King's College London

Research portal

Towards an AI Planning-Based Pipeline for the Management of Multimorbid Patients

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

Malvika Rao, Martin Michalowski, Szymon Wilk, Wojtek Michalowski, Amanda Coles, Marc Carrier

Original languageEnglish
Title of host publicationArtificial Intelligence in Medicine - 20th International Conference on Artificial Intelligence in Medicine, AIME 2022, Proceedings
EditorsMartin Michalowski, Syed Sibte Raza Abidi, Samina Abidi
PublisherSpringer Science and Business Media Deutschland GmbH
Pages14-23
Number of pages10
ISBN (Print)9783031093418
DOIs
Published2022
Event20th International Conference on Artificial Intelligence in Medicine, AIME 2022 - Halifax, Canada
Duration: 14 Jun 202217 Jun 2022

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume13263 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference20th International Conference on Artificial Intelligence in Medicine, AIME 2022
Country/TerritoryCanada
CityHalifax
Period14/06/202217/06/2022

Bibliographical note

Funding Information: Acknowledgements. We thank Jean-Luc Blais-Amyot and Maxime Côté-Gagnéfor their programming work on the automated translation component. We thank the reviewers for their helpful feedback. This research was supported by funding from the Telfer Health Transformation Exchange and the Natural Sciences and Engineering Research Council of Canada. Publisher Copyright: © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

King's Authors

Abstract

Treatment of patients with multimorbidity is one of the greatest challenges for clinical decision support. While evidence-based management of specific diseases is supported by clinical practice guidelines, concurrent application of multiple guidelines requires checking for possible adverse interactions between interventions and mitigating them, before a management plan is constructed. In earlier work, we developed an approach that casts the problem of multimorbidity management as an AI planning problem. In this paper we build on this earlier work and make progress towards creating a pipeline that inputs disease and patient-specific information and outputs a management plan. We describe research focused on selected aspects of pipeline development and illustrate these aspects with a clinical case implemented using the PDDL planning language and the OPTIC planner.

View graph of relations

© 2020 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454