Wavelet analysis is introduced in early the 1980s, and attracts attention as the method of changing for the conventional Fourier transformation.

On the other hand, Java is the object-oriented-programming language born in 1995, and Java is made strongly being conscious of Internet which is the leading role of an information network revolution and Java has developed with Internet.

Just Java and wavelet are considered to be positioned as core-technology of ** IT ** revolution. (If it is not illusion.)

This book is a very unique work by the author who is a clinician (physician) , and there is no other book of the similar contents.

The author of this book creates the software "Wavelet Analysis and Spectrum Analysis Software: MEM" by Java, and opens to the public by this book.

It divides and public presentation of the source code of wavelet packet analysis takes the initiative in the world. It is possible for this to associate the field of wavelet theory, and an information theory and fractal and chaos theory.

"Wavelet Analysis and Spectrum Analysis Software: MEM" by the author of this book are already broadly utilized for research and education not only in Japanese universities and Japanese research organizations but in overseas universities and overseas research organizations.

The 1st purpose of this book is to explained plainly the wavelet analysis that is difficult to undestand for a beginner or an outsider, in the range of high school mathematics.

First, a reader uses "Wavelet Analysis and Spectrum Analysis Software: MEM" in CD-ROM of this book, analyzes sample data, and grasps an image. After that, the concept and the definition formula are explained.

The reader aiming at applying wavelet transform to the analysis of time series data is borne in mind for this book.

It is also important to understand the concept (image) of wavelet transform rather than the details of each definition to a beginner in application.

If the contents which are holding the mathematical expression are understood as a feeling even if neither a beginner nor an outsider understands an orthodox operation-process from a purely mathematical standpoint, it is because it thinks that he can understand the rest by rereading only the item related with the book of basic mathematics or statistics if needed.

The 2nd purpose of this book is giving sufficient description about how the result of the analysis in "Wavelet Analysis and Spectrum Analysis Software: MEM" of time series data being interpreted.

If the existing commercial software is purchased and data is inputted, a certain display of a picture or numbers will be made in it. But, It is seldom described how this result should be interpreted and applied.

As time series data, circadian rhythm of blood-pressure, heart rate variability,electrocardiogram, etc. were used for this book, and in "Wavelet Analysis and Spectrum Analysis SoftWare: MEM", the analyzing methods, such as Continuation Wavelet Transformation (CWT), Discrete Wavelet Transformation (DWT), Wavelet Packet Analysis (WPA), FFT, the maximum entropy method (MEM),and an information theory etc., were used synthetically, and description was thoroughly added about the interpretation as a result of the analysis

The 3rd purpose of this book is that an entire beginner makes numerical computation and scientific research master indispensable knowledge at its minimum about programming so that programming by Java may be attained.

The most books of wavelet analysis are descriptions of Mathematica and S-Plus or MatLab which are the very expensive commercial software in Japan. With these books, it is only scripts. The source code which can be used by Java or C which is full-scale program language, is not found.

Even if it purchases expensive commercial software, wavelet analysis is not fully mastered to research only by it.

"Wavelet Analysis and Spectrum Analysis Software: MEM" by Java explained in this book exhibit the all source codes. moreover,Wavelet Packet Analysis and Best Base Algorithm are implemented.

Only the practical knowledge of Java for customizing "Wavelet Analysis and Spectrum Analysis Software: MEM" in this book, and fully utilizing for actual research is written.

In Chapter 1, the method of acquisition of JDK(Java Development Kits),installation of JDK, and how to compile of Java programming source code for an entire beginner, is described.

The installation of "Wavelet Analysis and Spectrum Analysis Software: MEM" is described.. The knowledge of required for numerical computation and customize of MEM is described.

The ideological subject of Design pattern, Object Orienteted Programming etc. is avoided in this book,and practical basic structure and grammar of Java applicable to research of a scientist and an engineer are described.

The usage of Java related site of SUN, sources of information of wavelet,PaperChaise which is the information database of medicine are described.

First,a reader operates sample data by the CWT of MEM, and has actually experienced the CWT. It is considered the shortcut to understanding of the concept that a reader actually experience the CWT.

