@techreport{oai:ipsj.ixsq.nii.ac.jp:00184617,
 author = {早野, 順一郎 and 吉田, 豊 and 古川, 由己 and 湯田, 恵美 and Emi, Yuda and Yutaka, Yoshida and Yuki, Furukawa and Junichiro, Hayano},
 issue = {1},
 month = {Nov},
 note = {近年の心拍や脈拍を検出するウェアラブルセンサの普及により，心拍時系列のビッグデータが形成される状況が生じているが，そこに潜むプライバシーリスクはあまり認識されていない．ALLSTAR プロジェクトの 24 時間心電図から得られた心臓の拍動間隔時系列のみの自動解析による実証実験で，突然の意識消失や突然死につながる 30 秒以上の持続性心室頻拍が 247,712 例の解析で 164 例 (0.066%) に，また日中の強い眠気をもたらす重症睡眠時無呼吸 (無呼吸低呼吸指数 > 30 / 時) が，196,091 件の解析で，男性の 7.2%，女性の 2.6% に発見された．これらの情報が個人や特定の集団の情報として流出すれば，本人や所属集団にとって大きなプライバーリスクとなる可能性がある．, With the rapid progress of wearable sensors, continuous monitoring of biological signal in daily life has become widespread. Particularly, for heart / pulse rate signals that can be monitored relatively easily, enormous continuous signals are accumulated and big data are being formed. However, given the healthcare information that can be extracted from continuous heart / pulse rate signals, our perceptions of the potential information security risks of their big data may be inadequate. In this study, we analyzed 24-hr heart rate data obtained in Allostatic State Mapping by Ambulatory ECG Repository (ALLSTAR) project. By the analysis of data in 247,712 subjects, we were able to find 164 (0.066%) subjects who showed episodes of ventricular tachycardia sustained for >30 sec that can lead sudden syncope or death. We also detected the typical heart rate pattern of severe sleep apnea with apnea-hypopnea index > 30 / h that can cause strong daytime sleepiness and increased health risk in 7.2% of males and 2.6% of females among 196,091 subjects. If such information leaks along with the subject or group identity information, it could be a major privacy risk for individuals and related groups.},
 title = {心拍時系列ビッグデータに潜在するプライバシーリスク},
 year = {2017}
}