KAIST XAI Tutorial 2024 | XAI for Clinical Decision Support | Jihyeon Seong (KAIST)

In the medical field, it is essential not only to achieve high performance in AI models but also to explain the reasons behind their predictions to ensure model reliability. Especially for applying AI in real hospitals, both high performance and the ability to explain disease causes tailored to individual patients are necessary. This session introduces prediction and patient-specific cause-explanation techniques using AI for Acute Kidney Injury (AKI). First, we present the development process of an AKI prediction model for practical clinical application. Next, we introduce explanation algorith…