University of Twente Student Theses
Assessing Doctor and Patient explanation and information needs of Explainable Artificial Intelligence
Kazokas, Matas (2024) Assessing Doctor and Patient explanation and information needs of Explainable Artificial Intelligence.
PDF
1MB |
Abstract: | Artificial Intelligence (AI) has proven to significantly increase healthcare instrument performance but faces reluctance from medical experts due to its 'black box' nature. Explainable Artificial Intelligence (XAI) aims to solve this problem by making AI models more transparent and tailoring explanations to specific user needs, including medical professionals and patients. This study explores how XAI can effectively communicate AI-based decisions in healthcare focusing on tailoring explanations to meet the user group‘s needs, thereby increasing the knowledge of user-centered needs in the medical domain. Applying a mixed model approach, participants first answered open-ended questions to identify their information needs, followed by a case scenario involving a low grade glioma, which was investigated using multiple choice questions. The thematic analysis identified nine themes, showing patient‘s interest in understanding their illness and treatment options, and the doctor‘s focus on background knowledge and differential diagnosis. The quantitative findings indicated a preference for long, descriptive mechanistic explanations for the patient, and brief, causal explanations among the doctor group, with the latter group expressing lower satisfaction overall. This research increases understanding of XAI in medicine, accentuating the importance of user-centered designs and the need for tailored explanations to meet the diverse expertise of different user groups. |
Item Type: | Essay (Bachelor) |
Faculty: | BMS: Behavioural, Management and Social Sciences |
Subject: | 77 psychology |
Programme: | Psychology BSc (56604) |
Link to this item: | https://purl.utwente.nl/essays/98067 |
Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Repository Staff Only: item control page