gms | German Medical Science

64. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e. V. (GMDS)

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie

08. - 11.09.2019, Dortmund

Semantic verification during BPMN modeling of healthcare processes by integrating Shapes Constraint Language (SHACL) graphs

Meeting Abstract

Search Medline for

  • Daniel Keuchel - University of Applied Sciences and Arts Dortmund, Dept. of Computer Science, Dortmund, Germany
  • Britta Böckmann - University of Applied Sciences and Arts Dortmund, Dept. of Computer Science, Dortmund, Germany
  • Nicolai Spicher - University of Applied Sciences and Arts Dortmund, Dept. of Computer Science, Dortmund, Germany

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie. 64. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e.V. (GMDS). Dortmund, 08.-11.09.2019. Düsseldorf: German Medical Science GMS Publishing House; 2019. DocAbstr. 227

doi: 10.3205/19gmds097, urn:nbn:de:0183-19gmds0978

Published: September 6, 2019

© 2019 Keuchel et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 License. See license information at http://creativecommons.org/licenses/by/4.0/.


Outline

Text

Introduction: Process models have a wide field of applications in healthcare, ranging from the visual representation of clinical guidelines to knowledge representation in decision-support systems [1], with the Business Process Model Notation (BPMN) being the de-facto standard tool for graphical process representation [2], [3].

The aim of our work is to integrate implicit expert knowledge during the process of BPMN modeling by non-expert users (e.g. physicians).

Related work: BPMN enforces syntax rules only, therefore several algorithms for enabling semantic verification have been proposed, e.g. using Petri nets [4], [5] or ontology-based approaches by integrating Web Ontology language (OWL) graphs [6], [7].

Method: In contrast, we aim for enabling semantic verification of BPMN models by using SHACL graphs [8]. Unlike OWL, SHACL is not based on the open world assumption and therefore allows strict constraints.

Our future goal is an algorithm running in parallel to a BPMN modeler software and giving the user meaningful real-time feedback when a semantic error occurs. This requires extraction of semantic knowledge from SHACL graphs with a fast runtime. In this work, we perform an initial proof-of-concept study of this approach.

Results: We developed a Java application which reads a BPMN model [9] and a SHACL graph [10] and checks if the first is semantically correct by using information from the latter. The algorithm processes all entities of the BPMN model and matches them to entities in the SHACL graph, instantiates manually-predefined “rule templates” and checks whether they are violated.

Evaluation: We chose the clinical algorithm of acute cough diagnosis accompanying a clinical guideline [11]. We manually defined two rule templates: “Check if user task A lies in the correct swim lane B” and “Check if user task C follows user task D” and an expert user created a SHACL graph covering the semantic knowledge (e.g. “An examination must be performed by a physician.” and “A patient is not allowed to stop medicine intake without a physician’s instruction”). On each run, the algorithm matches BPMN to SHACL entities and creates instances of the corresponding rule templates, for example: “Check if the user task abdominal examination lies in swim lane physician” and “Check if user task stop medicine intake follows user task instruction to stop medicine”.

Two non-expert users (computer science undergraduates) were asked to use an off-the-shelf software to complete a BPMN model of the clinical algorithm. Predefined BPMN elements were given but without any order or connection. The algorithm ran in parallel and displayed an error message if rules were violated. The user described this feedback as supportive and the algorithm showed short processing times (approx. 0.2s).

Discussion: The results show the potential of the proposed approach. In future work, we aim for algorithmic extraction of rule templates overcoming the need for manual rule definition and, evaluation of the pros and cons of our approach compared to others.

Acknowledgement: Das diesem Bericht zugrundeliegende Vorhaben wurde mit Mitteln des Bundesministeriums für Bildung und Forschung unter dem Förderkennzeichen 13GW0210B gefördert. Die Verantwortung für den Inhalt dieser Veröffentlichung liegt beim Autor.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.


References

1.
Yao W, Kumar A. CONFlexFlow: Integrating Flexible clinical pathways into clinical decision support systems using context and rules. Decision Support Systems. 2013;55(2):499-515.
2.
Müller R, Rogge-Solti A. BPMN for healthcare processes. In: 3rd Central-European Workshop on Services and their Composition, Services und ihre Komposition, ZEUS 2011; 2011 February 21-22; Karlsruhe, Germany. 2011; p. 65-72.
3.
Ruiz F, et al. Business process modeling in healthcare. Stud Health Technol Inform. 2012;179:75-87.
4.
Awad A, Decker G, and Lohmann N. Diagnosing and Repairing Data Anomalies in Process Models. Lecture Notes in Business Information Processing. 2009;43:5-16.
5.
Weber P, Filho J, Bordbar B, Lee M, Litchfield I, Backman R. Automated conflict detection between medical care pathways. Journal of Software: Evolution and Process. 2017;30(7):e1898.
6.
OWL - Semantic Web Standards. 2019 [Accessed 3 June 2019]. Available from: https://www.w3.org/OWL/ External link
7.
Fellmann M, Hogrebe F, Thomas O, Nüttgens M. An Ontology-driven Approach to Support Semantic Verification in Business Process Modeling. In: Werner Esswein, Klaus Turowski, Martin Juhrisch, editors. MobIS 2010 – Modellierung betrieblicher Informationssysteme. Bonn: Gesellschaft für Informatik; 2010. p. 99-110. (Lecture Notes in Informatics; P-171).
8.
Shapes Constraint Language (SHACL). 2019 [Accessed June 2019]. Available from: https://www.w3.org/TR/shacl/ External link
9.
GitHub. camunda BPMN model. 2019 [Accessed 3 June 2019]. Available from: https://github.com/camunda/camunda-bpmn-model External link
10.
GitHub. SHACL. 2019 [Accessed 3 June 2019]. Available from: https://github.com/TopQuadrant/shacl External link
11.
Kardos P, Dinh Q, Fuchs K. Leitlinie der Deutschen Gesellschaft für Pneumologie und Beatmungsmedizin zur Diagnostik und Therapie von erwachsenen Patienten mit Husten. Pneumologie. 2019;73(03):143-180.