A reference ontology for biomedical informatics: the Foundational Model of Anatomy

The Foundational Model of Anatomy (FMA), initially developed as an enhancement of the anatomical content of UMLS, is a domain ontology of the concepts and relationships that pertain to the structural organization of the human body. It encompasses the material objects from the molecular to the macroscopic levels that constitute the body and associates with them non-material entities (spaces, surfaces, lines, and points) required for describing structural relationships. The disciplined modeling approach employed for the development of the FMA relies on a set of declared principles, high level schemes, Aristotelian definitions and a frame-based authoring environment. We propose the FMA as a reference ontology in biomedical informatics for correlating different views of anatomy, aligning existing and emerging ontologies in bioinformatics ontologies and providing a structure-based template for representing biological functions.     

The Foundational Model of Anatomy in OWL 2 and its use

ObjectiveThe objective is to represent the Foundational Model of Anatomy (FMA) in the OWL 2 Web Ontology Language (informally OWL 2), and to use it in a European cross-lingual portal of health terminologies for indexing and searching Web resources. Formalizing the FMA in OWL 2 is essential for semantic interoperability, to improve its design, and to ensure its reliability and correctness, which is particularly important for medical applications.Method and materialThe native FMA was implemented in frames and stored in a MySQL database backend. The main strength of the method is to leverage OWL 2 expressiveness and to rely on the naming conventions of the FMA, to make explicit some implicit semantics, while improving its ontological model and fixing some errors. Doing so, the semantics (meaning) of the formal definitions and axioms are anatomically correct. A flexible tool enables the generation of a new version in OWL 2 at each Protégé FMA update. While it creates by default a ‘standard’ version of the FMA in OWL 2 (FMA-OWL), many options allow for producing other variants customized to users’ applications. Once formalized in OWL 2, it was possible to use an inference engine to check the ontology and detect inconsistencies. Next, the FMA-OWL was used to derive a lightweight FMA terminology for a European cross-lingual portal of terminologies/ontologies for indexing and searching resources. The transformation is mainly based on a reification process.ResultComplete representations of the entire FMA in OWL 1 or OWL 2 are now available. The formalization tool is flexible and easy to use, making it possible to obtain an OWL 2 version for all existing public FMA. A number of errors were detected in the native FMA and several patterns of recurrent errors were identified in the original FMA. This shows how the underlying OWL 2 ontology is essential to ensure that the lightweight derived terminology is reliable.The FMA OWL 2 ontology has been applied to derive an anatomy terminology that is used in a European cross-lingual portal of health terminologies. This portal is daily used by librarians to index Web health resources. In August 2011, 6481 out of 81,450 health resources of CISMeF catalog (http://www.chu-rouen.fr/cismef/ – accessed 29.08.12) (7.96%) were indexed with at least one FMA entity.ConclusionThe FMA is a central terminology used to index and search Web resources. To the best of our knowledge, neither a complete representation of the entire FMA in OWL 2, nor an anatomy terminology available in a cross-lingual portal, has been developed to date. The method designed to represent the FMA ontology in OWL 2 presented in this article is general and may be extended to other ontologies. Using a formal ontology for quality assurance and deriving a lightweight terminology for biomedical applications is a general and promising strategy.     

A formal theory for spatial representation and reasoning in biomedical ontologies

OBJECTIVE: The objective of this paper is to demonstrate how a formal spatial theory can be used as an important tool for disambiguating the spatial information embodied in biomedical ontologies and for enhancing their automatic reasoning capabilities. METHOD AND MATERIALS: This paper presents a formal theory of parthood and location relations among individuals, called Basic Inclusion Theory (BIT). Since biomedical ontologies are comprised of assertions about classes of individuals (rather than assertions about individuals), we define parthood and location relations among classes in the extended theory Basic Inclusion Theory for Classes (BIT+Cl). We then demonstrate the usefulness of this formal theory for making the logical structure of spatial information more precise in two ontologies concerned with human anatomy: the Foundational Model of Anatomy (FMA) and GALEN. RESULTS: We find that in both the FMA and GALEN, class-level spatial relations with different logical properties are not always explicitly distinguished. As a result, the spatial information included in these biomedical ontologies is often ambiguous and the possibilities for implementing consistent automatic reasoning within or across ontologies are limited. CONCLUSION: Precise formal characterizations of all spatial relations assumed by a biomedical ontology are necessary to ensure that the information embodied in the ontology can be fully and coherently utilized in a computational environment. This paper can be seen as an important beginning step toward achieving this goal, but much more work along these lines is required.     

