=Paper=
{{Paper
|id=Vol-3939/short3
|storemode=property
|title=Expanding the Ontology of Organizational Structures of Trauma Centers and Trauma
Systems
|pdfUrl=https://ceur-ws.org/Vol-3939/short3.pdf
|volume=Vol-3939
|authors=Diya Mehta,Justin M. Whorton,Reza Shahriari,Eric D Ragan,Jonathan P. Bona,William R. Hogan,Kevin W. Sexton,Mathias Brochhausen
|dblpUrl=https://dblp.org/rec/conf/icbo/MehtaWSRBHSB24
}}
==Expanding the Ontology of Organizational Structures of Trauma Centers and Trauma
Systems==
https://ceur-ws.org/Vol-3939/short3.pdf
Expanding the Ontology of Organizational Structures
of Trauma Centers and Trauma Systems
Diya Mehta1 , Justin M. Whorton2 , Reza Shahriari3 , Eric D Ragan3 , Jonathan P. Bona2 ,
William R. Hogan4 , Kevin W. Sexton2 and Mathias Brochhausen2,∗
1
Harvey Mudd College, 301 Platt Blvd, Claremont, CA 91711, USA
2
University of Arkansas for Medical Sciences, 4301 W Markham St, Little Rock, AR 72205, USA
3
University of Florida, 201 Criser Hall PO Box 114000 Gainesville, FL 32611, USA
4
Medical College of Wisconsin, 8701 W Watertown Plank Rd, Milwaukee, WI 53226, USA
Abstract
A knowledge gap exists regarding the impact of organizational parameters of trauma centers and patient
outcomes. This is partially due to such organizational parameters being understudied. The Ontology
of Organizational Structures of Trauma Centers and Trauma Systems (OOSTT) provides a controlled
vocabulary to study that specific area. It is used in tools created by the TIPTOE project to provide trauma
stakeholders with novel insights on role of organizational parameters and patient outcomes. This paper
reports the extension of OOSTT to cover relevant patient outcome measures.
Keywords
medical ontologies, trauma centers, organizational structures, patient outcomes
1. Introduction
In the United States in 2020, trauma is the leading cause of death for individuals under the
age of 45 [1]. Despite growing standardization of clinical trauma care, at Level 1 (L1) and
Level 2 (L2) trauma centers, there remains significant variability in patient outcomes across
trauma centers on both levels [2, 3]. We hypothesize that this variability in patient outcomes
is partially created by variability in organizational parameters of the trauma centers, which
is an understudied subject. By organizational parameters we mean parameters of a trauma
care environment, e.g., a trauma center, that describe how the care, documentation of care, and
quality improvement measures are organize. The organizational parameters include, but are
not restricted to key roles, e.g. trauma medical director, trauma program manager and trauma
registrar, the obligations and privileges associated with those roles, the staffing of the trauma
team, including credentials of participating providers, the availability of medical specialities and
sub-specialties in or to the trauma team. The Ontology of Organizational Structures of Trauma
Centers and Trauma Systems (OOSTT) is aimed to help address the knowledge gap regarding
15th International Conference on Biological and Biomedical Ontology, July 17-19 2024, Enschede, The Netherlands
∗
Corresponding author.
Envelope-Open dimehta@g.hmc.edu (D. Mehta); jmwhorton@uams.edu (J. M. Whorton); rshahriari@ufl.edu (R. Shahriari);
eragan@ufl.edu (E. D. Ragan); jpbona@uams.edu (J. P. Bona); hoganwr@mcw.edu (W. R. Hogan);
kevin.sexton@uams.edu (K. W. Sexton); mbrochhausen@uams.edu (M. Brochhausen)
Orcid 0000-0002-9881-1017 (W. R. Hogan); 0000-0002-1460-9867 (K. W. Sexton); 0000-0003-1834-3856 (M. Brochhausen)
© 2022 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR
ceur-ws.org
Workshop ISSN 1613-0073
Proceedings
Figure 1: Screenshot of the TIPTOE Knowledge Path Explorer. Showing an example how a user can
explore the knowledge graph about their trauma program.
organizational structures. Its initial releases cover representation of trauma centers and trauma
systems, their components, and the roles of professional and deontic roles that are part of these
organizations [4]. OOSTT has been tested and validated to provide a controlled vocabulary for
trauma centers and trauma systems organization [5]. It has been used to collect organizational
data of trauma centers and trauma systems for the Comparative Assessment Framework of
Environments of Trauma Care (CAFE) web service [6]. OOSTT is an OBO Foundry ontology,
that is open access and can be used by developers and other ontologies to represent medical
roles (e.g., trauma medical director), and organizational units (e.g., trauma team), and core
components of trauma care (e.g., trauma centers).
