Diagnostic Expert System Research Articles

Analysis of ultrasonographic prostate images for the detection of prostatic carcinoma: The Automated Urologic Diagnostic Expert system

This paper describes a study on the automated analysis of ultrasonographic prostate images. With image processing, tissue characterization in the prostate was performed to assess the probability of malignancy. During prostate examinations, images were recorded at the positions where biopsies were taken. The used samples were divided into three groups. Two of them were used for the construction of a classification tree, and the third was used for the evaluation of this classification. A sensitivity of 80.6% and specificity of 77.1% were reached retrospectively. In a prospective way, these results were 80.0% and 88.2%, respectively. The prospective predictive value for cancer detection was 85.7%. The presented prospective value for image analysis was almost twice as high as the values normally found for prostate examination.

Pitch expert: A productivity improvement tool for pulp production

The production of pulp from wood is a billion dollar international industry. Pitch deposits in pulp mills result in inferior quality or significant downtime which translates into hundreds of millions of dollars in lost revenue. Various techniques exist to combat pitch deposition during the pulping process, but they must be carefully applied in order to avoid complications which may arise. Pitch Expert is a diagnostic expert system developed by CRIM (Centre de recherche informatique de Montreal) and its associate member, PAPRICAN (Pulp and Paper Research Institute of Canada). Its development involved the use of a complex knowledge base development tool combined with many highly specialized and customized knowledge structure manipulation modules. In addition, the final system incorporates a framework for continued system maintenance and expansion by PAPRICAN personnel. The system was designed, developed and built over a 3-year span and was delivered as of April 1991. Following its delivery Pitch Expert was tested with the cooperation of several pulp mills with very successful results. As of this writing it is about to be released for general use to most kraft pulp mills in Canada.

Diagnosis of selected pulpal pathoses using an expert computer system

The diagnosis of dental pain that results from a pulpal pathosis may prove to be a confusing and complex issue for both dental students and experienced clinicians alike. A computerized diagnostic expert system, COMENDEX, was developed to aid in the clinical diagnosis of pulpal pathosis and to provide a rapid, accurate second opinion when a human consultant is not readily available. The vast majority of diagnostic expert systems use a single reasoning methodology as their inference mechanism. COMENDEX is a prototype hybrid expert system that combines and exploits the best aspects of both rule-based and statistical reasoning methodologies. The COMENDEX endodontic diagnostic system was tested to determine the validity and accuracy of its diagnoses by the use of a variant of Turing's test and weighted κ statistic. The promising result obtained from the initial tests suggests that an expert system using this type of hybrid reasoning methodology is well suited to the area of endodontic diagnosis and may prove highly successful if expanded to include other problem areas in oral diagnosis.

Towards competent information acquisition interactions between an expert system and its user

Expert systems are consultative, highly interactive systems, and hence the quality of interaction between system and user is important for the acceptability of the system. Acquisition of data about the problem in hand is a central feature of the interaction between system and user, and it makes a major contribution to the user's perception of the system. It is therefore crucial that the acquisition of this data (both as individual data items and as a sequence of data inputs) is perceived by the user as competent. This paper identifies central, domain-independent, design goals for the information-acquisition interactions between an expert system and its user, including mixed-initiative interaction, flexibility in user input, and system competence in querying the user. The paper discusses the realisation of these goals in a diagnostic expert system, Skeletal Dysplasias Diagnostician, through an explicit data model which allows for the representation of data with temporal and spatial (locality) aspects and the decide-status function which operates on the data model.

D‐KAT: a deep knowledge acquisition tool

Abstract: This paper describes a system of shallow and deep knowledge acquisition and representation for diagnostic expert systems. The acquisition system is integrated into a diagnostic expert system shell. Shallow knowledge is represented in a failure model as a set of cause‐effect relations among the possible faults, while deep knowledge is represented in three deep models: a functional, a deep causal and a taxonomic model. The acquisition and the representation of all the models are fully integrated. The deep knowledge is used by the final expert system in order to provide the user with deep explanations of the cause‐effect relations of the failure model.

Development of a shell for constructing a diagnostic expert system based on designer's knowledge

Conventionally, diagnostic expert systems have been developed using maintenance engineer's knowledge. However, these systems have many problems, such as the requirement for failure data and experience of maintenance for objective equipment, etc. This paper presents a method of knowledge representation and systematization of knowledge for diagnostic expert systems which can be applied to new equipment having no historical data. The use of design knowledge and the relations between components are considered in terms of necessary and possible relation. A method of knowledge representation, termed a dual direction diagnostic matrix, is introduced. As a result, it is confirmed that speedy and concise diagnosis can be implemented.

