@techreport{oai:ipsj.ixsq.nii.ac.jp:00211726,
 author = {江頭, 彩夏 and 船元, 望補 and 有地, 祐人 and 出田, 良輔 and 倉田, 博之 and 坂井, 宏旭 and Ayaka, Egashira and Miho, Funamoto and Yuto, Ariji and Ryosuke, Ideta and Hiroyuki, Kurata and Hiroaki, Sakai},
 issue = {23},
 month = {Jun},
 note = {脊髄損傷とは脊椎が損傷，圧迫されることで，脊髄が損傷することである．早期の段階から受傷者に合わせたリハビリテーションを行うことで ADL の改善を試みることが重要である．本研究では受傷後早期のデータから 6 か月後の重症度を予測することを目的とした．初めに 132 項目ある筋力評価・感覚評価から重要な因子を選ぶため，いくつかの手法を用いて特徴選択を行う．次に選ばれた因子から順序ロジスティック回帰を使用し，作成された比例オッズモデルから，重症度分類 AIS が A，B，C，D になる確率をそれぞれ求める．正解率約 98% の機能回復予測モデルを作成できた．, Spinal cord injury is the result of damage or compression of the spine. It is important to try to improve ADL by providing rehabilitation tailored to the injured person from the early stage. The purpose of this study is to predict the severity of the injury at 6 months from the early post-injury data. First, feature selection is performed using several methods to select important factors from 132 muscle strength and sensory assessments. Then, from the selected factors, we used ordered logistic regression to determine the probability that the severity classification AIS would be A, B, C, or D, respectively, from the created proportional odds model. We are able to create a functional recovery prediction model with a correct answer rate of about 98%.},
 title = {順序ロジスティック回帰を用いた外傷性脊髄損傷患者の機能回復予測},
 year = {2021}
}