An approach for sub-ontology evolution in a distributed health care enterprise
Highlights
► Sub-ontologies can be used as a way to manage the changes in health ontologies. ► We built some rules to transmit the changes of the base ontology to sub-ontologies. ► Our rules assure the validity of the produced sub-ontologies. ► Evaluation showed the consistency of sub-ontologies to the evolved base ontology. ► The approach outperforms direct re-extraction in terms of the number of operations.
Introduction
Health care providers range from hospital to specific care centers such as rehabilitation centers. Patients can choose or be recommended to different health care providers to achieve their health goals. The increased mobility of people often results in them receiving health treatment from caregivers who are geographically separated. In these conditions, interoperability among different health providers is vitally important so that the medical records for each patient can be preserved and exchanged between providers. Brailer in [1] believes that the consumer can suffer from a lack of interoperability and health information exchange because the health care enterprise hopes to gain a comparative advantage by imposing high costs when consumers change health care providers. According to Dogac et al. [2], full ‘share-ability’ of data and information requires two levels of interoperability: semantic interoperability and functional (syntactic) interoperability. Our focus is on semantic interoperability, which is defined as the ability for information shared by systems to be understood at the level of formally defined domain concepts so that the information is computer processable by the receiving system [3].
It is believed that ontologies are one way to overcome the semantic interoperability problem between different health care providers. Using ontology, the semantic meaning of each health term can be uniformly interpreted. An example of the use of ontologies for health care applications is the binding between the archetype terms and the ontology concepts. Several health ontologies such as SNOMED CT (Systematized Nomenclature of Medicine—Clinical Terms), LOINC (Logical Observation Identifiers Names and Codes) and UMLS (Unified Medical Language System) can be referred to by archetype terms. Archetype has been proposed by the openEHR1 Foundation as a model of specific domain knowledge. This model has also been adopted by the CEN TC/2512 in its Health informatics—Electronic Health Record Communication (EN 13606) European Standard. The binding between the archetype terms and the ontology concepts is aimed at achieving semantic interoperability between different health care institutions which may use different electronic health record standards. Health terminologies are also used in a similar way in the HL7 (Health Level Seven) standard. Externally defined terms and codes such as SNOMED CT can be utilized in HL7 as an Instance Identifier, which is used to give a unique identity to people, persons, organizations, things and information objects.2
Since ontologies aim for standardization, their size is usually very large. Many ontologies in the health domain such as SNOMED CT, LOINC and especially UMLS have hundreds of thousands and even millions of concepts. For a specific health application, the use of the whole ontology is not appropriate since actually, only a small part of the ontology is relevant to the application. For example, the health information system of a pharmacy, which uses SNOMED CT as the base ontology, requires only terms related to drugs, while there are many more terms in SNOMED CT, such as terms related to the examination of patients and medical procedures, which obviously are not relevant to the application. For an application with such a specific focus, sub-ontologies can be utilized instead of the whole ontology.
A sub-ontology is a subset of an ontology derived from that base ontology using a specific extraction process. A characteristic which differentiates a sub-ontology from a subset is that a sub-ontology should be a valid ontology in its own right [4]. A sub-ontology refers to a particular part of a base ontology which is appropriate to a specific context, user, specialty, etc. An example of the use of a specific context of knowledge in the clinical domain is the concept of archetype, which has been mentioned previously. An archetype describes a complete clinical knowledge concept such as ‘diagnosis’ or ‘test result’ [5]. An archetype may contain clinical terms which refer to the terms in health ontologies such as SNOMED CT and LOINC. Sari et al. propose the use of archetype sub-ontology to represent the semantic content of an archetype [6]. The archetype sub-ontology is extracted from the health ontology. Similarly, Yu et al. [7] has proposed a kind of sub-ontology referred to as the Terminological Shadow, which is derived from SNOMED CT, to represent the semantic content of an archetype. These studies show the applicability of the use of sub-ontologies in the health domain.
Ontology and sub-ontology should represent the current knowledge in the domain. When the knowledge changes, they should be adjusted. This process is known as ontology evolution and is one of the prominent issues in the use of ontologies in the health domain as it is one of the domains in which knowledge changes frequently. This is shown by the high frequency of health ontology changes. As an example, in each version of SNOMED CT, which is released twice a year, the average number of changes is more than 50,000 which consist of additions (45.45%), status changes (30.87%), and minor changes (23.68%) [8].
