Neural Networks and Artificial Intelligence for Biomedical Engineering

Name of the Book: Neural Networks and Artificial Intelligence for Biomedical Engineering.

Authors: Donna L. Hudson, Maurice E. Cohen

Publication: IEEE Press, New York, 2000

ISBN: 0-7803-3404-3

Pages: 306

‘Neural Networks and Artificial Intelligence for Biomedical Engineering’
[Reviewed by Prachi A Deshmukh]


The advent of technology in the twentieth century has changed the total scenario around us. The fiction stories which we were reading in books have now become the truth, the reality. Robotics and artificial intelligence are some of the examples.

Neural networks and artificial intelligence have created new horizons in the field of medicine and biomedical engineering.

The book ‘Neural Networks and Artificial Intelligence for Biomedical Engineering’ by Donna Hudson and Maurice Cohen is a good guide for those who are working in this field or for the students who are studying neural networks, artificial intelligence or biomedical engineering as a part of their course work.

The book is divided into three parts: The first part covers Basics Neural networks, the second discusses Artificial Intelligence and the third one comprises of Alternative approaches.

Before starting with the neural networks, the authors have introduced us with the basics of biological systems as well as medical and biological data. This makes easy for us to enter in the field.

In part one, there are eight chapters.

All these are related with the neural networks. Chapter one is the foundation of neural networks. This one introduces us with the basic things regarding the neural networks. Starting from the structure of a biological cell and neuron, central nervous system we move towards the early neural models and current models. Second chapter classifies the neural networks. In this chapter we also learn the techniques which are used in biomedical problems. In chapter 3, 4 ad 5 we learn the network structure, feature selection and types of learning. Chapter 4 and 5 are dedicated to supervised learning and unsupervised learning respectively.
Chapter 6 is regarding the design issues. In this we learn the input data types, structure of networks implementation of network structures and the choice of learning algorithms. In chapter 7, we do the comparative analysis. In this chapter we again deal with the supervised and unsupervised learning. In this chapter we also learn the network structures.
The last chapter of part one is chapter no 8: Validation and evaluation. Here we become familiar with some important concepts such as data checking, validation and learning algorithms and evaluation performance.

After going through the first part of the book, we have with ourselves a strong base on neural networks.

In second part we move towards the Artificial Intelligence. Artificial Intelligence is an important field in the history of science and technology.

The second part is divided into 5 chapters.

First chapter of this part is chapter no. 9: ‘Foundations of computer assisted decision making’. Here the authors introduce us with the databases and medical records, mathematical modeling and simulation, pattern recognition, decision theory, symbolic reasoning techniques. Chapter no. 10 and chapter no. 11 are knowledge representation and knowledge acquisition respectively. These two introduce us with the production rules, frames, databases, nets (predicate calculus and semantic nets),learned knowledge, meta knowledge and knowledge base maintenance.
Chapter 12 deals with the reasoning methodologies. Here we learn the problem representations, blind searching, trees-graphs and higher level reasoning methodologies. In chapter 13 we learn the validation and evaluation. Here we learn the algorithmic evaluation, knowledge base evaluation and system evaluation. The part two terminates with the end of chapter no. 13.

In third part there are 6 chapters which are arranged under the name ‘Alternate approaches’.

In chapter 14 we learn about the genetic algorithms. In this chapter we learn about the representation schemes, evaluation functions, genetic operators, evaluation strategies and some biomedical examples.

Chapter 15 and 16 are probabilistic systems and fuzzy systems respectively. In chapter 15: probabilistic systems, we learn about the Bayesian approaches in which we become familiar with the Baye’s rule and Baye’s decision theory, parameter estimation, discriminant analysis, statistical analysis, regression analysis. At the end of the chapter brief information about the medical applications is given.

Chapter no. 16 is related with the Fuzzy systems. Here we get the information of fuzzy logic and fuzzy set theory, representation of fuzzy variables, membership functions, fuzzy neural networks, fuzzy approaches of supervised learning networks, fuzzy generalizations of unsupervised learning methods, reasoning with uncertain information, pre-processing and post processing using fuzzy techniques etc.

At the end we find some applications in biomedical engineering.

In chapter 17 we become familiar with the hybrid systems. Here we learn about the hybrid system approaches, components of the hybrid systems, use of complex data structures, design methodologies etc.

Chapter 18 deals with the ‘HyperMerge’ which is a hybrid expert system. Here we learn about the knowledge based component, neural network component, analysis of time series data, combined system and lastly an application: Diagnosis of Heart Disease.

Chapter 19 is the concluding chapter of this part as well as the book. It is ‘Future Perspectives’. In this chapter different aspects are considered such as the effects of hardware advances, effects of increase in knowledge and the future of software. In this chapter different important parameters are discussed. Computing speed, memory, parallel machines, miniaturization etc. The different effects of increase in knowledge such as information explosion, human genome project, proliferation of databases, communication of information etc are studied. The end of the chapter the authors discuss about the future of the software. Here we discuss about the hybrid systems, parallel systems, non textual data, Neural network models and the artificial intelligence approaches.

This book provides useful guidance for the graduate and undergraduate students who are studying this subject and I strongly commend the book to those interested in this fascinating field.