Use of Protg-2000 to Encode Clinical Guidelines Ravi D. Shankar, MS, - PDF document

Use of Protégé-2000 to Encode Clinical Guidelines Ravi D. Shankar, MS, Samson W. Tu, MS, and Mark A. Musen, MD, PhD Stanford Medical Informatics, Stanford University School of Medicine, Stanford, California, USA base. Abstract Several encoding tools have been developed, and they vary in their architectural flexibility and feature A major step in building guideline-based clinical support. Some are dedicated authoring environments care systems is encoding medical knowledge in for specific guideline models (e.g., Arezzo for guideline documents for interpretation by the PRO forma’s guideline model and AsbruView for computer. Guideline models provide the structure to Asbru [3]). Protégé-2000 [4], on the other hand, is the encapsulate guideline knowledge. Developers who latest in a series of completely general -purpose are familiar with the guideline model and domain programs that help users build knowledge acquisition experts who have relevant medical knowledge work systems. Pro tégé-2000’s extensible component-based as a team to encode guidelines using knowledge architecture and configurable GUI greatly facilitate acquisition tools. Protégé-2000, developed in our customizing knowledge-acquisition for given laboratory, is an environment for building domains. Domain experts can use the custom systems knowledge bases. It facilitates knowledge to record, browse and maintain domain knowledge in acquisition and maintenance. We have used knowledge bases. P rotégé-2000 is currently used in Protégé-2000 to encode clinical guidelines such as varied situations from formalizing clinical guidelines the JNC6 hypertension guidelines and cancer trial to modeling aviation knowledge. protocols. Many other modeling groups are also using it in their work. In this paper, we show that Protégé development has been going on for several Protégé-2000’s general-purpose knowledge- years, and is still continuing. Today Protégé has evolved to a point where it has become attractive to acquisition framework can provide a rich environment for to encode clinical guidelines. several guideline-modeling groups (e.g., EON, Prodigy [5], GLIF [6]). This paper illustrates the INTRODUCTION unique features of Protégé-2000 that strongly support the knowledge acquisition tasks. We discuss the trade-offs using Protégé-2000, a general-purpose A clinical guideline typically contains the guideline’s framework, to encode guidelines. intentions, medical background, patient eligibility criteria, procedural statements such as clinical A BRIEF ANATOMY OF PROTÉGÉ -2000 algorithms and drug recommendations, evidence for the advisories, treatment cost–benefit analyses, and literature references. Guidelines are generally Protégé-2000’s knowledge model [7] is frame-based. published in textual format via print and electronic A Protégé knowledge base consists of frames that media. In recent years, the healthcare community has represent classes, slots, facets, instances and shown an increased interest in automated support for constraints. Classes are concepts in the domain of guideline-based care. An important step in building discourse. They are abstract conceptual entities in such systems is encoding guidelines. This process the domain (e.g., the concept of an ACE Inhibitor ), or entails gleaning clinical knowledge out of paper named collections of instances (e.g., sets of guidelines and using a guideline model to Guidelines ). Slots are binary -relationships that encapsulate the knowledge in a computable describe properties or attributes of classes (e.g., the formalism. The guideline model may be explicit (as in eligibility criteria in the Guideline class). Facets EON [1]) or implicit (as in PRO forma [2]). describe properties of slots (e.g., the data type of the eligibility criteria slot). Instances of classes have Knowledge acquisition tools provide a structured specific slot values (e.g., in EON, environment for domain specialists to build guideline hypertension_guideline is an instance of the knowledge bases. These tools support the encoding Guideline class). Constraints specify additional process by providing access to controlled relationships among instances. vocabulary and/or user-defined medical concepts, by representing guideline-model concepts, by enabling A primary feature of Protégé-2000 is its ability to domain-specialists to enter, edit and browse guideline automatically generate a GUI form for each class knowledge via graphical user interfaces (GUIs), and based on the class definition. Domain experts can use by allowing rapid testing of an evolving knowledge

Protégé-2000. These widgets are visual metaphors that can be used to display the slot values and allow editing them. For example, there is a text field widget to store string slot values, and a check box widget to store Boolean slot values. Sophisticated custom- tailored widgets can also be associated with any data type to generate intuitive user-interfaces. When a form is generated, every slot on a class is automatically associated with a default widget on the form. The forms are laid out using simple rules such as “all slots of similar data types are grouped together”. Developers can use the forms layout editor to override the system’s default presentation by custom tailoring the form’s elements. Users can change the default widget association by choosing a new one from an extensible list of appropriate widgets for each slot type. Users can also modify the default form layout by moving the slot widgets around, by Figure 1: Automatically generated form for resizing the widgets, and so on. Guideline experts can specifying drug information. This GUI was generated then use these domain-specific forms to enter based on the Drug_Usage class definition. instances of the classes, thereby building the these forms to enter instances of classes. Besides the guideline knowledge base. forms, complex special-purpose user-interface Custom User-Interface Widgets widgets can also be added to the system as plug-ins to simplify knowledge acquisition (e.g., a diagram Complex domain information such as guideline widget to encode clinical algorithms). There is also a clinical algorithms warrants specialized visual notion of a Project , which contains configuration metaphors for acquiring it and browsing through it. information and form layout information. Protégé-2000 allows easy expansion of the standard Configuration information includes descriptions of all set of widgets to include custom domain-specific the user-interface widgets that have been added to widgets (e.g. a map widget for visualizing the project and a list of other projects that have been geographical information). A complex widget that is included by the current project. useful in encoding guidelines is the diagram widget. ENCODING CLINICAL GUIDELINES The Diagram Widget presents information graphically as a network of nodes and arcs. It comprises a set of nodes with connectors that join A component-based architecture and a facility to them. Nodes map to domain objects, and connectors expand the core functionality form the foundation of represent relationships between the objects. Diagram Protégé-2000. Developers can build domain-specific widgets form a natural visual metaphor for encoding a components to solve different knowledge acquisition tasks or simplify them, and to easily integrate them with the system. Thus, rich knowledge acquisition tools custom-tailored for encoding clinical guidelines can be built by assembling, with the core system, special-purpose user-interface widgets, utility functions, and even whole applications. Automatic Generation of User Interfaces When building a knowledge base, most user interaction with the Protégé-2000 GUI is done via forms — users create instances of a class by filling out the form associated with it. Every concrete class in the knowledge base has an associated form. Protégé-2000 generates user-interface forms automatically based on class definitions, by associating visual metaphors to the data types of the Figure 2: The Diagram Widget. The hypertension slots in each class, and by using some general layout guideline algorithm shown was built using the palette on heuristics (Figure 1). A list of user -interface widgets the right. is associated with each data type supported by