When an ontology evolves, the sub-ontologies derived from it should be adjusted as well so that they are consistent with the base ontology. Re-extraction based on the evolved ontology can be the simplest method to maintain the consistency of the sub-ontologies. However, this approach is not practical when the number of sub-ontologies is high and the changes in health ontologies occur frequently. In this case, it is more appropriate to change the sub-ontologies according to the changes which have taken place in the base ontology. In the notion of ontology evolution, this process is known as change propagation.
In addition to ensuring consistency to the base ontology, another important reason for the change propagation process to occur in sub-ontologies is that sub-ontologies should be kept valid. In other words, it should be assured that the evolved sub-ontologies produced from the change propagation process are the same as the ones extracted directly from the evolved base ontology. Rules are needed to determine which changes should be propagated to the sub-ontologies to avoid the sub-ontologies from becoming too big or too small but, at the same time, keep its semantic content. Most of the existing change propagation approaches [9], [10], [11] do not consider this requirement because they have not been applied to sub-ontologies. Moreover, the approaches are usually based on the assumption that the semantics of the changes are already known from the version log which contains the list of changes which occur from a previous version of the ontology to the next one. This is not appropriate for health ontologies as most of them do not provide version logs containing a list of changes which are semantically meaningful. For example, SNOMED CT provides a list of basic change operations such as additions and deletions of concepts and descriptions, while they actually permit more complex change operations such as the movement of concept. To capture such types of changes, an approach to identify the semantics of change is needed which should be based on the nature of health ontologies.
In this paper, we propose the use of sub-ontologies as a way to simplify ontology evolution management in a distributed health enterprise. The main issue addressed in this work is the change propagation mechanism from the base ontology to the sub-ontologies derived from it which can be distributed among different health institutions. The goal is to maintain the consistency of the sub-ontologies with the base ontology while keeping the validity of the sub-ontologies. To simplify the propagation process, we also develop a change identification process which is based on the available change logs of health ontologies. A case study is used to evaluate the approach by comparing a sub-ontology produced from the proposed change propagation approach with the one directly extracted from the evolved ontology in terms of the consistency of the content and the number of operations carried out to produce them. The efficiency of the sub-ontology evolution process is also enhanced by isolating changes only to the sub-ontologies affected by the changes which occurred in the base ontology.
The rest of the paper is organized as follows. Section 2 presents a motivational scenario which shows the benefit of our approach. Section 3 elaborates previous work related to ontology change propagation. The formalization of a health ontology is presented in Section 4, followed by a discussion on change operations in Section 5 and a description of the identification process of the semantic change operations in Section 6. Section 7 discusses the change propagation approach using some rules. The evaluation of the performance of the approach is presented in Section 8, and finally, Section 9 concludes our paper.
Section snippets
Motivational scenario
Consider a situation where a patient is suspected to suffer from diabetes. He goes to hospital to see the doctor. The doctor advises him to go to the laboratory to have his blood glucose checked. In the laboratory, a lab analyst analyzes his blood glucose according to the standard procedure for blood glucose measurement and then sends the results to the doctor. The doctor will make a diagnosis based on his knowledge of diabetes by examining the results of the blood glucose test and other
Related work
Several frameworks of ontology evolution such as CHAO [12], KAON [13] and Evolva [14] include change propagation as one of the phases in the ontology evolution process. In [15], change propagation is the fifth phase of a six-phase ontology evolution process: change capturing, change representation, semantics of change, change implementation, change propagation and change validation. According to [16], the aim of the change propagation phase is to bring all dependent artifacts in a consistent
Formalization of health ontology
Before formalizing the change operations of health ontologies, in this section we first formalize the definition of a health ontology. Our definition is mainly based on two prominent health ontologies, i.e. UMLS and SNOMED CT. Since UMLS contains several different health ontologies which can be represented uniformly, we consider that by basing our definition on UMLS, we can accommodate the representation of other health ontologies. Some examples will be provided and taken from SNOMED CT to
Change operations
