[3S1-OS-7b] OS-7

[3S1-OS-7b-03] Challenges and Applications of Artificial Intelligence in Corpus Creation Using Case Reports of Rare and Intractable Diseases

〇Eisuke Dohi¹, Yuka Tateishi², Toyofumi Fujiwara³, Yasunori Yamamoto³ (1. National Center of Neurology and Psychiatry, 2. Japan Science and Technology Agency Department of NBDC Program, 3. Database Center for Life Science Joint Support-Center for Data Science Research Research Organization of Information and Systems)

Keywords:Medical Informatics, Textdata

For effective diagnostic support using Artificial Intelligence (AI), an accurate case-based corpus is essential, yet it presents several challenges. This abstract proposes solutions to these challenges utilizing AI.Sharing case reports is difficult due to privacy concerns, and the main challenges include text extraction from PDFs, variability in disease name notation, structuring clinical data, normalizing text data, and extracting and annotating information. Particularly, text extraction from PDFs is technically challenging, and variability in disease name notation is common. Systems like CaseSharing have proven effective for structuring clinical data, and normalization of text data has been somewhat resolved using Large Language Models (LLMs). Furthermore, LLMs enable extraction of information following a timeline, but annotation remains a challenge.From these experiences, it is believed that the application of AI plays a crucial role in dataset creation. Moving forward, we aim to deepen the discussion on more effective utilization of these technologies.