In 2022, the second phase of the CAFE project started, and was renamed Trauma Institutional
Priorities and Teams for Outcomes Efficacy (TIPTOE). The purpose of this phase is the evolution
of trauma center quality improvement fostering adding scientific evidence regarding impact
of organizational parameters on patient outcomes in L1 and L2 trauma centers. TIPTOE is
recruiting 230 L1 and L2 trauma centers to fill in the survey about organizational parameters,
similar to the CAFE web service [8], and provide their Trauma Quality Improvement Program
(TQIP) data. TQIP is an initiative by the American College of Surgeons, Committee on Trauma
aimed to improve the quality of care for trauma patients [7]. It collects data from trauma centers
and provides feedback about performance and identifies improvements to be implemented by
trauma center staff to improve outcomes [7].
One tool TIPTOE has developed is the Knowledge Path Explorer (KPE), that allows trauma
center stakeholders to explore a knowledge graph that links organizational parameters of
their institution to patient outcomes. The KPE graph is organized using OOSTT. Figure 1
shows the design of the KPE pilot that we are currently reviewing with medical staff for
enhancements to design and functionality. The current visual graph interface allows inspection
of specific parameters of interest while also providing the added benefit of showing context of
the ontological information and relationships to other related parameters. Through participatory
design with medical stakeholders and center leadership, the system will evolve to accommodate
a broad range of data exploration goals. In this paper, we report the extensions of OOSTT,
which are necessary to cover patient outcome data; something that was not necessary to the
first phase of the project. We also present early results on how the OOSTT extension allows
exploring TIPTOE data regarding two core competency questions:
1. How does the number of general surgeons with Advanced Trauma Life Support (ATLS)
certification at your trauma center affect the number of major complications including death?
2. How does the inter-correlation between neurosurgeons taking call exclusively and the num-
ber of neurosurgeons with certified 18 hours of trauma-specific Continued Medical Education
(CME) affect length of stay?
2. Methods
2.1. OBO Foundry and OBO Ontologies
The Open Biological and Biomedical Ontology Foundry (OBO Foundry) (http://obofoundry.org/)
is a library of open source, community developed biological and biomedical ontologies agreeing
to a set of overarching principles [8, 9]. The OBO Foundry aims at “facilitating the development,
harmonization, application and sharing of ontologies (...)”[9]. The following OBO Foundry
ontologies were used to expand OOSTT:
The Ontology for Modeling and Representation of Social Entities (OMRSE), initially names
Ontology for Medically Related Social Entities, is ”a realist representation of medically related
social entities”[10]. The scope has been expanded to cover ”that various entities that arise from
human social interactions, such as social acts, social roles, social groups, and organizations”[11].
The MONDO Disease Ontology ”provides a sustainable and fully-provenanced approach to
integrating disease concepts from numerous sources across disease categories”[12]. It currently
represents over 22.000 disease concepts that represent 90.000 source concept from 17 disease
resources[12].
The Infectious Disease Ontology (IDO) represents ”entities generally relevant to both the
biomedical and clinical aspects of infectious diseases, including terms such as pathogen, host,
vector, and vaccine”[13].
The Medical Action Ontology (MAxO) is ”a comprehensive open source computational
representation of medical diagnostics, preventions, procedures, interventions, and therapies”[14].
MAxO currently contains more than 1.700 terms representing medical actions, such as medical
procedures, interventions, therapies, and measurements[14].
2.2. OOSTT
OOSTT is a publicly available ontology that is part of the OBO Foundry and follows OBO
Foundry principles. OOSTT can be accessed via http://purl.obolibrary.org/obo/oostt.owl. Ad-
ditional information and tools, e.g., an issue tracker, can be found at OOSTT’s git repository:
http://github.com/OOSTT/OOSTT. OOSTT uses Basic Formal Ontology [15]as its top level
ontology and covers the domain of trauma center and trauma system organizational parameters.
In 2022, the design principles and coverage of OOSTT have been reviewed by WRH, who was
not involved in the initial OOSTT development. The adjustment and changes suggested by that
review have been implemented during 2023.
