Showing papers presented at "Knowledge Representation for Health-Care in 2011"

Towards the automated calculation of clinical quality indicators

Abstract: To measure the quality of care in order to identify whether and how it can be improved is of increasing importance, and several organisations define quality indicators as tools for such measurement. The values of these quality indicators should ideally be calculated automatically based on data that is being collected during the care process. The central idea behind this paper is that quality indicators can be regarded as semantic queries that retrieve patients who fulfil certain constraints, and that indicators that are formalised as semantic queries can be calculated automatically by being run against patient data. We report our experiences in manually formalising exemplary quality indicators from natural language into SPARQL queries, and prove the concept by running the resulting queries against self-generated synthetic patient data. Both the queries and the patient data make use of SNOMED CT to represent relevant concepts. Our experimental results are promising: we ran eight queries against a dataset of 300,000 synthetically generated patients, and retrieved consistent results within acceptable time.

Patterns of clinical trial eligibility criteria

Abstract: Medical research would benefit from automatic methods that support eligibility evaluation for patient enrollment in clinical trials and design of eligibility criteria. In this study we addressed the problem of formalizing eligibility criteria. By analyzing a large set of breast cancer clinical trials we derived a set of patterns, that capture typical structure of conditions, pertaining to syntax and semantics. We qualitatively analyzed their expressivity and evaluated coverage using regular expressions, running experiments on a few thousands of clinical trials also related to other diseases. Based on an early evaluation we conclude that derived patterns cover the language of eligibility criteria to a large extent and may serve as a semi-formal representation. We expect that extending the presented method for pattern recognition with recognition of ontology concepts will facilitate generating computable queries and automated reasoning for various applications.

Task network based modeling, dynamic generation and adaptive execution of patient-tailored treatment plans based on smart process management technologies

Abstract: In this paper we present a knowledge-based, Clinical Decision Support System (OncoTheraper2.0) that provides support to the full life-cycle of both clinical decisions and clinical processes execution in the field of pediatric oncology treatments. The system builds on a previous proof of concept devoted to demonstrate that Hierarchical Planning and Scheduling is an enabling technology to support clinical decisions. The present work describes new issues about the engineering process carried out in the development and deployment of the system in a hospital environment (supported by a knowledge engineering suite named IActive Knowledge Studio, devoted to the development of intelligent systems based on Smart Process Management technologies). New techniques that support the execution and monitoring of patient-tailored treatment plans, as well as, the adaptive response to exceptions during execution are described.

A systematic analysis of medical decisions: how to store knowledge and experience in decision tables

Abstract: Diagnosis, treatment and prognosis are three different stages in the practice of medicine. All of these require some kind of decisions that are based on medical knowledge and the physician's experience, among other factors. In this paper, we first present a set of key questions that involve clinical decisions and that are faced by medical practitioners in a daily basis. We then discuss the type and codification of the information needed to answer these questions. Finally, we introduce a family of knowledge-experience decision tables to store this information and to provide answers to the aforementioned questions.

Diaflux: a graphical language for computer-interpretable guidelines

Abstract: In this paper, we introduce the formal representation language DiaFlux, that is simple and easy to use on the one hand. On the other hand it allows for the definition of Computer-Interpretable Guidelines (CIGs), that can solve valuable tasks being executed in the clinical context. Further, we describe a wiki-driven development process using the stepwise formalization and allowing for almost self-acquisition by the domain specialists. The applicability of the approach is demonstrated by a protocol for sepsis monitoring and treatment developed by a collaboration of clinicians.

Computing problem oriented medical records

Abstract: Problem Oriented Medical Record (POMR) is a medical record approach that provides a quick and structured acquisition of the patient's history. POMR, unlike classical health records, focuses on patient's problems, their evolution, and the relations between the clinical events. This approach provides the physician a view of the patients' history as an orderly process to solve their problems, giving the opportunity to make explicit hypotheses and clinical decisions. Most efforts regarding POMR focus on the implementation of information systems as an alternative of classical health records. Results reveal that POMR information systems provide a better organisation of patients' information but unsuitable mechanisms to perform other basic issues (e.g. administrative reports). Due to its features, POMR can help to bridge the gap between the traditional clinical information process and knowledge management. Despite the potential advantages of POMR, only few efforts have been done to exploit its capacities as a knowledge representation model and a further automatic reasoning. In this work, we propose the Problem Flow, a computational model based on the POMR. This proposal has a double objective: (1) to make explicit the knowledge included in the POMR for reasoning purposes and (2) to allow the coexistence between classical health records and the POMR. We also present PLOW, a knowledge acquisition tool which supports the proposed model. We illustrate its application in the Intensive Care Unit domain.

The human cli-knowme project: building a universal, formal, procedural and declarative clinical knowledge base, for the automation of therapy and research

Abstract: Currently, most clinical knowledge is in free text and is not easily accessible to clinicians and medical researchers. A major grand challenge for medical informatics is the creation of a distributed, universal, formal, sharable, reusable, and computationally accessible medical knowledge base. The required knowledge consists of both procedural knowledge, such as clinical guidelines, and declarative knowledge, such as context-sensitive interpretations of longitudinal patterns of raw clinical data accumulating from several sources. In this position paper, I first demonstrate the feasibility of such an enterprise, and explain in detail the overall lifecycle of a clinical guideline, by reviewing the main current components and their respective evaluations of one such comprehensive architecture for management of clinical guidelines: The Digital Electronic Guideline Library(DeGeL), a Web-based, modular, distributed architecture that facilitates gradual conversion of clinical guidelines from text to a formal representation in chosen target guideline ontology. The architecture supports guideline classification, semantic markup, context-sensitive search, browsing, run-time application to a specific patient at the point of care, and retrospective quality assessment. The DeGeL architecture operates closely with a declarative-knowledge temporal-abstraction architecture, IDAN. Thus, there is significant evidence that building a distributed, multiple-ontology architecture that caters for the full life cycle of a significant portion of current clinical procedural and declarative knowledge, which I refer to as "theHuman Clin-knowme Project," has become a feasible task for a joint, coordinated, international effort involving clinicians and medical informaticians.

