A Verified, Executable Formalism for Resilient and Pervasive Guideline-Based Decision Support for Patients

Nick L.S. Fung*, Marten J. van Sinderen, Valerie M. Jones, Hermie J. Hermens

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

Abstract

We present an executable formalism for clinical practice guidelines, with the aim of providing pervasive and evidence-based decision support to patients. Unlike traditional formalisms that capture the control flow between tasks, we focus on data flow, with tasks modeled as processes that execute in parallel. By parallelizing and distributing guideline knowledge, each device that constitutes the patient’s pervasive healthcare system can provide decision support independently, avoiding single points of failure. This distribution also enables dynamic system re-configurations, increasing its resilience against evolving requirements and changing communications environments. Our model recognizes four types of processes: Monitoring, Analysis, Decision and Effectuation. These processes were specified using (axiomatic) set theory and implemented as a set of libraries on top of Rosette, which supports execution of the formalism and verification of it using constraint solvers. The formalism was also tested by formalizing a complete clinical guideline for diabetes management, which yielded a Rosette program that was then tested on simulated patient data. The major point of clinical relevance is enhancing the quality and safety of decision support delivered to patients.

Original languageEnglish
Title of host publicationArtificial Intelligence in Medicine
Subtitle of host publication18th International Conference on Artificial Intelligence in Medicine, AIME 2020, Proceedings
EditorsMartin Michalowski, Robert Moskovitch
PublisherSpringer
Pages427-439
Number of pages13
ISBN (Electronic)978-3-030-59137-3
ISBN (Print)978-3-030-59136-6
DOIs
Publication statusPublished - 26 Sep 2020
Event18th International Conference on Artificial Intelligence in Medicine, AIME 2020 - Virtual Conference, United States
Duration: 25 Aug 202028 Aug 2020
Conference number: 18

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12299 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference18th International Conference on Artificial Intelligence in Medicine, AIME 2020
Abbreviated titleAIME 2020
CountryUnited States
CityVirtual Conference
Period25/08/2028/08/20

Keywords

  • Computerized clinical practice guidelines
  • Data flow modeling
  • Diabetes management
  • Formal specification
  • Knowledge representation
  • Pervasive healthcare
  • Verification and validation

Fingerprint Dive into the research topics of 'A Verified, Executable Formalism for Resilient and Pervasive Guideline-Based Decision Support for Patients'. Together they form a unique fingerprint.

Cite this