2.3. OOSTT Extensions
This current ontology development step aims to provide ontological representation for TQIP
data elements to enable the integration of TQIP data with data on organizational structures in
the TIPTOE project and, specifically in the KPE . This extension was done using two different
approaches: a) terminology-driven to broaden OOSTT coverage, b) data-driven providing
representation for the 3 patient outcomes TIPTOE focuses on.
2.3.1. Terminology-driven Extension
To foster integration with trauma outcome data nationwide, the study started with definitions
and labels from TQIP’s data dictionary, the National Trauma Data Standard (NTDS)[16]. ”The
NTDS Dictionary is designed to establish a national standard for the exchange of trauma registry
data, and to serve as the operational definitions for the National Trauma Data Bank (NTDB)”[16].
It is a crucial component of TQIP, since the standardization provided by the NTDS allowed the
addition of assessment of patient outcomes to the trauma center verification process[17]. Our
project was done as a Summer Research Internship by DM. First, 20 terms from the NTDS were
identified for implementation in the Web Ontology Language (OWL) and inclusion in OOSTT.
All 20 terms are listed in the first column of Table 1. These terms were picked based on priority
regarding project needs. A spreadsheet was created to account for changes made to each term
and its curation status.
First, the label of each term from NTSD was reviewed for consistency with the label format
suggested by [18]. Each label that did not follow the format was edited to their singular form, the
expanded version of abbreviations/symbols, and in lowercase lettering. For instance, “ICD-10
INJURY DIAGNOSIS” became “international classification of diseases tenth revision injury
diagnosis”. Additionally, all acronyms were expanded, to prevent misunderstandings, following
the OBO Foundry principle on naming[19].
There was one instance where one NTDS term, required three proposed OOSTT terms, to
specifically represent the NTDS term’s specified values: According to the NTDS database, the
term “alternative home residence” represents individuals that are either homeless, living at a
temporary residence, or are undocumented. Since those three values represent situations that
are not easily represented by one superclass, we decided to discard “alternative home residence”.
It was replaced by three terms capturing its respective values: “homeless”, “temporary address”,
and “undocumented immigrant”. By following these guidelines, we are preventing incorrect
hierarchical structures, such as claiming that an instance of an undocumented immigrant is
also a member of the class ‘alternative home residence’, once we build the OWL hierarchy. This
complies with the requirement to build taxonomies on the basis that every member of the child
class is also a member of the parent class[20].
Second, we reviewed the definitions of the NTDS terms and found that some of them are
defined in a circular manner, viz. the label or parts thereof are used as the definition. For instance,
NTDS defines the term ICD-10 INJURY DIAGNOSIS as “diagnosis related to all identified
Table 1
OWL implementation status of new OOSTT terms based onNTDS terms. Terms marked * were imple-
mented during data-driven extension; terms marked ! were imported from IDO.
NTSD term OOSTT term In OWL?
Age Units age unit Yes
ICD-10 Injury Diagnosis international classification of disease Yes
tenth revision injury diagnosis
Drug Screen drug screening Yes
ED Discharge emergency department discharge date Yes
Initial ED/Hospital Oxygen Saturation initial oxygen saturation Yes
Initial ED/Hospital Temperature initial temperature Yes
Trauma Surgeon Arrival Time trauma surgeon arrival time Yes
Primary Method of Payment primary payment method Yes
Deep Vein Thrombosis (DVT) deep vein thrombosis Yes
Acute Kidney Injury (AKI) acute kidney injury No
Severe Sepsis severe sepsis Yes *!
Unplanned Admission To ICU unplanned intensive care unit admission Yes*
Unplanned Intubation unplanned endotracheal intubation Yes*
ICD-10 Hospital Procedures international classification of diseases No
tenth revision hospital procedure
Incident Location Zip/Postal Code incident zip/postal code No
Patient’s Occupational Industry patient’s occupation No
Protective Devices protective device No
Total ICU Length Of Stay intensive care unit length of stay data item Yes
Advance Directive Limiting Care advance directive limiting care No
Alternate Home Residence homeless status Yes
temporary address Yes
undocumented status No
injuries”. The definition does not explicate what the words ”diagnosis” and ”injury” actually
mean. Additionally, the phrase ICD-10 is addressed by the definition. We propose an alternative
definition: “An information content entity that is about an injury borne by a patient and that
expresses the diagnosis in an ICD 10 code”. This definition uses the next superclass (genus),
”information content entity”, and gives differentiating characteristics. Thus, the term is defined
by it being a member of a specific superclass and its specific, defining characteristics, following
the format suggested by OBO Foundry principles[21]. Each definition was rewritten following
this format.
