| 研究実績の概要 |
At the final year of the study period, our team has successfully completed all the proposed research tasks according to predefined schedule. Specifically, we have developed and validated annotation guidelines for medication errors' incident reports, and made annotation training materials available for further use. In terms of AI modeling tasks, we have developed and evaluated various rule-based and deep learning models (including BiLSTM, BERT, GPT-3) for incident reports' NER automatic labeling. Furthermore, we have launched the prototype AI-empowered incident report system on AWS server (accessible through https://med-incident-report.link/). In this reporting year, we have also published our findings to various reputable international conferences (including Medinfo 2021) and achieved more journal publication, such as.
-Wong ZS, Rigby M. Identifying and addressing digital health risks associated with emergency pandemic response: Problem identification, scoping review, and directions toward evidence-based evaluation. International journal of medical informatics. 2022;157:104639.
Our AI patient safety research also gained local and international attentions. For instance, Dr Shin Ushiro continued to communicate our ongoing project outcomes as an example project at WHO Policy Makers’ Forum Feb 23rd-24th 2022. He presented our research on application of AI for analysis of adverse event on medication therapy in the global presentations. Furthermore, the research outcome is planned to communicate at other future global events.
|
