1.1       Introduction

From the very earliest moments in the modern history of the computer, scientists have dreamed of creating an ‘electronic brain’. Of all the modern technological quests, this search to create artificially intelligent (AI) computer systems has been one of the most ambitious and, not surprisingly, controversial.

It also seems that very early on, scientists and doctors alike were captivated by the potential such a technology might have in medicine (e.g. Ledley and Lusted, 1959). With intelligent computers able to store and process vast stores of knowledge, the hope was that they would become perfect ‘doctors in a box’, assisting or surpassing clinicians with tasks like diagnosis. With such motivations, a small but talented community of computer scientists and healthcare professionals set about shaping a research program for a new discipline called Artificial Intelligence in Medicine (AIM). These researchers had a bold vision of the way AIM would revolutionize medicine, and push forward the frontiers of technology.

In reviewing this new field in 1984, Clancey and Shortliffe provided the following definition: ‘Medical artificial intelligence is primarily concerned with the construction of AI programs that perform diagnosis and make therapy recommendations. Unlike medical applications based on other programming methods, such as purely statistical and probabilistic methods, medical AI programs are based on symbolic models of disease entities and their relationship to patient factors and clinical manifestations.


1.3       Problem Statement

It is unwise saying that the hospital is an exemption of the numerous life applications where manual operations pose some limitations with respect to effective service and productivity. Most women today face a lot of breast diseases due to lack of information or ignorance about several breast diseases, because there are no easy access to diagnose this disease on time. Therefore there is need for an implementation of an expert system on breast diseases diagnosis to detect these diseases on time and for early treatment.


1.4       Aims and Objectives the Study

The aim of the project is to design software that can be used to diagnose breast diseases on line.

The objective could be summarized as follows:

·         Aid health care providers responsible for large patient, the expert system can answer basic or general questions, leaving more time for individuals to patience with peculiar situations

·         Offering prompt feedback and self evaluation.

·         Providing potentially infinite array of information of the steps to take in the eventuality of a particular occurrence.

·         Aiding the nurses and other staff to know what to in the case of emergency if the human expert is not present at that point in time.

  • Providing potentially infinite array of information of the steps to take in the eventuality of a particular occurrence.
  • Aiding the nurses and other staff to know what to in the case of emergency if the human expert is not present at that point in time.

1.5       Justification of Study.

This research is justified because in real-time situations, an expert system attached to a monitor can warn of changes in a patient’s condition. In less acute circumstances, it might scan laboratory test results or drug orders and send reminders or warnings, thus alerting the physician of new developments in a patient’s conditions.

1.6       Scope of Study

This research work expert system on breast disease diagnosis system concentrates only on the diagonis of some breast diseses. Will be concernrd in registering patients and saving their records to the database.






Ever since man first learnt to communicate, knowledge that are to be shared and used by humans is most likely to be confined to what is stored in a person’s head or what the person can learn from another (Swett, 1991).Consider the following words by Sir William Osier (Wood, 1999): “Medicine is a science of uncertainty and an art of probability”.

Important components of the art of medicine are skills in repeatedly making decisions, formulating appropriate judgments and being comfortable with risk and uncertainty. Medical training, with its heavy emphasis on factual learning, often assigns a lesser priority to the study of decision making. Our own history of medicine contributes to dismissive attitudes about decision making. Before the later part of the 19th century, medical treatment was largely a matter of tradition, spurred on by a physician’s need to do something for the patient.



The future of MI as a profession is thus very promising. In other words, MI means managing medical and health care through information science and engineering technology. Like medicine, MI is also multidisciplinary. MI

deals with the entire domain of medicine and health care, from computer-based patient records to applications of image processing and from primary care practices to hospitals and regions of health care.

A few years ago, only a handful doctors had even heard of the term “health informatics.” Health informatics is a relatively new sub-speciality of medicine which uses information technology to manage clinical information. At a three-day eHealth Asia 2004 conference held at Kuala Lumpur in early April 2004, a local expert, Dr H.M.