Table 2
Examples of classes imported to OOSTT using MIREOT
Source Ontology Terms
OMRSE patient discharge, admission process
MONDO respiratory failure, acute respiratory distress syndrome, myocardial
infarction, cardiac arrest, pulmonary embolism, stroke disorder
IDO severe sepsis
MAXO endotracheal intubation
The revised and edited terms yielded 22 potential new ontology classes. Three of those terms,
severe sepsis, unplanned intensive care unit admission, and unplanned endotracheal intubation,
were pushed to the data-driven extension process for consistency. The other 19 terms were
manually checked against OBO Foundry ontologies, to prevent duplication. No duplication
with those 19 terms has been detected. Hence, they were implemented in OWL[22] using the
Protege ontology editor [23]. A complete list of all 20 NTSD terms, the proposed OOSTT label,
and the OWL implementation status can be found in Table 1. The implementation was done by
manually creating a novel OWL file. This file imported BFO [15]and the Information Artifact
Ontology (IAO)[24]. The development team (DM and MB) decided to import IAO, since 10
of the initial potential new OOSTT classes were subclass of ’information content entity’ [24],
which is not part of BFO, nor does BFO contain another class representing information. The
resulting OWL file was merged with the latest release of OOSTT resulting in OOSTT release
version 2024-01-25 (https://github.com/OOSTT/OOSTT/tree/2024-01-25).
2.3.2. Data-driven Extension
In parallel to the terminology-driven approach, we needed to extend OOSTT to cover the 3
patient outcomes the TIPTOE project focuses on: mortality, length of stay, and major com-
plications. For major complications the representation of clinical conditions and situations
was also needed. To represent the various major complications, we required classes from
MONDO[12], IDO[25], and MAxO[14] using a MIREOT[26] Protege Plugin. The details on the
imported classes can be found in Table 2. To represent mortality using discharge disposition
data and length of stay, we used MIREOT to import two classes from OMRSE[10] (see Table
2). The minimum information to reference an external ontology term (MIREOT) guidelines
were initially created to develop the Ontology for Biomedical Investigation (OBI)[27]. MIREOT
enables the reuse of pre-existing ontology resources, e.g., classes and object properties to avoid
duplication[26]. MIREOT is ”independent of any design principle, and provides a mechanism by
which external ontology terms can be selectively imported, even if they do not use a particular
upper ontology(...)”[26]. This is particularly relevant for developing multiple ontologies in
one unified environment, for instance, in the OBO Foundry[8, 9]. Hanna et al. implemented
the MIREOT guidelines into a Protege plug in that allows to drag and drop classes and object
properties from existing OBO Foundry ontologies in ontology project developed in Protege[28].
Examples of the classes proposed based on the NTDS are given in Table 1. In total, 36 classes
were imported to OOSTT in this step. In addition, 8 classes were created newly in OOSTT:
Figure 2: Visual representation of the created OWL file. BFO and IAO subclasses represented by white
boxes and newly created NTDS terms categorized highlighted in blue.
unplanned intensive care unit admission process, intensive care unit admission process, un-
planned surgical procedure, unplanned endotracheal intubation, cardiopulmonary resuscitation,
patient discharge disposition information, total intensive care unit length of stay data item.
Three of these classes had also been identified in the terminology-driven extension approach.
3. Results
All terms, their NTDS definitions, and the definitions revised in accordance with the principles
and practices mentioned above can be found here: https://tinyurl.com/OOSTTe. Figure 2 shows
how 11 terms implemented in OWL as part of the terminology-driven extension approach
extend BFO. The definitions have been revised by KWS, our trauma surgery expert. In total,
OOSTT was expanded by 55 classes; 36 imported classes and 19 new classes. The OOSTT
release that includes all extensions discussed in this paper can be accessed at: http://purl.oboli-
brary.org/obo/oostt/release/2024-01-25/oostt.owl.
The ontology development described here makes these outcome patient measures available
in the KPE. We are currently in the process of conducting a usability study of the KPE. Due to
ongoing data collection and analysis, this utilizes with a virtual data set that includes instance
data on patient outcomes. While the usability study is still ongoing and results from it are not
yet available, in preparation we internally tested the functionality of the tool with the newly
created OOSTT classes. We can report that the ontology extension described in this paper,
allows successfully retrieving information on the following two topics from the TIPTOE triple
store using the KPE:
1. How does the number of general surgeons with ATLS certification at your trauma center
affect the number of major complications including death?
