Katharina Kaiser

How can information extraction ease formalizing treatment processes in clinical practice guidelines

OBJECTIVE Formalizing clinical practice guidelines (CPGs) for a subsequent computer-supported processing is a challenging, but burdensome and time-consuming task. Existing methods and tools to support this task demand detailed medical knowledge, knowledge about the formal representations, and a manual modeling. Furthermore, formalized guideline documents mostly fall far short in terms of readability and understandability for the human domain modeler. METHODS AND MATERIAL We propose a new multi-step approach using information extraction methods to support the human modeler by both automating parts of the modeling process and making the modeling process traceable and comprehensible. This paper addresses the first steps to obtain a representation containing processes which is independent of the final guideline representation language. RESULTS We have developed and evaluated several heuristics without the need to apply natural language understanding and implemented them in a framework to apply them to several guidelines from the medical subject of otolaryngology. Findings in the evaluation indicate that using semi-automatic, step-wise information extraction methods are a valuable instrument to formalize CPGs. CONCLUSION Our evaluation shows that a heuristic-based approach can achieve good results, especially for guidelines with a major portion of semi-structured text. It can be applied to guidelines irrespective to the final guideline representation format.

Modeling Treatment Processes Using Information Extraction

Clinical Practice Guidelines (CPGs) are important means to improve the quality of care by supporting medical staff. Modeling CPGs in a computerinterpretable form is a prerequisite for various computer applications to support their application. However, transforming guidelines in a formal guideline representation is a difficult task. Existing methods and tools demand detailed medical knowledge, knowledge about the formal representations, and a manual modeling. In this chapter we introduce methods and tools for formalizing CPGs and we propose a methodology to reduce the human effort needed in the translation from original textual guidelines to formalized processable knowledge bases. The idea of our methodology is to use Information Extraction methods to help in the semi-automation of guideline content formalization of treatment processes. Thereby, the human modeler will be supported by both automating parts of the modeling process and making the modeling process traceable and comprehensible. Our methodology, called LASSIE, represents a novel method applying a stepwise procedure. The general idea is to use this method to formalize guidelines in any guideline representation language by applying both general steps (i.e., languageindependent) and language-specific steps. In order to evaluate both the methodology and the Information Extraction system, a framework was implemented and applied to several guidelines from the medical subject of otolaryngology. The framework has been applied to formalize the guidelines in the formal Asbru plan representation. Findings in the evaluation indicate that using semi-automatic, stepwise Information Extraction methods are a valuable instrument to formalize CPGs.

Easing semantically enriched information retrieval-An interactive semi-automatic annotation system for medical documents

Mapping medical concepts from a terminology system to the concepts in the narrative text of a medical document is necessary to provide semantically accurate information for further processing steps. The MetaMap Transfer (MMTx) program is a semantic annotation system that generates a rough mapping of concepts from the Unified Medical Language System (UMLS) Metathesaurus to free medical text, but this mapping still contains erroneous and ambiguous bits of information. Since manually correcting the mapping is an extremely cumbersome and time-consuming task, we have developed the MapFace editor.The editor provides a convenient way of navigating the annotated information gained from the MMTx output, and enables users to correct this information on both a conceptual and a syntactical level, and thus it greatly facilitates the handling of the MMTx program. Additionally, the editor provides enhanced visualization features to support the correct interpretation of medical concepts within the text. We paid special attention to ensure that the MapFace editor is an intuitive and convenient tool to work with. Therefore, we recently conducted a usability study in order to create a well founded background serving as a starting point for further improvement of the editors usability.

Identifying treatment activities for modeling computer-interpretable clinical practice guidelines

Clinical practice guidelines are important instruments to support clinical care. In this work we analysed how activities are formulated in these documents and we tried to represent the activities using patterns based on semantic relations. For this we used the Unified Medical Language System (UMLS) and in particular its Semantic Network. Out of it we generated a collection of semantic patterns that can be used to automatically identify activities. In a study we showed that these semantic patterns can cover a large part of the control flow. Using such patterns cannot only support the modelling of computer-interpretable clinical practice guidelines, but can also improve the general comprehension which treatment procedures have to be accomplished. This can also lead to improved compliance of clinical practice guidelines.