Content-specific auditing of a large scale anatomy ontology

Biomedical ontologies are envisioned to be useable in a range of research and clinical applications. The requirements for such uses include formal consistency, adequacy of coverage, and possibly other domain specific constraints. In this report we describe a case study that illustrates how application specific requirements may be used to identify modeling problems as well as data entry errors in ontology building and evolution. We have begun a project to use the UW Foundational Model of Anatomy (FMA) in a clinical application in radiation therapy planning. This application focuses mainly (but not exclusively) on the representation of the lymphatic system in the FMA, in order to predict the spread of tumor cells to regional metastatic sites. This application requires that the downstream relations associated with lymphatic system components must only be to other lymphatic chains or vessels, must be at the appropriate level of granularity, and that every path through the lymphatic system must terminate at one of the two well known trunks of the lymphatic system. It is possible through a programmable query interface to the FMA to write small programs that systematically audit the FMA for compliance with these constraints. We report on the design of some of these programs, and the results we obtained by applying them to the lymphatic system. The algorithms and approach are generalizable to other network organ systems in the FMA such as arteries and veins. In addition to illustrating exact constraint checking methods, this work illustrates how the details of an application may reflect back a requirement to revise the design of the ontology itself.     

Dynamic sub-ontology evolution for traditional Chinese medicine web ontology

As a form of important domain knowledge, large-scale ontologies play a critical role in building a large variety of knowledge-based systems. To overcome the problem of semantic heterogeneity and encode domain knowledge in reusable format, a large-scale and well-defined ontology is also required in the traditional Chinese medicine discipline. We argue that to meet the on-demand and scalability requirement ontology-based systems should go beyond the use of static ontology and be able to self-evolve and specialize for the domain knowledge they possess. In particular, we refer to the context-specific portions from large-scale ontologies like the traditional Chinese medicine ontology as sub-ontologies. Ontology-based systems are able to reuse sub-ontologies in local repository called ontology cache. In order to improve the overall performance of ontology cache, we propose to evolve sub-ontologies in ontology cache to optimize the knowledge structure of sub-ontologies. Moreover, we present the sub-ontology evolution approach based on a genetic algorithm for reusing large-scale ontologies. We evaluate the proposed evolution approach with the traditional Chinese medicine ontology and obtain promising results.     

Combining expert knowledge and knowledge automatically acquired from electronic data sources for continued ontology evaluation and improvement

IntroductionA common bottleneck during ontology evaluation is knowledge acquisition from domain experts for gold standard creation. This paper contributes a novel semi-automated method for evaluating the concept coverage and accuracy of biomedical ontologies by complementing expert knowledge with knowledge automatically extracted from clinical practice guidelines and electronic health records, which minimizes reliance on expensive domain expertise for gold standards generation.MethodsWe developed a bacterial clinical infectious diseases ontology (BCIDO) to assist clinical infectious disease treatment decision support. Using a semi-automated method we integrated diverse knowledge sources, including publically available infectious disease guidelines from international repositories, electronic health records, and expert-generated infectious disease case scenarios, to generate a compendium of infectious disease knowledge and use it to evaluate the accuracy and coverage of BCIDO.ResultsBCIDO has three classes (i.e., infectious disease, antibiotic, bacteria) containing 593 distinct concepts and 2345 distinct concept relationships. Our semi-automated method generated an ID knowledge compendium consisting of 637 concepts and 1554 concept relationships. Overall, BCIDO covered 79% (504/637) of the concepts and 89% (1378/1554) of the concept relationships in the ID compendium. BCIDO coverage of ID compendium concepts was 92% (121/131) for antibiotic, 80% (205/257) for infectious disease, and 72% (178/249) for bacteria. The low coverage of bacterial concepts in BCIDO was due to a difference in concept granularity between BCIDO and infectious disease guidelines. Guidelines and expert generated scenarios were the richest source of ID concepts and relationships while patient records provided relatively fewer concepts and relationships.ConclusionsOur semi-automated method was cost-effective for generating a useful knowledge compendium with minimal reliance on domain experts. This method can be useful for continued development and evaluation of biomedical ontologies for better accuracy and coverage.     