A method for putting strategic common sense into expert systems

Strategic common sense is defined as the heuristic considerations human experts often apply to decide which choice is the most opportune to make from a set of possible alternatives. A method for acquiring modeling, and representing human strategic common sense in diagnostic expert systems is presented. The preference parameter and preference criterion tools, which are used to define a basic level of strategic knowledge, are reviewed. Strategic common sense and its representation using general rules and metarules modeling are discussed. An interview with a physician who provided a medical case and the related solutions is summarized. It is shown that the solutions given by the physician match the conclusions obtained by applying the strategic common sense rules to the medical example.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

A transputer-based parallel expert diagnostic system

In this paper we present an expert system to make the real time diagnosis of an experimental apparatus. To overcome the drawbacks due to the time constraints required by a real time expert system a diagnostic apparatus based on a transputer network has been developed. The parallelization involves different parts of the diagnostic system: (1) the sensor data acquisition (2) the data interpretation, (3) the expert diagnosis. VME and PC boards have been used and the software has been written in Occam, parallel C and parallel Prolog. Experimental results of the system performances are presented.

A customized network diagnostic expert system based on general diagnostic strategies

The objective of this research is to investigate the common link in diagnostics of different networks. Are there any underlying common expertise and strategies that are applicable to various connection-based networks? If so, how is the general expertise related to network-specific expertise? Why is it significant to identify and figure out the general expertise?

Graphical knowledge acquisition for medical diagnostic expert systems

Like many textbook authors use text systems for writing their books, expert system authors should have easy to use knowledge acquisition systems for entering and testing their knowledge bases by themselves without much help from ‘knowledge engineers’. In this paper, we report on a graphical knowledge acquisition tool (CLASSIKA) based on an expert system shell for heuristic classification (MED2) and designed for direct use by domain experts. We demonstrate how the system has been used for building a rather large expert system for diagnosing rheumatology diseases which is now being tested in clinical use.

Generation of Diagnostic Rules via Inductive Machine Learning

A learning algorithm called CLILP2 (Cover Learning Using Integer Linear Programming) is applied to medical data to generate rules to recognize patients with coronary artery disease. The algorithm partitions a data set into subsets using features which best describe and distinguish a particular subset from all other subsets. These features are used to form the rules which can be used as the knowledge base of a diagnostic expert system. Results from the application of the algorithm to coronary artery stenosis data are compared with the results obtained from the existing expert system.

Logic based feature detection on incore neutron spectra

A general framework for detecting features of incore neutron spectra with a rule-based methodology is presented. As an example, we determine the meaningful peaks in the APSD-s. This work is part of a larger project, aimed at developing a noise diagnostic expert system.

Diagnostic Expert Systems: Encoding Geological Knowledge for an Exploration Play Analysis

The U.S. Geological Survey is currently exploring the feasibility of applying diagnostic expert systems and knowledge-based acquisition techniques to the geologic analysis of sedimentary basins and petroleum exploration plays. This paper describes a unique approach to the design and application of a rule-based expert system to the analysis of exploration plays, a methodology commonly used to assess petroleum resources in a basin. This technique consists of using various geologic models and Monte Carlo simulation or probabilistic methods for analyzing geologic and reservoir conditions favorable to the occurrence of petroleum resources. Work is in progress to adapt the geologic model used in the play analysis to an expert-systems and knowledge base, the object being to capture the logic and reasoning used to characterize the geologic model and to evaluate the play. Expert systems techniques now being developed are capable of dealing with inexact reasoning or reasoning under uncertainty by incorporating degrees of uncertainty to deal effectively with incomplete, inferred, or interpretive data involved in the knowledge base. The goal of this study is to provide an embedded diagnostic expert-system approach for characterizing the geologic model to expand upon the play-analysis system. Such a system provides the geologist with capabilities to documentmore » major basin components, such as stratigraphy, structural geology, and sedimentation, and to analyze the traditional concepts of source, reservoir, and trapping mechanisms. This expert system will assist the geologist in compiling and interpreting the geologic and reservoir engineering data necessary for running the probabilistic methods to calculate the amount of probable petroleum resources within each play. Diagnostic expert-systems technology can provide challenging new tools in knowledge acquisition and data interpretation in petroleum geology.« less

Fuzzy Diagnostic Expert System for Ultrasonic Testing. Construction of Personal-Computer-Aided System and Supplemental Knowledge for Parastic Echo.

A diagnostic fuzzy expert system for ultrasonic testing (UT) which has been constructed by the authors shows good agreement between the results obtained by the system and by UT experts. In this study, we attempt to construct a personal-computer-aided system and to supplement knowledge of the parastic echo. In addition, the computational time required for inference is reduced because the blackboard model used in the proposed system is more improved than authors' former models. The results obtained with the proposed fuzzy expert system are compared with those obtained by UT experts, and it is confirmed that the system has the capacity to discriminate between parastic echo and defects.

Converting a rule-based expert system into a belief network.

