Skip to main content
main-content

Über dieses Buch

A detailed description of a new approach to perceptual analysis and processing of medical images is given. Instead of traditional pattern recognition a new method of image analysis is presented, based on a syntactic description of the shapes selected on the image and graph-grammar parsing algorithms. This method of "Image Understanding" can be found as a model of mans' cognitive image understanding processes. The usefulness for the automatic understanding of the merit of medical images is demonstrated as well as the ability for giving useful diagnostic descriptions of the illnesses. As an application, the production of a content-based, automatically generated index for arranging and for searching medical images in multimedia medical databases is presented.

Inhaltsverzeichnis

Frontmatter

Introduction

Abstract
This book proposes a new approach to the processing and analysis of medical images. We introduce the term (and methodology) medical data understanding, as a new step in the way of starting from image processing, and followed by analysis and classification (recognition). The general view of the situation of new technology under consideration in the context of the more well known techniques of image processing, analysis, segmentation and classification is shown below.
Ryszard Tadeusiewicz, Marek R. Ogiela

1. What is Image Understanding Technology and why do we need it?

Abstract
Medical procedures which involve the use of various images constitute a particularly interesting and important field of informatics, especially as contemporary medical diagnosis is to a large extent based on images. Medical Imaging Technology has now become one of the major sources of information for therapy and also for research in medicine, biology and in other related fields. This information covers the morphology of examined organs and their function (though in a limited degree), which affords us the means to infer whether their functioning is correct and whether the structures are pathological. This allows well-founded conclusions to be drawn or helps in modern diagnostic reasoning processes.
Ryszard Tadeusiewicz, Marek R. Ogiela

2. A General Description of the Fundamental Ideas Behind Automatic Image Understanding

Abstract
In this chapter we shall try to explain what automatic image understanding is and how we can force the computer to understand the image content. Before we get down to details we must demonstrate that there is in fact a fundamental difference between a formal description of an image (typically obtained by means of a variety of computer viewing methods) and the content meaning of the image, which can be discovered by an intelligent entity, capable of understanding the profound sense of the image in question. Although the problem under consideration is rather serious, we shall now use a joke to show how weak traditional computer vision technology can be in applications in which image understanding is the bottleneck of the proper solution.
Ryszard Tadeusiewicz, Marek R. Ogiela

3. Formal Bases for the Semantic Approach to Medical Image Processing Leading to Image Understanding Technology

Abstract
This chapter presents the basic terms of the formal language and finite state automata theory that are associated with syntactic pattern recognition methods. The symbols and terms introduced here will be used throughout this book; the formalisms introduced here will be used by the Authors in cognitive analyses of images presented in this book.
Ryszard Tadeusiewicz, Marek R. Ogiela

4. Examples of Structural Pattern Analysis and Medical Image Understanding Application to Medical Diagnosis

Abstract
This chapter will present the results of actual medical applications of the new approach described in previous chapters. We shall try to demonstrate that the structural pattern analysis can be regarded as an effective tool for medical image understanding, replacing simple recognition (Figure 4.1). Structural image analysis can be considered as a totally new approach to the analysis and description of shapes of selected organs in medical imaging in general. Examples of syntactic methods of pattern recognition application for the understanding and analysis of selected medical images presented in this chapter show their usefulness for early diagnosis of some diseases of selected organs. Analysis results of investigations based on structural analysis of selected types of medical images confirm very good properties of the proposed methodology and algorithms under consideration.
Ryszard Tadeusiewicz, Marek R. Ogiela

5. The application of the Image Understanding Technology to Semantic Organisation and Content-Based Searching in Multimedia Medical Data Bases

Abstract
In recent years the knowledge of engineering has become one of the most dynamically developing branches of intelligent IT systems, including the PACS, Computer-Aided Diagnosis and medical data bases. Specialist medical data bases storing data in visual form constitute a large group of multimedia data bases; those patterns originate from numerous diagnostic examinations of practically all organs of the human body. One of the main problems with accessing and analysing information collected in this way is how to transform efficiently the visual information transfer of those patterns into a form enabling intelligent analysis and understanding of medical meaning of these patterns.
Ryszard Tadeusiewicz, Marek R. Ogiela

6. Strengths and Weaknesses of the Image Understanding Technology Compared to Previously Known Approaches

Abstract
The techniques of automatic image understanding, presented in this book, utilising the linguistic approach have several major advantages over classical image recognition algorithms. It is readily apparent that for many types of images, in particular medical images, it is difficult to interpret and define the representative vector of numerical features required in the classical approach applied in theoretical decision-based methods. This means that a certain type of images containing structural information can be extremely difficult or even impossible to classify on the basis of selected features represented in numerical form. This is so because the structures ought to be described in such a manner that some relationships and constituent elements of the structure are first defined while the structure itself can be described in general terms with a use of a model or strictly specified. The presence of semantic information requires therefore that analysis be made, both of the classification and description (meaning) sense. The classification task is based chiefly on operations of seeking similarities (usually referred to as the grammar derivation path); yet this operation has some generalisation properties and together with semantic information obtained in the course of analysis, it allows us to recognise a practically unlimited number of classes and objects.
Ryszard Tadeusiewicz, Marek R. Ogiela

Backmatter

Weitere Informationen