Subsequently, the concept of CWT was explained using the model and the theoretical formula was shown. Process until it results in implement from theory is carefully explained about the typical wavelets of Gabor and Morlet.

From a relative scale, how to specify the frequency in wavelet is explained.

The theory of DWT (discrete wavelet transformation) is explained about the orthonormal wavelet of Daubechies. The construction of wavelet of Daubechies is described and implementation by Java is described.

For the reader, it pointed out clearly it would not be necessary what he should understand and what to understand. if you understand multiresolution analysis(MRA,multiresolution decomposition),Two-scale relation and orthogonality, it will come out enough.

Moreover, for this book, the author newly implemeted interpolatory graphical display algorithm, and released it. In case bi-orthogonal wavelet is studied, it is thought that it is useful.

How to use DWT in MEM are explained using sample data.

Public presentation of the source code of WPT takes the initiative in the world. It is possible to associate the field of wavelet theory, an information theory, and fractal and chaos theory by WPT.

Moreover, in CWT and DWT, although limited to the analysis in a certain specific frequency domain, the analysis in all frequency domains is attained by WPT.

About WPT and BBA based on information cost fanctional which made Information-entropy, Gauss-Markov entropy,a theoretical dimension, etc. the cost, sample data is used and the usage is explained using MEM.

In this chapter, it learns through the analysis of the waveform of an electrocardiogram about the actual application of WPT and BBA by the information cost functional.

In the software MEM of this book, it is possible to interlock CWT, DWT and WPT by scllorbar.

CWT and MRA are also used together and existence of p wave which cannot be checked in a raw signal is checked . The advantage and fault of CWT, DWT, and WPT were also explained.

This chapter describes concretely the usage in MEM of CWT, DWT, WPA, and Information Cost Functional through the analysis of SAECG.

The synthetic usage of software MEM is mastered and a reader can understand how the result is interpreted.

Moreover, the method of denoise which makes theoretical dimension an index and sets the multiple of the standard deviation of wavelet coefficient to threshold is introduced.

This chapter describes FFT, Maximum entropy method and System Identification which makes an index Infomation Ccriterion such as AIC and MDL explained in Chapter 8. The data of measurement of R-R interval of ECG is analyzed by CWT,WPT,FFT and Maximum Entropy Method and the result is explained.

I am pleased if a reader realize how many important information is acquired on clinical by hitting the light of wavelet analysis and information theory to the data which drew the histogram and the trend, only carried out easy statistics processing of standard deviation etc., and was left .

In this chapter, a reader understands FFT, Maximum Entropy Method and Cosinor method which have been performed in the conventional time series analysis. The model identification which makes information criterion, such as AIC,FPE and MDL, an index is described. And a reader studies the synthetic usage of "software MEM".

FFT and Maximum Entropy Method which are regarded as required for time series analysis are explained, and source codes of Java are implemented. In relation to Cosinor method, an input signal is considered that sum of cosine waves,each of which has different frequency, and how to calculate such frequency by least square method is explained. The source code is implemented.

The circadian rhythm of blood pressure by actual 24 hour blood-pressure automatic measurement equipment is analyzed using Wavelet Packet Transformation and information cost functional.

It is thought that the sample data of the sum of cosine and the pseudo-random numbers which have been conventionally considered to be right,it is not suitable as simulation of an actual blood-pressure flucuation from the viewpoint of wavelet analysis and information theory.

It is thought that two dimensional wavelet transformation (2DWT) will become the important method of the next generation in the fields of image processing . A fundamental theory of 2DWT was explained. It opened to the public about the fundamental source codes of 2DWT.

ISBN4-7678-0000-2

to WebSite of Medical Publication (IGAKU-SHUPPAN)
: "Wavelet Analysis for Clinical Medicine"

Publisher's Name: IGAKU-SHUPPAN

Price: 16,000YEN

Address:

2-16-12 Hongou

Bunkyou-ku Tokyo

113-0033 Japan

phone:+81-3-3813-8722

fax:+81-3-3818-7888

E-mail to Publisher: info@igaku.co.jp

This documents were auto-tranlated by excite.honyaku and Gist