The theory of belief networks offers a relatively new approach for dealing with uncertain information in knowledge-based (expert) systems. In contrast with the heuristic techniques for reasoning with uncertainty employed in many rule-based expert systems, the theory of belief networks is mathematically sound, based on techniques from probability theory. It therefore seems attractive to convert existing rule-based expert systems into belief networks. In this article we discuss the design of a belief network reformulation of the diagnostic rule-based expert system HEPAR. For the purpose of this experiment we have studied several typical pieces of medical knowledge represented in the HEPAR system. It turned out that, due to the differences in the type of knowledge represented and in the formalism used to represent uncertainty, much of the medical knowledge required for building the belief network concerned could not be extracted from HEPAR. As a consequence, significant additional knowledge acquisition was required. However, the objects and attributes defined in the HEPAR system, as well as the conditions in production rules mentioning these objects and attributes, were useful for guiding the selection of the statistical variables for building the belief network. The mapping of objects and attributes in HEPAR to statistical variables is discussed in detail.

ENIGMA: a system that learns diagnostic knowledge

The results of extensive experimentation aimed at assessing the concrete possibilities of automatically building a diagnostic expert system, to be used in-field in an industrial domain, by means of machine learning techniques, are described. The system, ENIGMA, is an incremental version of the ML-SMART system, which acquires a network of first-order logic rules, starting from a set of classified examples and a domain theory. An application is described that consists of discovering malfunctions in electromechanical apparatus. ENIGMA's efficacy in acquiring sophisticated knowledge and handling complex structured examples is largely due to its underlying database management system, which supports the learning operators, defined at the abstract level, with a set of primitives, taken from the field of deductive databases. An expert system, MEPS, devoted to the same task, has also been manually developed. A number of comparisons along different dimensions of the manual and automatic development process have been possible, allowing some practical indications to be suggested. >

Effects of Using a Diagnostic Rule-Based Expert System Developed for the Nuclear Industry

This experiment, conducted at the OECD Halden Reactor Project, Halden, Norway in the spring 1991, aimed to assess the effect on nuclear power plant operators diagnostic behaviour when using a rule based diagnostic expert system. The rule based expert system used in the experiment is called DISKET (Diagnosis System Using Knowledge Engineering Technique) and was originally developed by the Japan Atomic Energy Research Institute (JAERI). The experiment was performed in the Halden man-machine laboratory using a full scope pressurized water reactor simulator. Operator performance in terms of quality of diagnosis is improved by the use of DISKET. The use of the DISKET system also influences operators problem solving behaviour. The main difference between the two experimental conditions can be characterised as while the DISKET users during the diagnosis process are following a strategy which is direct and narrowed, the non-DISKET users are using a much broader and less focused search when trying to diagnose a disturbance.

KASE project toward effective diagnosis system development

The Knowledge Acquisition Support Environment (KASE) and some case studies on expert system development are described. DiPROS is a diagnostic expert system building tool, which has a knowledge base editor and an inference engine specialized for diagnostic tasks. DiKAST is an interactive knowledge-acquisition (KA) tool, which guides repairs on defects in the knowledge base. In order to build this KA facility, many kinds of KA interview strategies were formalized. We applied KASE tools in building a defect diagnosis system for a color picture tube manufacturing plant. The prototype system was able to be developed in a short time by the domain experts alone, who had no previous experience in computer programming.

Performance of a diagnostic system (Iliad) as a tool for quality assurance

Quality assurance improves health care through detection of quality problems and feedback to the care giver. Current review procedures employed by the Peer Review Organizations (PROs), however, appear to underdetect quality problems, particularly those arising from diagnostic errors. We studied the use of an expert diagnostic system, Iliad, to detect quality problems arising from diagnostic errors. One hundred cases were selected from among those Medicare cases reviewed by the Utah PRO (UPRO) and which contained diagnoses recognized by Iliad. Iliad flagged 28 cases out of the 100 as containing diagnostic errors, and a gold standard physician review confirmed quality problems in 17 cases (60.7%). The UPRO review found 28 cases with quality problems, mostly treatment and documentation errors. The quality problems detected by Iliad appeared to be more serious than those detected by the UPRO review. Among the six cases with quality problems detected by both the UPRO and the Iliad review, there was none for which the same quality problem was detected by the two procedures. The two review procedures were therefore complementary.

Explanation-based learning as justification of knowledge

Abstract: This paper describes the use of the explanation-based learning (EBL) machine learning technique in the practical domain of knowledge acquisition for expert systems. A knowledge acquisition tool, EBKAT (Explanation-Based Knowledge Acquisition Tool), is described, which may be used in the development of knowledge bases for diagnostic expert systems. The functioning of EBKAT attempts to combine the full potential of a domain expert's skills and the power of explanation-based machine learning techniques. The EBL component is not employed in the acquisition of the knowledge base rules but is used to justify the knowledge entered and to relate it to the knowledge already in the system. It is suggested that the EBKAT tool goes some way towards overcoming the knowledge acquisition bottleneck and results in the acquisition of knowledge which is rich in contextual information.