Toolbox
• Chapter 13 Image Processing: Filters, Transformations, and Registration
• Chapter 14 Image Segmentation
• Chapter 15 Image Acquisition and Reconstruction
• Chapter 16 Classification I: Linear Discriminant Analysis and Support Vector Machines
• Chapter 17 Classification II: Adaptive Neural Nets
• Appendix A: Numerical Integration in MATLAB
• Appendix B: Useful MATLAB Functions
• Bibliography
• Index