Visualizing Complex Process Hierarchies during the Modeling Process

Clinical practice guidelines are documents that include recommendations describing appropriate care for the management of patients with a specific clinical condition, such as diabetes or chronic heart failure. Several representation languages exist to model these documents in a computer-interpretable and -executable form with the intention of integrating them into clinical information systems. Asbru is one of these representation languages that is able to model the complex hierarchies of these medical processes (called plans in Asbru). To allow their efficient evaluation and manipulation, they must be visualized in a compact and still clear form. This visualization must be integrated into an editing environment which makes changes to the process hierarchy easy and gives immediate feedback on the changes.

Identifying Condition-Action Sentences Using a Heuristic-Based Information Extraction Method

Translating clinical practice guidelines into a computer-interpretable format is a challenging and laborious task. In this project we focus on supporting the early steps of the modeling process by automatically identifying conditional activities in guideline documents in order to model them automatically in further consequence. Therefore, we developed a rule-based, heuristic method that combines domain-independent information extraction rules and semantic pattern rules. The classification also uses a weighting coefficient to verify the relevance of the sentence in the context of other information aspects, such as effects, intentions, etc. Our evaluation results show that even with a small set of training data, we achieved a recall of 75 % and a precision of 88 %. This outcome shows that this method supports the modeling task and eases the translation of CPGs into a semi-formal model.

Gaining process information from clinical practice guidelines using information extraction

Formalizing Clinical Practice Guidelines for subsequent computer-supported processing is a cumbersome, challenging, and time-consuming task. But currently available tools and methods do not satisfactorily support this task. We propose a new multi-step approach using Information Extraction and Transformation. This paper addresses the Information Extraction task. We have developed several heuristics, which do not take Natural Language Understanding into account. We implemented our heuristics in a framework to apply them to several guidelines from the specialty of otolaryngology. Our evaluation shows that a heuristic-based approach can achieve good results, especially for guidelines with a major portion of semi-structured text.

Formalizing `Living Guidelines' Using LASSIE: A Multi-step Information Extraction Method

Living guidelines are documents presenting up-to-date and state-of-the-art knowledge to practitioners. To have guidelines implemented by computer-support they firstly have to be formalized in a computer-interpretable form. Due to the complexity of such formats the formalization process is challenging, but burdensome and time-consuming. The LASSIE methodology supports this task by formalizing guidelines in several steps from the textual form to the guideline representation language Asbru using a document-centric approach. LASSIE uses Information Extraction technique to semi-automatically accomplish these steps. We apply LASSIE to support the implementation of living guidelines. Based on a living guideline published by the Scottish Intercollegiate Guidelines Network (SIGN) we show that adaptations of previously formalized guidelines can be accomplished easily and fast. By using this new approach only new and changed text parts have to be modeled. Furthermore, models can be inherited from previously modeled guideline versions that were added by domain experts.

Information requisition is the core of guideline-based medical care: which information is needed for whom?

RATIONALE, AIMS AND OBJECTIVES It is mandatory for the design of an efficient software product to know the different groups of users of a software tool, the tasks the users want to perform with it, and the information that is required for it. Our goal is to establish a comprehensive information source for the development of a consistent software environment supporting all tasks emerging from the creation to the execution of a computerized clinical practice guideline (CPG) for different user groups. METHODS We conducted a comprehensive literature review to investigate the different user groups of a computerized CPG as well as their specific information needs. RESULTS We provide a complete catalogue of every single aspect that may be related to information needs of any party concerned. In particular, we give detailed information on the tasks of guideline modellers on the one hand, and clinical information needs (i.e. information needs of physicians, nurses, nurse practitioners and patients) on the other hand. CONCLUSION By providing categorized information from several studies and publications, we establish an exhaustive information basis for the design of a useful software tool facilitating the formalization and the execution of a CPG.

Embedding the Evidence Information in Guideline Representation Languages

Clinical practice guidelines are widely used to support medical staff in treatment planning and decision-making, whereas, the classification of different recommendations in the CPGs are one of the most important information sources to use. However, there is a lack of consensus amongst guideline developers, regarding those classification schemes. To address this problem, we mapped the different graded and ungraded evidence information used by different guideline developing organizations into our meta schema. In this paper we describe how guideline representation languages, such as Asbru and PROforma can be extended to model our meta schema.