Goh, council secretary of the Malaysian Health Informatics Association (MHIA) stated that there is space for growth in the local health informatics scene since few public and private hospitals have significant health management systems in place. The extraordinary thing about eHealth Asia 2004 was that it was attended by 350

participants (compared to 250 participants in 2001) which featured 54 speakers from over 20 countries around the world. This indicates that the field of health informatics has made itself felt throughout the world. Globally, health informatics include change management, artificial intelligence, messaging, mobile technology and the like. Only 10 of the 120 government hospitals are computerized, and only the Putrajaya and delayang hospitals have been fully-enabled with health informatics (The Star, 2004). With this upcoming awareness, the field of health informatics is very relevant to the Malaysian market. It is very strongly felt and believed that the research involvement in this study addresses a portion of and fits into this niche of health informatics.





The rapid evolution of technology and clinical research makes it difficult even for the specialist to keep up. In the light of this ‘information explosion’, it has been demonstrated that physicians do not always make optimal decisions. It has been mentioned earlier in the introduction that immense knowledge needs to be dissipated amongst health providers through in-depth training. Specifically in radiology, this strategy has been fairly effective in large academic centers but realistically, much has to be done by radiologists to practice state-of-the-art radiology at the forefront of radiological practice, especially in Malaysia. Although computers have proven to be very efficient and helpful in carrying out mundane tasks and the processing of data into useful information, its potential as a powerful technology can be further exploited to assist radiologists in knowledge processing.

The utilization of computers in decision-making can be employed in many different forms. However, the basic understanding to be realized and engraved in each and everyone’s mind is that these tools in decision-making have never been and are never intended in the first place to camouflage or belittle the decision makers in health care. Computers can be made as slaves to record huge amounts of detailed information. Simultaneously, these vast and abundant accumulated wealth of knowledge and information can be made available to radiologists at their disposal, put to use wherever or whenever abnormalities are encountered and ultimately arrive at a

more consistent decision-making. Some diagnoses can be made in a more quantitative, algebraic fashion although it cannot be denied that most radiological decision-making is very subjective. An expert is usually consulted for solving a difficult diagnostic problem. This situation Some diagnoses can be made in a more quantitative, algebraic fashion although it cannot be denied that most radiological decision-making is very subjective.

An expert is usually consulted for solving a difficult diagnostic problem. This situation and paradigm has served as a model for the birth of a class of computer systems that are known as expert systems. A KBS is designed to meet the knowledge gaps of the individual physician with specific patient problems. KBS and such other ES can be a boon to the rural health centres because even general medical practitioners can operate the systems. These are ideal examples of AI. has grown steadily since their introduction. Pham & Chen (2002) used applications of fuzzy logic in rule-based expert systems involving the problem of autofocusing camera lens system and also another on a financial decision system. Tocatlidou et al. (2002) built an ES that was capable of diagnosing plant diseases and disorders while Park & Storch (2002) shared a representation of ES in the shipbuilding industry which was able to downsize sizable development costs.Craker & Coenen (2006) proposed Knowledge Bazaar, the concept of which a paradigm for the development of ES and knowledge bases are created dynamically using knowledge supplied by self appointed internet communities. The philosophy underpinning the Knowledge Bazaar is the observation that knowledge can be accumulated, not from a limited number of experts or expert sources, but dynamically from internet users as they solve problems and offer advice. Mahmod et al. (2000) had shown the usage of neural networks combined with an expert system environment. Perhaps, all the relevant studies are best encapsulated in the paper by Liao (2005) where ES methodologies in almost all applications have been reviewed by the author for a span of a decade beginning from the year 1995.





Expert or knowledge-based systems are the most common type of artificial intelligence in medicine (AIM) system in routine clinical use. Indeed, it was in the medical area that expert systems have made their presence felt in the first place. AIMs contain medical knowledge, usually about a very specifically defined task and are able to reason with data from individual patients to eventually emerge with reasoned conclusions. Although there are many variations, the knowledge within an expert system is typically represented in the form of a set of rules (Keles & Keles, 2006).



Expert systems emerged as a branch of artificial intelligence – an amalgam of disciplines such as computer science, mathematics, engineering, philosophy and psychology. From the efforts of AI researchers, computer programs are developed that can reason as humans. ES are one of the most commercially viable branches of AI and although there have been reports of ES failures, surveys show that many parties have remained enthusiastic proponents of the technology and continue to develop important and successful applications in various fields (Duan et al., 2005).