Mammographic knowledge representation in description logic

Abstract: We present an advanced approach to representing knowledge about breast radiographs or mammograms which has advantages in terms of both usability and software engineering. The approach uses ontologies to create not merely a class hierarchy for a vocabulary but a full formal representation and, further, takes advantage of reasoning with description logic to provide application behaviour. The ontologies support a disjoint representation of graphical features and their interpretation in terms of medical findings. This separation of image features and medical findings allows the representation of different conceptual interpretations of the same graphical object, allowing different opinions of radiologists to be used in reasoning, which makes the approach useful for describing images to be used in computer-based learning and other applications. Three applications are discussed in detail: assessment of overlap in annotations, a conceptual consistency check in radiology training, and modelling temporal changes in parenchymal patterns. Reasoner usage, software testing, and implementation in Java are presented. The results show that, despite performance problems using the current implementations of reasoners, the description logic approach can be useful in practical applications.

Reasoning with effects of clinical guideline actions using OWL: AL amyloidosis as a case study

Abstract: We developed an ontology that allows representation and reasoning with effects of clinical actions. The ontology can support three important use-cases: (1) summarization and explanation of observed clinical states, (2) enhancing patient safety using safety rules, and (3) assessing guideline compliance. In this paper we focus on explanation of observed clinical states based on abductive reasoning that utilizes a causal network. We demonstrate our approach using examples taken from a guideline for management of amyloidosis.

Detecting dominant alternative interventions to reduce treatment costs

Abstract: Medical interventions can be compared attending to their health benefits and costs but also considering the similarity of the clinical actions involved. An intervention is a dominant alternative with respect to another intervention if it is better and cheaper. In this paper we introduce a hierarchy of medical actions that provides the semantics required by a methodology to detect dominant alternative interventions. After a formal introduction of this methodology, it is applied to analyze the data about the long term treatment of hypertension in the health care center ABS Vandellos-l'Hospitalet de l'Infant (Spain) in the years 2005-2009 in order to analyze feasible cost reductions after replacing medical interventions by their corresponding optimal, observed, dominant alternatives. This study shows that the use of this methodology reduces the average cost of each clinical encounter in €1.37.

Careflow personalization services: concepts and tool for the evaluation of computer-interpretable guidelines

Abstract: Healthcare professionals often need to adapt Clinical Guidelines to obtain personalized treatment plans for their patients. This adaptation is performed according to the patient's health state, but also takes into account implicit information (e.g., patients' preferences, local restrictions, etc). It may demand huge efforts, is error prone and the resulting guideline can contain contradictions. In this sense, Computer-Interpretable Guidelines associated with intelligent systems, offer additional support for healthcare professionals throughout the definition of a treatment plan. In this paper, we present an approach using Semantic Web Technologies to evaluate care actions and detect potential contradictions in the personalization process, taking into account the restrictions and needs for adaptation. This approach is based on the HL7 Reference Information Model, the UMLS Semantic Network and is compatible with most of Computer-Interpretable Guidelines formalisms. The prototype supporting our approach is also presented.

Analysis of treatment compliance of patients with diabetes

Abstract: The prevalence of diabetes is increasing worldwide. Despite the advances in evidence based therapies, patients with diabetes continue to encounter ongoing morbidity and diminished health-related quality of life. One of the reasons for the diminished benefit from therapy is medication noncompliance. Considerable evidence shows that a combination of therapeutic lifestyle changes (increased exercise and diet modification) and drug treatment can control and, if detected early enough, even prevent the development of diabetes and its harmful effects on health. However, despite the fact that type-2 diabetes is treatable and reversible with appropriate management, patients frequently do not comply with treatment recommendations. In this paper, we use a combination of Expectation Maximization (EM) clustering and Artificial Neural Network (ANN) modeling to determine factors influencing compliance rates, as measured in terms of medication possession ratio (MPR), among patients prescribed fixed dose combination therapy for type 2 diabetes.

Towards a Case-Based Reasoning Method for openEHR-Based Clinical Decision Support

Abstract: In 2007, a team of informaticians and specialists in dentistry in Sweden started a project to develop a CDSS based on openEHR for an oral disease named dry mouth Since openEHR is an emerging standard, designing a clinical decision support system (CDSS) based on it is an un-explored research area According to our findings, so far, very few (almost none) openEHR-based CDSSs have been released The methodological approach applied in developing an openEHR-based CDSS is presented in this paper This includes typical activities in developing CDSSs in addition to the activities one needs to carry out in order to develop an openEHR-based system In the first phase of this project, the focus has been on openEHR archetype design, knowledge acquisition, and choosing a suitable KRR method based on the available legacy patient records, ie a knowledge intensive case-based reasoning method, and the extracted general domain knowledge We also propose an architecture for such a system with the aim of benefiting from the structured openEHR-based patient data in reasoning