Ontology change operations can be observed from two points of view: user requirements and implementation. Changes based on these two points of view are different to each other in their execution. For example, from the user requirement view, a deletion operation of a concept from the hierarchy should also involve, other than the deletion of the concept itself, the deletion operations to the relationships connecting the concepts to other concepts and the deletion operation to its description
Identifying the semantic change operations
As previously mentioned, most health ontologies present their change logs based on the basic change operations. From the change logs as well as the comparison of the versions of the ontologies, we can derive the following lists of basic changes: list of concept additions, list of concept deletions, list of relationship additions, list of relationship deletions, list of description additions, list of description deletions, list of description alterations, list of description mapping additions
Propagation of changes to sub-ontologies
The changes which occurred to the main ontology must be propagated to the sub-ontologies to ensure that the sub-ontologies are kept updated. Since the changes are reflected by the change operations, they are the ones which must be propagated to the sub-ontology. However, there are two issues which must be considered in the propagation process. Firstly, the change propagation must not cause the unnecessary growth of the number of concepts in the sub-ontology. Only changes which affect the
Evaluation and discussion
To evaluate our proposed method of change propagation, we conduct an experiment using a case study. We perform the evolution of some sub-ontologies using both our method and direct re-extraction. Then, we compare the sub-ontologies produced using both methods in terms of the content and the number of operations executed in both methods to determine the advantage of our approach. The following sub-section is the description of the experimental design. It is followed by the discussion on the
Conclusion
This paper proposed a new approach to manage the evolution of standardized ontologies used in health enterprises. The use of distributed sub-ontologies in such an environment substantially improves the efficiency of update propagation to interconnected ontologies within a distributed enterprise.
A formal definition of a health ontology, as well as a valid sub-ontology, is presented to give a theoretical foundation to the approach. Based on this, several rules have been built in propagating the
Acknowledgment
This work is partially supported by the Ministry of National Education of the Republic of Indonesia through the scholarship granted to the first author. We also would like to express our gratitude to any anonymous reviewers of our manuscript for their valuable suggestions to improve this paper.
References (29)
- et al.
Archetype sub-ontology: improving constraint-based clinical knowledge model in electronic health records
Knowledge Based Systems
(2012) - et al.
Representation of change in controlled medical terminologies
Artificial Intelligence in Medicine
(1999) - et al.
Understanding ontology evolution: a change detection approach
Web Semantics: Science, Services and Agents on the World Wide Web
(2007) - et al.
A comparison of two methods for retrieving icd-9-cm data: the effect of using an ontology-based method for handling terminology changes
Journal of Biomedical Informatics
(2011) - et al.
Evolution of the sequence ontology terms and relationship
Journal of Biomedical Informatics
(2011) - et al.
Ontology driven semantic profiling and retrieval in medical information systems
Web Semantics: Science, Services and Agents on the World Wide Web
(2009) Interoperability: the key to the future health care system
Health Affairs Web Exclusive
(2005)- et al.
Exploiting ebXML registry semantic constructs for handling archetype metadata in healthcare informatics
International Journal of Metadata
(2005) - V. Bicer, G.B. Laleci, A. Dogac, Y. Kabak, Providing semantic interoperability in the healthcare domain through...
- C. Wouters, A Formalization and Application of Ontology Extraction, Ph.D. Thesis, La Trobe University,...
Archetypes 101
Cited by (10)
Ontology Maintenance System for Rheumatoid Disease
2018, Procedia Computer ScienceModeling ontology evolution via Pi-Calculus
2016, Information SciencesCitation Excerpt :Ochs et al. proposed to use compact networks to summarize and visualize the global structural changes in OE and defined the Diff Abstraction Network [26]. Sari et al. provided a way to treat OE in sub-ontologies by isolating the required change propagation and optimizing the propagation mechanism [37]. However, such attempts regarded the changes as passive results from a subjective administration point of view, that neglected the initiative aspects in the operational semantics.
Enhancing Ontology Integration in Medical Texts through Advanced Mechanisms
2023, Research SquareThe ontology-based modeling and evolution of digital twin for assembly workshop
2021, International Journal of Advanced Manufacturing TechnologyIdentification and formal representation of change operations in LOINC evolution
2019, International Journal of Advanced Computer Science and ApplicationsRole-Preserving Redaction of Medical Records to Enable Ontology-Driven Processing
2017, BioNLP 2017 - SIGBioMed Workshop on Biomedical Natural Language Processing, Proceedings of the 16th BioNLP Workshop