Skip to main content
Disciplines
Automotive Business IT + Informatics Construction + Real Estate Electrical Engineering + Electronics Energy + Sustainability Insurance + Risk Finance + Banking Management + Leadership Marketing + Sales Mechanical Engineering + Materials
Events
DE
EN
Books
Journals
Topic Page
Marketing
Start single access now
Access for companies
EXTENDED SEARCH
Log in
Top

2011 | Book

Ontology-Based Retrospective and Prospective Diagnosis and Medical Knowledge Personalization Read chapter Read first chapter

Knowledge Representation for Health-Care

ECAI 2010 Workshop KR4HC 2010, Lisbon, Portugal, August 17, 2010, Revised Selected Papers

Editors: David Riaño, Annette ten Teije, Silvia Miksch, Mor Peleg

Publisher: Springer Berlin Heidelberg

Book Series : Lecture Notes in Computer Science

Part of: Springer Professional "Wirtschaft+Technik" , Springer Professional "Technik" , Springer Professional "Wirtschaft"

Login to get access
insite
SEARCH

About this book

This book constitutes the proceedings of the KR4HC 2010 workshop held at ECAI in Lisbon, Portugal, in August 2010. The 11 extended papers presented were carefully reviewed and selected from 19 submissions. The papers cover topics like ontologies, patient data, records, and guidelines, and clinical practice guidelines.

Table of Contents

Frontmatter

Ontologies

Ontology-Based Retrospective and Prospective Diagnosis and Medical Knowledge Personalization
Abstract
Computers can be helpful to support physicians in medical diagnosis and health care personalization. Here, a health care ontology for the care of chronically ill patients that was created and validated in the k4care project is used in prospective and retrospective diagnoses, and also in the personalization of medical knowledge. This paper describes the technical aspects of these three ontology-based tasks and the successful experiences in their application to deal with wrong diagnoses, comorbidites, missing data, and prevention.
Cristina Romero-Tris, David Riaño, Francis Real
A Semantic Web Approach to Integrate Phenotype Descriptions and Clinical Data
Abstract
Integrating phenotype descriptions from text-rich research resources, such as OMIM, and data from experimental and clinical practice is one of the current challenges to promote translational research. Exploring new technologies to uniformly represent biomedical information is needed to support integration of information drawn from disparate sources. Positive progress to integrate data requires to propose solutions supporting fully semantic translations. The Semantic Web is a promising technology, so international efforts, such as the OBO Foundry, are developing ontologies to support annotation and integration of scientific data. In this paper, we show an approach to get concordances between phenotype descriptions and clinical data, supported by knowledge adapters based on description logic and semantic web rules. This integration provides a valuable resource for researchers in order to infer new data for statistical analysis.
María Taboada, María Jesús Sobrido, Verónica Colombo, Belén Pilo
Ontology-Based Knowledge Modeling to Provide Decision Support for Comorbid Diseases
Abstract
Handling comorbid diseases in a decision support framework is a challenging problem as it demands the synthesis of clinical procedures for two or more diseases whilst maintaining clinical pragmatics. In this paper we present a knowledge management approach for handling comorbid diseases by the systematic alignment of the Clinical Pathways (CP) of comorbid diseases. Our approach entails: (a) knowledge synthesis to derive disease-specific CP from evidence-bases sources; (b) knowledge modeling to abstract medical and procedural knowledge from the CP; (c) knowledge representation to computerize the CP in terms of a CP ontology; and (d) knowledge alignment by aligning multiple CP to develop a unified CP knowledge model for comorbid diseases. We present the COMET system that provides decision support to handle comorbid cardiac heart failure and atrial fibrillation.
Samina Raza Abidi

