Using Artificial Intelligence-Based Argument Theory To Generate Automated Patient Education Dialogues: An Interactive Educational Dialogue System For Families Of Children With Juvenile Idiopathic Arthritis

Abstract

Juvenile Idiopathic Arthritis (JIA) is a chronic rheumatic disease affecting between 1 and 4 out of 1000 children in Canada, with outcomes including pain, prolonged dependence on medications, and disability. To allow families to effectively self-manage chronic conditions, such as JIA, patient education is necessary. Families of children with JIA need access to the Patient Education Materials (PEM) provided by their Rheumatology clinic outside of clinic visits in a way that allows them control over what content they receive and when they receive it. This work aims to address this educational gap by using a knowledge management approach to explore how an artificial intelligence method based on the Toulmin model of argument could be used to construct an interactive dialogue that allows users to engage with PEM. Content from the PEM was computerized using a knowledge model based on the Toulmin model of argument which was also leveraged to manage the structure of the dialogue. We have evaluated the dialogue of the resulting system through cognitive walkthroughs and semi-structured interviews with JIA domain experts. The results of this study show that these methods show great promise for providing quality information to families of children with JIA when they need it.