2. How does the intercorrelation between neurosurgeons taking call exclusively and the
number of neurosurgeons with certified 18 hours of trauma-specific CME affect length of stay?
In addition to those two data exploration scenarios, the extension also enable multiple other
scenarios related to patient outcomes and more detail on patient demographics, diagnostics,
and treatment. Once the usability testing, consisting of two formative usability studies and one
summative study at the end of the project are completed, the tool will be made available.
4. Discussion
In previous development of OOSTT, we added definitions for domain experts in addition to the
genus-differentia definitions and we validated those definitions with domain experts. This step
has not yet been undertaken with this extension, but we plan to address this issue in the next
project year.
At this point, we have not yet conducted statistical analyses to assess which features of
trauma centers affect patient outcomes. As the results of those statistical analyses become
available, specifically the relationships between organizational features and patient outcomes
those will be added to OOSTT, too.
5. Conclusions
The integration of extensions into the Ontology of Organizational Structures in Trauma Centers
and Trauma Systems (OOSTT) significantly enhances our ability to discern the nuanced effects
that organizational structures and parameters exert on patient outcomes. This methodologi-
cal advancement facilitates a novel approach to examining the intricate relationships within
healthcare delivery systems. With the new extension OOSTT covers not only the multispecialty
composition of trauma care teams, but also key patient outcomes along the care pathway,
such as admission, readmission, and discharge. In addition, we also capture diagnoses that
represent unintended medical comlications, such as sever sepsis. By deploying this enhanced
ontology, our representation can encompass both multispecialty and single clinical service
outcomes. For instance, it enables a thorough examination of multispecialty outcomes, such
as readmission rates to the Intensive Care Unit (ICU), alongside the analysis of outcomes for
specific conditions, like the care pathway for isolated femur fractures. This dual perspective
permits a comprehensive exploration of complex system dynamics, specifically focusing on
their impact on clinical care. Through this lens, we gain a more profound understanding of
the interplay between organizational structures and patient health results, providing valuable
insights into potential areas for improvement.
Acknowledgments
Research reported in this paper was partially supported by the National Institute of General
Medical Sciences of the National Institutes of Health under award number R01GM111324.
DM’s contribution stems from 2023 Summer Research Internship Program, UAMS, funded by
the Health Resources and Services Administration (HRSA) of the U.S. Department of Health and
Human Services (HHS). The contents are those of the author(s) and do not necessarily represent
the official views of, nor an endorsement, by HRSA, HHS or the U.S. Government.
We are grateful for the recommendations we received from 3 anonymous reviewers.
References
[1] S. C. Curtin, B. Tejada-Vera, B. A. Bastian, Deaths: leading causes 2020, National Vital
Statistics Report 72 (2023).
[2] S. Shafi, R. M. Stewart, A. B. Nathens, R. S. Friese, H. Frankel, L. M. Gentilello, Significant
variations in mortality occur at similarly designated trauma centers, Arch Surg 144 (2008)
64–68. doi:10.1001/archsurg.2008.509 .
[3] M. T. Cudnik, M. R. Sayre, S. M. S. Brian Hiestand, Are all trauma centers created equally?
a statewide analysis., Acad Emerg Med 17 (2010) 701–708. doi:10.1111/j.1553- 2712.
2010.00786.x .
[4] J. Utecht, et al., Oostt: a resource for analyzing the organizational structures of trauma
centers and trauma systems, in: P. Jaiswal, R. Hoehndorf, C. Arighi, A. Meier (Eds.),
Proceedings of the Joint International Conference on Biological Ontology and BioCreative,
volume 1747 of CEUR-WS, 2016.
[5] J. Utecht, et al., , development and validation of a controlled vocabulary: an owl repre-
sentation of organizational structures of trauma centers and trauma systems. stud health
technol inform 2019; 264:403-407., Stud Health Technol Inform 264 (2019) 403–407.
[6] M. Brochhausen, et al., Collecting data on organizational structures of trauma centers:
The cafe web service. 2020 jul 29;5(1):e000473. doi: 10.1136/tsaco-2020-000473. ecollec-
tion 2020, Journal of Trauma Surgery and Acute Care Open 29 (2020). doi:10.1136/
tsaco- 2020- 000473 .