An ontology of randomized controlled trials for evidence-based practice: content specification and evaluation using the competency decomposition method

Randomized controlled trials (RCTs) are one of the least biased sources of clinical research evidence, and are therefore a critical resource for the practice of evidence-based medicine. With over 10,000 new RCTs indexed in Medline each year, knowledge systems are needed to help clinicians translate evidence into practice. Common ontologies for RCTs and other domains would facilitate the development of these knowledge systems. However, no standard method exists for developing domain ontologies. In this paper, we describe a new systematic approach to specifying and evaluating the conceptual content of ontologies. In this method, called competency decomposition, the target task for an ontology is hierarchically decomposed into subtasks and methods, and the ontology content is specified by identifying the domain information required to complete each of the subtasks. We illustrate the use of this competency decomposition approach for the content specification and evaluation of an RCT ontology for evidence-based practice.     

Relationship auditing of the FMA ontology

The Foundational Model of Anatomy (FMA) ontology is a domain reference ontology based on a disciplined modeling approach. Due to its large size, semantic complexity and manual data entry process, errors and inconsistencies are unavoidable and might remain within the FMA structure without detection. In this paper, we present computable methods to highlight candidate concepts for various relationship assignment errors. The process starts with locating structures formed by transitive structural relationships (part_of, tributary_of, branch_of) and examine their assignments in the context of the IS-A hierarchy. The algorithms were designed to detect five major categories of possible incorrect relationship assignments: circular, mutually exclusive, redundant, inconsistent, and missed entries. A domain expert reviewed samples of these presumptive errors to confirm the findings. Seven thousand and fifty-two presumptive errors were detected, the largest proportion related to part_of relationship assignments. The results highlight the fact that errors are unavoidable in complex ontologies and that well designed algorithms can help domain experts to focus on concepts with high likelihood of errors and maximize their effort to ensure consistency and reliability. In the future similar methods might be integrated with data entry processes to offer real-time error detection.     

Anatomical reasoning in the informatics age: Principles, ontologies, and agendas

Reasoning about anatomy shares historical scientific roots with formal logic and artificial intelligence. With advances in computer-based intelligent programming, high-level biological structural knowledge may be exploited directly for biomedical research, clinical tasks, and educational applications. We consider the special nature of anatomical domain knowledge, emphasizing the complex concepts and semantics that must be represented in the development of ontologies, formally structured databases of biological information. We review the evolution of the fundamental scientific principles of logic and artificial intelligence needed for building machines that can make use of anatomical knowledge. We look at methods for compiling ontologies and compare the structural designs of the Foundational Model of Anatomy and Open GALEN ontologies. We further consider issues related to mapping developing anatomy resources with other biological ontologies in genomics, proteomics, and physiology. Although early results are promising, considerable resources and continuing effort must be committed to completing and extending anatomical ontologies for the ultimate success of computer-based anatomical reasoning. Anat Rec (Part B: New Anat) 289B:72-84, 2006. (c) 2006 Wiley-Liss, Inc.     