From its early days of infancy when MYCIN (Negnevitsky, 2005) was first pioneered, ES have been developed in broad walks of life, in various areas and disciplines ranging from geology, statistics, electronics to medicine. In fact, the sky has no limit! To emphasize on this matter, a kaleidoscope of the expert systems developed in their respective fields is mentioned here. Williams (1991) suggested a prototype expert system for the design of complex statistical experiments. GEOPLAY (GEOPLAY, 2003) is a knowledge based expert system developed by the U.S. Geological Survey that is available for explorations in the oil and gas industry. Yang et al. (2005) developed an ES for vibration fault diagnosis of rotating machinery using decision tree and decision table and Duan et al. (2005) addressed the issues associated with the design, development and use of web-based ES from a standpoint of the benefits and challenges of developing and using them. Wagner et al. (2001) and Mak & Blanning (2003) applied ES to various problem domains and for the entry decisions of new products in business applications. The use of ES in business Other areas of specific medical applications of expert systems are in Obstetrics and Gynaecology (Medical Decisions, 2003), for leukemia management (Chae et al., 1998), for estimating the prognosis of head injured patients in intensive care unit (Sakellaropoulos & Nikiforidis, 2000), for heart valve diseases (Turkoglu et al., 2002), applied to brain MRI (Zhang & Maeda, 2000), even as early as the 1980’s to determine the irreversible cessation of all functions of the entire brain before any other organ transplantation (Pfurtscheller et al., 1988). Alonso et al. (2002) developed a medical diagnosis system, obtained by combining the expertise of a physician specialized in isokinetic and data mining techniques where patients may exercise one of their knee joints using basically a physical support machine according to different ranges of movement and at a constant speed. Lee et al. (1999), introduced a holistic system, which amalgamates case-based reasoning, rule-based reasoning, causal-based reasoning and an ontological knowledge base for managing clinical incidents in general practice enabling health professionals to share medical incident information, which has caused harm and may or can cause potential harm. The re-use of such information may prevent or mitigate human or medical errors. Morelli et al. (1987) and Solano et al. (2006) both described computational systems for automated diagnosis of depression and as an aid to clinical decision making in the mental health field. Verdaguer et al. (1992) investigated the application of ES in patients suffering from pneumonia, while Leong (1987) developed a system to detect irregularities by analyzing heart sounds through the interpretation and analysis of ausculatory findings. Costly and sometimes deadly clinical incidents may occur during the provision of health care, such as errors in dispensing inappropriate drugs due to the similarity of medication names to a patient for example. Lee et al. (1999) developed a prototype for this situation. In the same year, Li (1999) proposed a system to diagnose AIDS risky patients. While Lee & Lee (1991) suggested that future medical E.S. be specifically developed having at least one, if not all of these three characteristics i.e. simulate the performance of group of human experts, deal with chronic diseases and deal with several diseases simultaneously, Lhotska et al. (2001) focused on efficiency enhancements on rule based systems.



Breast cancer is among the leading causes of deaths in women worldwide. Its incidences have been rising at an alarming rate. More and more women have been subjected to the misery, suffering and pain caused by the disease. In Malaysia alone, approximately one in 20 women will be afflicted with breast cancer by the age of seventy, and by the age of 85, women have a one in eight chance of developing breast tumour. In the year 2000, almost 4,000 newly diagnosed cases emerge in the country. Of these, nearly 45% result in deaths, making it the number one cause of cancerrelated deaths among Malaysian women (Sunday Star, 2003). In a technical report drafted by the Ministry of Health Malaysia in the year 2001, 20% of patients afflicted by all kinds of the 1392 cancer cases have died from breast cancer alone. In its first report, the National Cancer Registry stated that 26,089 people were diagnosed with cancer in Peninsular Malaysia in the year 2002, of which 14,274 (55%) cases were cancers among women and 30.4% of it, were cancer of the breast (Mat Sakim, 2004). In Europe, 2004 estimates indicated 371,000 new cases with 129,900 breast cancer deaths. Mortality rates rose from 1951 until 1990 but fell noticeably in Western Europe, especially in the United Kingdom. However, this is not the case in Eastern and central Europe. Although rates in Hong Kong and Japan have been lower than those in Europe, they have also been increasing. Rates in North and South America are similar to Western Europe and so is Australia. The reasons for this decline in mortality rates in Western Europe, Australia and the Americas include the widespread practice of mammographic screening (Boyle et al., 2005).