[7] M. R. Hemmila, et al., The trauma quality improvement program: pilot study and
initial demonstration of feasibility, J Trauma 68 (2010) 253–262. doi:10.1097/TA.
0b013e3181cfc8e6 .
[8] B. Smith, et al., The obo foundry: coordinated evolution of ontologies to support biomedical
data integration, Nat Biotechnol. 25 (2007) 1251–1255. doi:10.1038/nbt1346 .
[9] R. Jackson, et al., Obo foundry in 2021: operationalizing open data principles to evaluate
ontologies, Database (Oxford) (2021). doi:10.1093/database/baab069 .
[10] A. Hicks, J. Hanna, D. Welch, M. Brochhausen, W. R. Hogan, The ontology of medically
related social entities: recent developments, J Biomed Semantics 12 (2016). doi:10.1186/
s13326- 016- 0087- 8 .
[11] Ontology for Modeling and Representation of Social Entities, OMRSE, 2024. URL: http:
//purl.obolibrary.org/obo/omrse.owl.
[12] N. A. Vasilevsky, et al., Mondo: Unifying diseases for the world, by the world, MedRxiv
(2022). doi:10.1101/2022.04.13.22273750 .
[13] Infectious Disease Ontology, IDO, 2024. URL: http://purl.obolibrary.org/obo/ido.owl.
[14] L. C. Carmody, et al., The medical action ontology: A tool for annotating and analyzing
treatments and clinical management of human disease, MedRxiv (2023). doi:10.1016/j.
medj.2023.10.003 .
[15] International Organization for Standardization, Information technology — top-level ontolo-
gies (tlo) — part 2: Basic formal ontology (bfo) (iso standard no. iso/iec 21838-2:2021(en),
2021. URL: http://www.iso.org/obp/ui/en/#iso:std:iso-iec:21838:-2:ed-1:v1:en.
[16] American College of Surgeons, National trauma data standard. data dictionary,
2023 admissions, 2022. URL: http://www.facs.org/quality-programs/trauma/quality/
national-trauma-data-bank/national-trauma-data-standard/.
[17] M. A. Hornor, C. Hoeft, A. B. Nathens, Quality benchmarking in trauma: from
the ntdb to tqip, Current Trauma Reports 4 (2018) 160–69. doi:doi.org/10.1007/
s40719- 018- 0127- 1 .
[18] D. Schober, et al., Survey-based naming conventions for use in obo foundry ontology
development, BMC Bioinformatics 27 (2009). doi:10.1186/1471- 2105- 10- 125 .
[19] Open Biological and Biomedical Ontology Foundry (OBO Foundry), Principle:
Naming conventions (principle 12), 2024. URL: https://obofoundry.org/principles/
fp-012-naming-conventions.html.
[20] R. Arp, B. Smith, A. D. Spear, Building Ontologies with Basic Formal Ontologies. Cam-
bridge/Mas., 2015., MIT Press, Cambridge, MA, 2015.
[21] Open Biological and Biomedical Ontology Foundry (OBO Foundry), Principle:
Textual definitions (principle 6), 2024. URL: https://obofoundry.org/principles/
fp-006-textual-definitions.html.
[22] World Wide Web Consortium (W3C), Owl 2 web ontology language primer (second edition),
2012. URL: http://www.w3.org/TR/owl2-primer/.
[23] M. Musen, The protégé project: A look back and a look forward, AI Matters. Association
of Computing Machinery Specific Interest Group in Artificial Intelligence 1 (2015). doi:10.
1145/2757001.2757003 .
[24] information-artifact-ontolgy, Iao, 2024. URL: https://github.com/
information-artifact-ontology/IAO.
[25] S. Babcock, J. Beverley, L. G. Cowell, B. Smith, The infectious disease ontology in the age
of covid-19, J Biomed Semantics 12 (2021). doi:10.1186/s13326- 021- 00245- 1 .
[26] M. Courtot, et al., Mireot: the minimum information to reference an external ontology
term, Applied Ontology 6 (2011) 23–33. doi:DOI10.3233/AO- 2011- 0087 .
[27] A. Bandrowski, et al., The ontology for biomedical investigations, PLoS One 11 (2016)
e0154556. doi:10.1371/journal.pone.0154556 .
[28] J. Hanna, et al., Simplifying mireot; a mireot protege plugin, CEUR-WS 914 (2016).