キタノ トモヤ   KITANO TOMOYA   北野　倫哉    所属   朝日大学  歯学部 口腔病態医療学講座 歯科放射線学    職種   助教
言語種別	英語
発行・発表の年月	2023/07
形態種別	研究論文
査読	査読あり
標題	Detection of aspiration from images of a videofluoroscopic swallowing study adopting deep learning
執筆形態	共著
掲載誌名	Oral Radiology
掲載区分	国外
国際共著	国際共著
著者・共著者	Yukihiro Iida · Janne Näppi · Tomoya Kitano · Toru Hironaka · Akitoshi Katsumata · Hiroyuki Yoshida
概要	Deep learning was adopted to detect aspiration in dys- phagia from images obtained from VFSS videos. Three deep-learning models were implemented using an inter- active tool for the rapid prototyping of the models. Our results based on 18,333 video frame images showed that deep learning can be used for classifying the presence or absence of aspiration in a VFSS with a high accuracy exceeding 90%.