Looking at the previous facts, we are immersed in a war, where the latent ‘enemies’ that we are confronted with in the battle against this killer disease is lying dormant out there, lurking and striking from unexpected corners. Indeed, we find ourselves in a difficult situation. In formulating the strategy best taken in this ‘war’, certain points as in the following, are noteworthy. Diagnosticians with the training and experience to interpret mammographic images are scarce. Therefore, there is an emphasis in training new radiologists to be able to interpret the mammographic images. The situation would be more crucial if mass screening were to be adopted as a national policy in this country as has been practiced in certain countries in the west. In the early period of a doctor’s professional activity, an expert system would prove valuable in minimizing the troubles that he or she might face due to inexperience. The existence of such facilities could be helpful especially for young radiologists or non-specialists. The existence of a diagnostic tool to aid in the interpretation process has been proven to be more useful for the junior than for the senior radiologists (Baileyguier et al., 2005). It is also a valuable teaching tool for the junior radiologists. Sensitivity also improved slightly for the senior radiologists. However, specificity remained unchanged in the study. An expert system for this application would make diagnostic expertise more widely and readily available in the clinical community. Therefore, the success of medical imaging depends on subjective factors that influence the ability of the observer to ‘ interpret the information’. These factors can be summarized into two broad classifications:

1. Those factors that are image dependent and relate to the visual conspicuity of features relevant to the clinical problem; and

2. Those that are image independent; are primarily cognitive in nature and relate to what the observer knows about the visual information in front of him.

Variation between readers was greater than the differences between imaging techniques (Manning et al., 2005). There are many image acquisition, display and processing parameters, and their effects on optimizing images for human interpretation are largely unknown. But we know less still, allowing the observer to structure the task

of interpreting image features; perhaps a better understanding of these factors now deserves our research attention so that we can achieve a better match of image displays to cognitive/perceptual skills. The availability of Computer Aided Detection(CAD) and Computer Aided Diagnosis(CADx) should be employed with a word of caution. That is, it is important that the development and availability of such systems do not detract from quality and the need for radiological skills across the imaging workforce. In other words, the skill of radiologists using such CAD and CADx remains paramount. Maintaining high radiological skill levels whilst using technology efficiently and effectively to formulate correct diagnostic decisions quickly is a key issue for the future (Manning et al., 2005).

Even though the Breast Imaging Reporting and Data System (BI-RADS) was introduced to help standardize feature analysis and final management of breast modality findings, there still exists variations in their interpretations. Continued efforts to educate radiologists to promote maximum consistency still need to be carried out (Lehman et al., 2002). The risk of breast cancer increases with age. Considerable evidence indicates that older women frequently do not undergo mammography. Offering on-site mammography at community-based sites where older women gather is an effective method for increasing breast cancer screening rates among older women (Reuben et al., 2002). It is hoped that this work may be useful in filtering only the abnormal cases to be further scrutinized by specialists. Routine and repetitive use of computer-based systems developed for experiments would bring several benefits. Radiologists could be trained to evaluate the perceptual features appropriately (D’Orsi et al., 1992). In clinical practice, only 15-30% of patient referred for biopsy are found to have a malignancy (Hadjiiski, 2004). Unnecessary biopsies increase health care costs and may cause patient anxiety and morbidity. It is therefore important to improve the accuracy of interpreting mammographic lesions (Hadjiiski, 2004), thereby improving the positive predictive values of detection modalities.

