Model-based prediction of oncotherapy risks and side effects in bladder cancer

Barki, Chamseddine; Rahmouni, Hanene; Labidi, Salam

Abstract

The prediction of cancer treatment side-effects requires the capturing of complex biophysical therapy parameters and the integration of different medical knowledge elements. In relation with radiotherapy, it is widely observed that the uncontrolled processes or undefined radiation therapy dose can decline the state of treatment. Precisely, the inability to manage the flow of available information, usually provided in heterogeneous formats, made it complicated to oversee and predict risks and effects of a prescribed treatment protocol. We think that, the optimization of knowledge representation and modelling in the context of evidence-based medicine can support the automated prediction of risks and side effects in oncotherapy. The following manuscript describes our methodology used for the design of a bladder cancer treatment side effects ontology embedded with evidence-based semantic rules and queries. Treatment knowledge is represented along with a particular consideration to the modelling of its referred risks and side effects. Our ontology model helps in improving the streamlining of medical practices and clinical decision-making. Within our semantic web approach, better strategies are applied for treatment selection with reference to possible side effects. Our ontology depicts real world scenario of developing treatment-related side effects. Furthermore, it is a clinical decision support system founding tool that highlights treatments efficiency. Our model shares treatment knowledge, facts and effects. Moreover, it includes medical evidence and incorporates a semantic rule base for systemic prediction results.

• Journal Article Type: Article
• Acceptance Date: Jul 6, 2020
• Online Publication Date: Feb 22, 2021
• Publication Date: 2021
• Deposit Date: Jan 1, 2021
• Publicly Available Date: Mar 26, 2021
• Journal: Procedia Computer Science
• Print ISSN: 1877-0509
• Publisher: Elsevier
• Peer Reviewed: Peer Reviewed
• Pages: 818-826
• DOI: https://doi.org/10.1016/j.procs.2021.01.235
• Public URL: https://uwe-repository.worktribe.com/output/6966092