Patient Data, Records, and Guidelines

Inducing Decision Trees from Medical Decision Processes
Abstract
In medicine, decision processes are correct not only if they conclude with a right final decision, but also if the sequence of observations that drive the whole process to the final decision defines a sequence with a medical sense. Decision trees are formal structures that have been successfully applied to make decisions in medicine; however, the traditional machine learning algorithms used to induce these trees use information gain or cost ratios that cannot guarantee that the sequences of observations described by the induced trees have a medical sense. Here, we propose a slight variation of classical decision tree structures, provide four quality ratios to measure the medical correctness of a decision tree, and introduce a machine learning algorithm to induce medical decision trees whose final decisions are both correct and the result of a sequence of observations with a medical sense. The algorithm has been tested with four medical decision problems, and the successful results discussed.
Pere Torres, David Riaño, Joan Albert López-Vallverdú
Critiquing Knowledge Representation in Medical Image Interpretation Using Structure Learning
Abstract
Medical image interpretation is a difficult problem for which human interpreters, radiologists in this case, are normally better equipped than computers. However, there are many clinical situations where radiologist’s performance is suboptimal, yielding a need for exploitation of computer-based interpretation for assistance. A typical example of such a problem is the interpretation of mammograms for breast-cancer detection. For this paper, we investigated the use of Bayesian networks as a knowledge-representation formalism, where the structure was drafted by hand and the probabilistic parameters learnt from image data. Although this method allowed for explicitly taking into account expert knowledge from radiologists, the performance was suboptimal. We subsequently carried out extensive experiments with Bayesian-network structure learning, for critiquing the Bayesian network. Through these experiments we have gained much insight into the problem of knowledge representation and concluded that structure learning results can be conceptually clear and of help in designing a Bayesian network for medical image interpretation.
Niels Radstake, Peter J. F. Lucas, Marina Velikova, Maurice Samulski
Linguistic and Temporal Processing for Discovering Hospital Acquired Infection from Patient Records
Abstract
This paper describes the first steps of development of a rule-based system that automatically processes medical records in order to discover possible cases of hospital acquired infections (HAI). The system takes as input a set of patient records in electronic format and gives as output, for each document, information regarding HAI. In order to achieve this goal, a temporal processing together with a deep syntactic and semantic analysis of the patient records is performed. Medical knowledge used by the rules is derived from a set of documents that have been annotated by medical doctors. After a brief description of the context of this work, we present the general architecture of our document processing chain and explain how we perform our temporal and linguistic analysis. Finally, we report our preliminary results and we lay out the next steps of the project.
Caroline Hagège, Pierre Marchal, Quentin Gicquel, Stefan Darmoni, Suzanne Pereira, Marie-Hélène Metzger
A Markov Analysis of Patients Developing Sepsis Using Clusters
Abstract
Sepsis is a significant cause of mortality and morbidity. There are now aggressive goal oriented treatments that can be used to help patients suffering from sepsis. By predicting which patients are more likely to develop sepsis, early treatment can potentially reduce their risks. However, diagnosing sepsis is difficult since there is no “standard” presentation, despite many published definitions of this condition.
In this work, data from a large observational cohort of patients – with variables collected at varying time periods – are observed in order to determine whether sepsis develops or not. A cluster analysis approach is used to form groups of correlated datapoints. This sequence of datapoints is then categorized on a per person basis and the frequency of transitioning from one grouping to another is computed. The result is posed as a Markov model which can accurately estimate the likelihood of a patient developing sepsis. A discussion of the implications and uses of this model is presented.
Femida Gwadry-Sridhar, Michael Bauer, Benoit Lewden, Ali Hamou
Towards the Interoperability of Computerised Guidelines and Electronic Health Records: An Experiment with openEHR Archetypes and a Chronic Heart Failure Guideline
Abstract
Clinical guidelines contain recommendations based on the best empirical evidence available at the moment. There is a wide consensus about the benefits of guidelines and about the fact that they should be deployed through clinical information systems, making them available during clinical consultations. However, one of the main obstacles to this integration is the interaction with the electronic health record system. With the aim of solving the interoperability problems of guideline systems, we have investigated the utilisation of the openEHR standardisation proposal in the context of one of the existing guideline representation languages. Concretely, we have designed a collection of archetypes to be used within a chronic heart failure guideline. The main contribution of our work is the utilisation of openEHR archetypes in the framework of guideline representation languages. Other contributions include both the concrete set of archetypes that we have selected and the methodological approach that we have followed to obtain it.
Mar Marcos, Begoña Martínez-Salvador

Clinical Practice Guidelines

Identifying Treatment Activities for Modelling Computer-Interpretable Clinical Practice Guidelines
Abstract
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.
Katharina Kaiser, Andreas Seyfang, Silvia Miksch
Updating a Protocol-Based Decision-Support System’s Knowledge Base: A Breast Cancer Case Study
Abstract
Modelling clinical guidelines or protocols in a computer-executable form is a prerequisite to support their execution by Decision Support Systems. Progress of medical knowledge requires frequent updates of the encoded knowledge model. Moreover, user perception of the decision process and user preferences regarding the presentation of choices require modifications of the model.
In this paper, we describe these two maintenance requirements using a protocol for the medical therapy of breast cancer and the lessons learnt in the process. The protocol was modeled in Asbru and is used at the S. Chiara Hospital in Trento.
Claudio Eccher, Andreas Seyfang, Antonella Ferro, Silvia Miksch
Toward Probabilistic Analysis of Guidelines
Abstract
In the formal analysis of health-care, there is little work that combines probabilistic and temporal reasoning. On the one hand, there are those that aim to support the clinical thinking process, which is characterised by trade-off decision making taking into account uncertainty and preferences, i.e., the process has a probabilistic and decision-theoretic flavour. On the other hand, the management of care, e.g., guidelines and planning of tasks, is typically modelled symbolically using temporal, non-probabilistic, methods. This paper proposes a new framework for combining temporal reasoning with probabilistic decision making. The framework is instantiated with a guideline modelling language combined with probabilistic pharmokinetics and applied to the treatment of diabetes mellitus type 2.
Arjen Hommersom
Backmatter
Metadata
Title
Knowledge Representation for Health-Care
Editors
David Riaño
Annette ten Teije
Silvia Miksch
Mor Peleg
Copyright Year
2011
Publisher
Springer Berlin Heidelberg
Electronic ISBN
978-3-642-18050-7
Print ISBN
978-3-642-18049-1
DOI
https://doi.org/10.1007/978-3-642-18050-7

Premium Partner

    Image Credits