As the expert system contains specific rule base for the differentiation of breast diseases, it may be utilized both to help train physicians in breast cancer modalities and to promote a more consistent mammographic and ultrasound interpretation. The criterion for interpreting imagery is subjective and variable. With the help of an expert

system, the diagnostic criteria can be made more explicit. This would serve as a basis for consistent and reproducible diagnoses. At the same time, it would also form the basis for discussion and further research to improve the validity of the diagnostic criteria. Expert systems would serve as models with intelligent behavior in cognitive and perceptual realms and skills to solve problems thought to require human intelligence. People are better at clarifying a problem, suggesting kinds of procedures to follow, judging the reliability of facts and deciding if a solution is reasonable. The problem solver must know how to use knowledge and see patterns in the signals

presented. To sum up, the following points are relevant:

• The human heuristic approach of combining evidence to reach a prognosis can deal successfully with a limited amount of evidence. The proliferation of large databases of patient findings, due to the increased use of computers in clinical settings, offers an abundance of available data, challenging the limited human capacity for indirect inference. Decision support systems that are able to model uncertainty and analyze diverse sources of information can therefore become auseful tool for medical experts (Sakellaropoulos & Nikiforidis, 2000).

• Some of the most successful applications have been for instruction e.g. use of a medical expert system to develop diagnostic skills thus encouraging students to structure knowledge and process it systematically in response to a problem or abnormality. Also, as precise analytical models of knowledge and through the ways in which they are used, expert systems can enhance our understanding of human decision-making processes.

• As clinical decision making inherently requires reasoning under uncertainty, expert systems will be suitable techniques for dealing with partial evidence and with uncertainty regarding the effects of proposed interventions (Shortliffe,1987).

• Radiology is gradually developing a more systematic approach to training, replacing the traditional mixture of ad hoc apprenticeship and formal lectures with a combination of structured tuition and case-based experiential learning. This is intended to meet a long-recognized need for clinicians to encapsulate general medical knowledge within the development of skills through diagnostic practice. A structured approach to training can have the additional benefit of equipping learners with a coherent ‘conceptual framework’: an appropriately defined and organized notation that enables them to externalize, reflect on and share diagnostic knowledge (Structured Computer-based Training, 2005).

• Radiological expertise is based on two kinds of skills: the swift and accurate processing of normal appearance, and the ability to distinguish disease from normal variation in appearance. Thus, skill development in radiology requires exposure to, and reporting of a large range of images, so that recognition of varied normal anatomy are firmly etched in the minds of the skilled interpreters and cognitive resources can be devoted to the process of describing abnormal appearances (Structured Computer-based Training, 2005).

• Despite the wide applications of AI techniques to a range of clinical activities, few expert systems have been implemented in the field of medical imaging; its scarcity possibly due to the inherent difficulty in high-level vision. The data acquired from medical scanners can be noisy and ambiguous. Nevertheless, the potential benefits make it tempting to aim at designing expert systems using the digital images provided by the various modalities, especially with the advent of networking of medical images through PACS and the DICOM format. DICOM is an international standard, recognized by most hardware and software manufacturers for the storing and transmission of medical images acquired with all modalities (Chabat et al., 2000).   

This work is an attempt to fulfill or partially fulfill the dearth of imaging expert systems and with the inadvertent and inevitable emergence of digital mammography, radiologists would need to undergo pertinent retraining (Digital Imaging, 2004). Hence, this work will all the more be relevant.



Man strives to augment his abilities by building tools, from the invention of the club to lengthen his reach and strengthen his blow to the refinement of the electron microscope to sharpen his vision, tools have extended his ability to sense and to manipulate the world about him. (Szolovits, P. 1982) Today we stand on the threshold of new technical developments which will augment man’s reasoning, the computer and the programming methods being devised for it are the new tools to effect this change.

According to Schwartz, 1970 in his book AI and Medicine he opined that medicine is a field in which such help is critically needed, our increasing expectations of the highest quality health care and the rapid growth of ever more detailed medical knowledge leave the physician without adequate time to devote to each case and struggling to keep up with the newest developments in his field. For lack of time, most medical decisions must be based on rapid judgments of the case relying on the physician’s unaided memory. Only in rare situations can a literature search or other extended investigation be undertaken to assure the doctor (and the patient) that the latest knowledge is brought to bear on any particular case. Continued training and recertification procedures encourage the physician to keep more of the relevant information constantly in mind, but fundamental limitations of human memory and recall coupled with the growth of knowledge assure that most of what is known cannot be known by most individuals, it  is the opportunity for new computer tools: to help organize, store, and retrieve appropriate medical knowledge needed by the practitioner in dealing with each difficult case, and to suggest appropriate diagnostic, prognostic and therapeutic decisions and decision making techniques. (Schwartz, 1970)

The key technical developments leading to this reshaping will almost certainly involve exploitation of the computer as an ‘intellectual,’ ‘deductive’ instrument–a consultant that is built into the very structure of the medical-care system and that augments or replaces many traditional activities of the physician. Indeed, it seems probable that in the not too distant future the physician and the computer will engage in frequent dialogue, the computer continuously taking note of history, physical findings, laboratory data, and the like, alerting the physician to the most probable diagnoses and suggesting the appropriate, safest course of action (Doyle J, 1979).