Using ontologies linked with geometric models to reason about penetrating injuries

OBJECTIVE: Medical assessment of penetrating injuries is a difficult and knowledge-intensive task, and rapid determination of the extent of internal injuries is vital for triage and for determining the appropriate treatment. Physical examination and computed tomographic (CT) imaging data must be combined with detailed anatomic, physiologic, and biomechanical knowledge to assess the injured subject. We are developing a methodology to automate reasoning about penetrating injuries using canonical knowledge combined with specific subject image data. METHODS AND MATERIAL: In our approach, we build a three-dimensional geometric model of a subject from segmented images. We link regions in this model to entities in two knowledge sources: (1) a comprehensive ontology of anatomy containing organ identities, adjacencies, and other information useful for anatomic reasoning and (2) an ontology of regional perfusion containing formal definitions of arterial anatomy and corresponding regions of perfusion. We created computer reasoning services ("problem solvers") that use the ontologies to evaluate the geometric model of the subject and deduce the consequences of penetrating injuries. RESULTS: We developed and tested our methods using data from the Visible Human. Our problem solvers can determine the organs that are injured given particular trajectories of projectiles, whether vital structures--such as a coronary artery--are injured, and they can predict the propagation of injury ensuing after vital structures are injured. CONCLUSION: We have demonstrated the capability of using ontologies with medical images to support computer reasoning about injury based on those images. Our methodology demonstrates an approach to creating intelligent computer applications that reason with image data, and it may have value in helping practitioners in the assessment of penetrating injury.     

From frames to OWL2: Converting the Foundational Model of Anatomy

ObjectiveThe Foundational Model of Anatomy (FMA) [Rosse C, Mejino JLV. A reference ontology for bioinformatics: the Foundational Model of Anatomy. J. Biomed. Inform. 2003;36:478–500] is an ontology that represents canonical anatomy at levels ranging from the entire body to biological macromolecules, and has rapidly become the primary reference ontology for human anatomy, and a template for model organisms. Prior to this work, the FMA was developed in a knowledge modeling language known as Protégé Frames. Frames is an intuitive representational language, but is no longer the industry standard. Recognizing the need for an official version of the FMA in the more modern semantic web language OWL2 (hereafter referred to as OWL), the objective of this work was to create a generalizable Frames-to-OWL conversion tool, to use the tool to convert the FMA to OWL, to “clean up” the converted FMA so that it classifies under an EL reasoner, and then to do all further development in OWL.MethodsThe conversion tool is a Java application that uses the Protégé knowledge representation API for interacting with the initial Frames ontology, and uses the OWL-API for producing new statements (axioms, etc.) in OWL. The converter is relation centric. The conversion is configurable, on a property-by-property basis, via user-specifiable XML configuration files. The best conversion, for each property, was determined in conjunction with the FMA knowledge author. The convertor is potentially generalizable, which we partially demonstrate by using it to convert our Ontology of Craniofacial Development and Malformation as well as the FMA. Post-conversion cleanup involved using the Explain feature of Protégé to trace classification errors under the ELK reasoner in Protégé, fixing the errors, then re-running the reasoner.ResultsWe are currently doing all our development in the converted and cleaned-up version of the FMA. The FMA (updated every 3 months) is available via our FMA web page http://si.washington.edu/projects/fma, which also provides access to mailing lists, an issue tracker, a SPARQL endpoint (updated every week), and an online browser. The converted OCDM is available at http://www.si.washington.edu/projects/ocdm. The conversion code is open source, and available at http://purl.org/sig/software/frames2owl. Prior to the post-conversion cleanup 73% of the more than 100,000 classes were unsatisfiable. After correction of six types of errors no classes remained unsatisfiable.ConclusionBecause our FMA conversion captures all or most of the information in the Frames version, is the only complete OWL version that classifies under an EL reasoner, and is maintained by the FMA authors themselves, we propose that this version should be the only official release version of the FMA in OWL, supplanting all other versions. Although several issues remain to be resolved post-conversion, release of a single, standardized version of the FMA in OWL will greatly facilitate its use in informatics research and in the development of a global knowledge base within the semantic web. Because of the fundamental nature of anatomy in both understanding and organizing biomedical information, and because of the importance of the FMA in particular in representing human anatomy, the FMA in OWL should greatly accelerate the development of an anatomically based structural information framework for organizing and linking a large amount of biomedical information.     

