الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 174 |  |  |
| | | from | 174 | | |
Abstract
The so-called information society demands for complete access to available information, which is often heterogeneous and distributed especially in the medical domain. Years of studies in the areas of human disease have resulted in millions of papers, articles and journals whose accumulated volume is terabytes in size. This information is available on the Web through journal databases, medical databases, e-health systems etc. and it covers various types of diseases, symptoms, treatment and disease-causing factors such as genetics, environment and microorganisms. It is difficult for medical researchers, doctors, patients and the general public to search for, retrieve and interpret the relevant information pertaining to human disease through these databases, and there is no tool available to help access, filter, integrate and present such information.
The sheer volume of information presents a significant obstacle to medical researchers (identification of the causes of diseases), doctors (diagnosis of diseases), patients (treatment of diseases), and the general public (prevention of disease). The underlying issues are not the information or database itself, but: a) the existence of a highly distributed, heterogeneous, unstructured and complex body of information; b) the unavailability of any knowledge structure or meta-data for human disease information, and c) the absence of an effective tool to support human disease information retrieval and knowledge sharing. To overcome these problems, we need an appropriated way to describe these distributed knowledge in a machine understandable way which can be achieved through ontologies.
One of these human diseases, that is the second leading cause of death in most developed countries and claims millions of lives worldwide each year, is cancer. The word cancer refers collectively to more than 100 different diseases, all characterized by the rampant and abnormal growth and spread of cells.
This thesis presents an introduction to medical informatics then it gives an overview on cancer informatics (chapter 2). Then it provides the main concepts of ontological engineering in medicine which includes: the meaning of ontologies, the main components of ontologies, the goals and benefits of ontologies and the different applications that ontologies can be used for, the different methodologies, languages and tools that are used for building ontologies and some of the medical ontologies currently exist (chapter 3). It also presents the process of building a web based liver cancer ontology (chapter 4) then it discusses the ontological engineering approach for creating a semantic web service for determining the stage of breast cancer (chapter 5). The thesis proposes an ontology-based multi-agent tool for cancer diseases knowledge management (chapter 6) and finally it describes the process of building a proposed ontology-based cancer diseases management system (chapter 7).
8.2.Conclusions
Through the studying of ontologies and their based systems implemented in this thesis, we can reach the following conclusions:
Concerning the methodologies used for building ontologies, it is found that many of the methodologies take a task as a starting point. This is obviously very useful because it focuses the acquisition, provides the potential for evaluation and provides a useful description of the capabilities of the ontology, expressed as the ability to answer well defined competency questions. On the down side, however, it seems to provide limitations to the re-use of the ontology. A complete methodology must provide guidelines to assist the ontological engineer in making choices at a variety of levels, from the high level structure of the ontology, to the fine detail of whether or not to include some particular distinction. Finally, it is quite clear that building ontologies is still a matter of craft skill rather than an understood engineering process. If ontologies are to realize their potential, it is important to clarify this practice, taking into account the variety of experience that is available, rather than basing the methodology too much on the experience of one or two projects.
Concerning ontology development languages, different ontology languages have different expressiveness and inference mechanisms. Knowledge representation paradigms underlying all these languages are diverse: frames, description logics, first (and second) order logic, conceptual graphs, semantic networks, production rules, etc. In many cases, they are even based on combinations of several formalisms.
Concerning ontology development tools, we have made a distinction between three groups of tools: web-based, computer-based, and client-server tools. Some of them are created for editing ontologies in a specific ontology language, and others are created as integrated extensible tool suites, most of which are language independent.
As for the liver cancer ontology, it is built using the Protégé-OWL editing environment and encoded in OWL-DL format. It was built using Top-Down approach. The main goals behind building this ontology are to allow finding and locating information about the liver cancer needed for interested users and domain experts, integrating information about the liver cancer to be accessed in an easy manner and providing the availability and accessibility of the liver cancer knowledge over the web. This ontology can be used by experts or medical researchers who want the liver cancer knowledge to be represented in a semantic way that allows reasoning.
Concerning the semantic web service for determining the stage of breast cancer, this service created using the OWL-S editor based on the breast cancer ontology which is encoded in OWL-DL format and has been built using the Protégé OWL ontology editor. This service can be used by medical students, patients as well as physicians to allow them to determine the stage of the breast cancer based on the symptoms of the patient.
One of the interesting multi-agent system applications can be found in health care. It can be used in patient scheduling and management, senior and community care, medical information access and management, and medical decision support. A new ontology-based multi-agent tool for cancer diseases knowledge management is proposed. It was developed using JADE framework. It can be adapted to be used for the diagnosis of cancer type or determining the stage of a specific cancer according to the provided symptoms. It also can provide treatment solutions. The tool allows patients, physicians and students to develop their own cancer-specific semantic management system that can be used for diagnosis, staging and treatment of cancer diseases.
Concerning the proposed ontology-based cancer diseases management system, it is implemented in Java using the Eclipse IDE for Java developers, the OWL-API and the FaCT++ reasoner. It can be used by patients, students and physicians to determine the cancer type, the stage of the cancer and recommends the treatment solutions. It contains three types of cancers, which are the lung cancer, the breast cancer and the liver cancer. It give about 92% correct classification.
8.3.Future Work
The thesis provides a new ontology-based multi-agent tool for cancer diseases knowledge management and an ontology-based cancer diseases management system. Currently the author has been developed three web based cancer ontologies which are the lung cancer ontology, the breast cancer ontology and the liver cancer ontology and there are more than 100 cancer types so the author aims, in future, to cover all the types of cancers by providing a web based ontology for each one (through a research team) to provide a complete medical system that can allow the management of cancer diseases in terms of diagnosis, staging and treatment.