This vision is only slowly coming to reality. The techniques needed to implement computer programs to achieve these goals are still elusive, and many other factors influence the acceptability of the programs.


Bleich 1999, defined Artificial Intelligence as the study of ideas which enable computers to do the things that make people seem intelligent. The central goals of Artificial Intelligence are to make computers more useful and to understand the principles which make intelligence possible. (Bleich, 1999).

This is a rather straightforward definition, but it embodies certain assumptions about the idea of intelligence and the relationship between human reasoning and computation which are, in some circles, quite controversial. The coupling of the study of how to make computers useful with the study of the principles which underlie human intelligence clearly implies that the researcher expects the two to be related. Indeed, in the newly-developing field of cognitive science, computer models of thought are explicitly used to describe human capabilities (Ituma C.2010). Is it Possible for Computing Machines to Think?

(Gorry, G. A., “On the Mechanization of Clinical Judgment,” 1976,) Nor if one defines thinking as an activity peculiarly and exclusively human, any such behavior in machines, therefore, would have to be called thinking-like behavior. Nor if one postulates that there is something in the essence of thinking which is inscrutable, mysterious, mystical. Yes if one admits that the question is to be answered by experiment and observation, comparing the behavior of the computer with that behavior of human beings to which the term “thinking” is generally applied.

AI in Medicine (AIM) is AI specialized to medical applications, researchers in AIM need not engage in the controversy introduced above. Although we employ human- like reasoning methods in the programs we write, we may justify that choice either as a commitment to a human/computer equivalence sought by some or as a good engineering technique for capturing the best-understood source of existing expertise on medicine–the practice of human experts. Most researchers adopt the latter view.

The choice to model the behavior of a computer expert in medicine on the expertise of human consultants is by no means logically necessary. If we could understand the functioning in health and in disease of the human body in sufficient depth to model the detailed disease processes which disturb health, then, at least in principle, we could perform diagnosis by fitting our model to the actually observable characteristics of the patient at hand. McCorduck, P., in his book “Computers Who Think” 1980 argued further, that we could try out possible therapies on the model to select the optimum one to use on the patient. Unfortunately, although biomedical research strives for such a depth of understanding, it has not been achieved in virtually any area of medical practice. The AIM methodology does not dogmatically reject the use of non-human modes of expertise in the computer. Indeed, accurate computations of probabilities and solutions of simple differential equations tasks at which human experts are rather poor without special training play a role in some of our programs. Nevertheless, most of what we know about the practice of medicine we know from interrogating the best human practitioners; therefore, the techniques we tend to build into our programs mimic those used by our clinician informants. (McCorduck, P., 1980)

Relying on the knowledge of human experts to build expert computer programs is actually helpful for several additional reasons: First, the decisions and recommendations of a program can be explained to its users and evaluators in terms which are familiar to the experts. Second, because we hope to duplicate the expertise of human specialists, we can measure the extent to which our goal is achieved by a direct comparison of the program’s behavior to that of the experts. Finally, within the collaborative group of computer scientists and physicians engaged in AIM research, basing the logic of the programs on human models supports each of the three somewhat disparate goals that the researchers may hold: (Schwartz, 1970)

  • To develop expert computer programs for clinical use, making possible the inexpensive dissemination of the best medical expertise to geographical regions where that expertise is lacking, and making consultation help available to non-specialists who are not within easy reach of expert human consultants.
  • To formalize medical expertise, to enable physicians to understand better what they know and td give them a systematic structure for teaching their expertise to medical students.
  • To test AI theories in a “real world” domain and to use that domain to suggest novel problems for further AI research.