Biomedical Ontologies to Guide AI Development in Radiology

The advent of deep learning has engendered renewed and rapidly growing interest in artificial intelligence (AI) in radiology to analyze images, manipulate textual reports, and plan interventions. Applications of deep learning and other AI approaches must be guided by sound medical knowledge to assure that they are developed successfully and that they address important problems in biomedical research or patient care. To date, AI has been applied to a limited number of real-world radiology applications. As AI systems become more pervasive and are applied more broadly, they will benefit from medical knowledge on a larger scale, such as that available through computer-based approaches. A key approach to represent computer-based knowledge in a particular domain is an ontology. As defined in informatics, an ontology defines a domain’s terms through their relationships with other terms in the ontology. Those relationships, then, define the terms’ semantics, or “meaning.” Biomedical ontologies commonly define the relationships between terms and more general terms, and can express causal, part-whole, and anatomic relationships. Ontologies express knowledge in a form that is both human-readable and machine-computable. Some ontologies, such as RSNA’s RadLex radiology lexicon, have been applied to applications in clinical practice and research, and may be familiar to many radiologists. This article describes how ontologies can support research and guide emerging applications of AI in radiology, including natural language processing, image–based machine learning, radiomics, and planning.

     

COnto–Diff: generation of complex evolution mappings for life science ontologies

Life science ontologies evolve frequently to meet new requirements or to better reflect the current domain knowledge. The development and adaptation of large and complex ontologies is typically performed collaboratively by several curators. To effectively manage the evolution of ontologies it is essential to identify the difference (Diff) between ontology versions. Such a Diff supports the synchronization of changes in collaborative curation, the adaptation of dependent data such as annotations, and ontology version management. We propose a novel approach COnto–Diff to determine an expressive and invertible diff evolution mapping between given versions of an ontology. Our approach first matches the ontology versions and determines an initial evolution mapping consisting of basic change operations (insert/update/delete). To semantically enrich the evolution mapping we adopt a rule-based approach to transform the basic change operations into a smaller set of more complex change operations, such as merge, split, or changes of entire subgraphs. The proposed algorithm is customizable in different ways to meet the requirements of diverse ontologies and application scenarios. We evaluate the proposed approach for large life science ontologies including the Gene Ontology and the NCI Thesaurus and compare it with PromptDiff. We further show how the Diff results can be used for version management and annotation migration in collaborative curation.     

An Ontology for PACS Integration

An ontology describes a set of classes and the relationships among them. We explored the use of an ontology to integrate picture archiving and communication systems (PACS) with other information systems in the clinical enterprise. We created an ontological model of thoracic radiology that contained knowledge of anatomy, imaging procedures, and performed procedure steps. We explored the use of the model in two use cases: (1) to determine examination completeness and (2) to identify reference (comparison) images obtained in the same imaging projection. The model incorporated a total of 138 classes, including radiology orderables, procedures, procedure steps, imaging modalities, patient positions, and imaging planes. Radiological knowledge was encoded as relationships among these classes. The ontology successfully met the information requirements of the two use-case scenarios. Ontologies can represent radiological and clinical knowledge to integrate PACS with the clinical enterprise and to support the radiology interpretation process.     

Law and order: assessing and enforcing compliance with ontological modeling principles in the Foundational Model of Anatomy

The objective of this study is to provide an operational definition of principles with which well-formed ontologies should comply. We define 15 such principles, related to classification (e.g., no hierarchical cycles are allowed; concepts have a reasonable number of children), incompatible relationships (e.g., two concepts cannot stand both in a taxonomic and partitive relation), dependence among concepts, and the co-dependence of equivalent sets of relations. Implicit relations--embedded in concept names or inferred from a combination of explicit relations--are used in this process in addition to the relations explicitly represented. As a case study, we investigate the degree to which the Foundational Model of Anatomy (FMA)--a large ontology of anatomy--complies with these 15 principles. The FMA succeeds in complying with all the principles: totally with one and mostly with the others. Reasons for non-compliance are analyzed and suggestions are made for implementing effective enforcement mechanisms in ontology development environments. The limitations of this study are also discussed.     

An analysis of FMA using structural self-bisimilarity

As ontologies are mostly manually created, they tend to contain errors and inconsistencies. In this paper, we present an automated computational method to audit symmetric concepts in ontologies by leveraging self-bisimilarity and linguistic structure in the concept names. Two concepts A and B are symmetric if concept B can be obtained from concept A by replacing a single modifier such as “left” with its symmetric modifier such as “right.” All possible local structural types for symmetric concept pairs are enumerated according to their local subsumption hierarchy, and the pairs are further classified into Non-Matches and Matches. To test the feasibility and validate the benefits of this method, we computed all the symmetric modifier pairs in the Foundational Model of Anatomy (FMA) and selected six of them for experimentation. 9893 Non-Matches and 221 abnormal Matches with potential errors were discovered by our algorithm. Manual evaluation by FMA domain experts on 176 selected Non-Matches and all the 221 abnormal Matches found 102 missing concepts and 40 misaligned concepts. Corrections for them have currently been implemented in the latest version of FMA. Our result demonstrates that self-bisimilarity can be a valuable method for ontology quality assurance, particularly in uncovering missing concepts and misaligned concepts. Our approach is computationally scalable and can be applied to other ontologies that are rich in symmetric concepts.     

An empirical analysis of ontology reuse in BioPortal

Biomedical ontologies often reuse content (i.e., classes and properties) from other ontologies. Content reuse enables a consistent representation of a domain and reusing content can save an ontology author significant time and effort. Prior studies have investigated the existence of reused terms among the ontologies in the NCBO BioPortal, but as of yet there has not been a study investigating how the ontologies in BioPortal utilize reused content in the modeling of their own content. In this study we investigate how 355 ontologies hosted in the NCBO BioPortal reuse content from other ontologies for the purposes of creating new ontology content. We identified 197 ontologies that reuse content. Among these ontologies, 108 utilize reused classes in the modeling of their own classes and 116 utilize reused properties in class restrictions. Current utilization of reuse and quality issues related to reuse are discussed.     

Semantic enrichment for medical ontologies

The Unified Medical Language System (UMLS) contains two separate but interconnected knowledge structures, the Semantic Network (upper level) and the Metathesaurus (lower level). In this paper, we have attempted to work out better how the use of such a two-level structure in the medical field has led to notable advances in terminologies and ontologies. However, most ontologies and terminologies do not have such a two-level structure. Therefore, we present a method, called semantic enrichment, which generates a two-level ontology from a given one-level terminology and an auxiliary two-level ontology. During semantic enrichment, concepts of the one-level terminology are assigned to semantic types, which are the building blocks of the upper level of the auxiliary two-level ontology. The result of this process is the desired new two-level ontology. We discuss semantic enrichment of two example terminologies and how we approach the implementation of semantic enrichment in the medical domain. This implementation performs a major part of the semantic enrichment process with the medical terminologies, with difficult cases left to a human expert.     

ONTOFUSION: ontology-based integration of genomic and clinical databases

ONTOFUSION is an ontology-based system designed for biomedical database integration. It is based on two processes: mapping and unification. Mapping is a semi-automated process that uses ontologies to link a database schema with a conceptual framework-named virtual schema. There are three methodologies for creating virtual schemas, according to the origin of the domain ontology used: (1) top-down--e.g. using an existing ontology, such as the UMLS or Gene Ontology--, (2) bottom-up--building a new domain ontology-- and (3) a hybrid combination. Unification is an automated process for integrating ontologies and hence the database to which they are linked. Using these methods, we employed ONTOFUSION to integrate a large number of public genomic and clinical databases, as well as biomedical ontologies.