睡眠在人的生活中扮演很重要的腳色，然而人們可能患有各式各樣的睡眠障礙，其 中最常見的就是失眠，根據台灣睡眠醫學學會在 2017 年的調查，全台每十位就有一位受慢性失眠症所苦，失眠儼然已成為非常嚴重的心理健康問題。傳統上評估睡眠 的方法是使用多通道生理記錄儀 (PSG) 來記錄整晚病患的生理訊號進行分析，然而 這需要在醫院或是睡眠中心量測，並且因為黏貼許多電擊可能造成不適。睡眠的狀 況也需要長期的評估，因此，我們開發一腕式活動紀錄器用來偵測活動，該裝置充 飽電後可以長達一個月的紀錄，並可將演算法特徵儲存在裝置裡頭，透過無線的方 式傳送給行動裝置，手機端可輸入上下床的時間即可進行 Wake­Sleep 演算法獲得客 觀的睡眠品質之評估。我們錄製了 116 個受試者整晚的睡眠活動訊號，並透過演算 法計算清醒­睡眠的結果與多通道生理監控儀 (PSG) 的判讀結果比較，整體準確度、 靈敏度以及特異度分別為 90.84%、55.47%、94.71%，在睡眠效率、入睡時間、睡後 覺醒時間以及總睡覺時間上，與 PSG 判讀結果的平均絕對誤差分別為 4.45%、6.92 分鐘、17.68 分鐘，18.92 分鐘。為了防止使用者沒戴手環造成演算法的誤判，亦配 有未戴事件偵測，準確率達到了 91.74%。本系統與 PSG 的判讀結果有九成以上的一 致性，適合提供長期的睡眠評估。

Sleep plays an important role in the daily activities of humans. However, humans may suffer from various sleep disorders such as insomnia, which is the most common specific sleep disorder. According to a survey conducted by the Taiwan Society of Sleep Medicine in 2017, one tenth of people in the Taiwan suffers from chronic insomnia. So insomnia have become a very serious mental health problem in Taiwan. Traditionally, the evaluation method for sleep quality is use PSG. However, the subject must stay in sleep laboratory or sleep center and place electrodes on he/his body. It may make subject feel uncomfortable.Sleep also need long term evaluation. Therefore, we develop a wrist actigraphy for detect activity, the device can record data for a month without charging. The algorithm features can be stored in the device and transmitted to the mobile device wirelessly. The mobile phone can input the in and out of bed time to perform Wake­Sleep algorithm. We recorded 116 subject all night about actigraphy signal for measure our algorithm with polysomnography(PSG), the average accuracy, sensitivity and specificity are 90.84%, 55.47% and 94.71%, respectively. The MAEs of SE, SOT, WASO and TST are 4.45%, 6.92 min, 17.68 min and 18.92 min, respectively. In order to avoid user is not wearing status and misjudgment sleep quality. We also develop he not wearing event detection. The accuracy of the not wearing event detection reached 91.74%. Our system could provide the identical result as traditional PSG above 90% agreement and is suitable for long­term sleep assessment.

[1]Da­Wei Chang, Sheng­Fu Liang, Chung­Ping Young, Fu­Zen Shaw, Alvin WY Su, You­De Liu, Yu­Lin Wang, Yi­Che Liu, Jing­Jhong Chen, and Chun­Yu Chen. A versa­ tile wireless portable monitoring system for brain–behavior approaches. IEEE Journal on Emerging and Selected Topics in Circuits and Systems, 1(4):440–450, 2011.
[2] Da­Wei Chang, You­De Liu, Chung­Ping Young, Jing­Jhong Chen, Ying­Huang Chen, Chun­Yu Chen, Yu­Cheng Hsu, Fu­Zen Shaw, and Sheng­Fu Liang. Design and imple­ mentation of a modularized polysomnography system. IEEE Transactions on Instru­ mentation and Measurement, 61(7):1933–1944, 2012.
[3] Jacob Cohen. A coefficient of agreement for nominal scales. Educational and psycho­ logical measurement, 20(1):37–46, 1960.
[4] Heidi Danker­Hopfe, Dieter Kunz, Georg Gruber, Gerhard Klösch, José L Lorenzo, Sari­Leena Himanen, Bob Kemp, Thomas Penzel, Joachim Röschke, Hans Dorn, et al. Interrater reliability between scorers from eight european sleep laboratories in subjects with different sleep disorders. Journal of sleep research, 13(1):63–69, 2004.
[5] Marco Filardi, Fabio Pizza, Monica Martoni, Stefano Vandi, Giuseppe Plazzi, and Vin­ cenzo Natale. Actigraphic assessment of sleep/wake behavior in central disorders of hypersomnolence. Sleep medicine, 16(1):126–130, 2015.
[6] SLEEPCOMPUTINGCOMMITTEEOFTHEJAPANESESOCIETYOFSLEEPRE­ SEARCH SOCIETY (JSSR):, Tadao Hori, Yoshio Sugita, Einosuke Koga, Shuichiro Shirakawa, Katuhiro Inoue, Sunao Uchida, Hiroo Kuwahara, Masako Kousaka, Toshi­ nori Kobayashi, et al. Proposed supplements and amendments to`a manual of standard­ ized terminology, techniques and scoring system for sleep stages of human subjects' , the rechtschaffen & kales (1968) standard. Psychiatry and clinical neurosciences, 55(3):305–310, 2001.
[7] Ron Kohavi and Foster Provost. Glossary of terms journal of machine learning. Mach. Learn, 1998.
[8] Daniel F Kripke, Elizabeth K Hahn, Alexandra P Grizas, Kep H Wadiak, Richard T Loving, J Steven Poceta, Farhad F Shadan, John W Cronin, and Lawrence E Kline. Wrist actigraphic scoring for sleep laboratory patients: algorithm development. Journal of sleep research, 19(4):612–619, 2010.
[9] Chih­En Kuo, Yi­Che Liu, Da­Wei Chang, Chung­Ping Young, Fu­Zen Shaw, and Sheng­Fu Liang. Development and evaluation of a wearable device for sleep quality assessment. IEEE Transactions on Biomedical Engineering, 64(7):1547–1557, 2016.
[10] Sheng­Fu Liang, Chin­En Kuo, Yi­Chieh Lee, Wen­Chieh Lin, Yen­Chen Liu, Peng­ Yu Chen, Fu­Yin Cherng, and Fu­Zen Shaw. Development of an eog­based automatic sleep­monitoring eye mask. IEEE Transactions on Instrumentation and Measurement, 64(11):2977–2985, 2015.
[11] Miguel Marino, Yi Li, Michael N Rueschman, John W Winkelman, JM Ellenbogen, Jo M Solet, Hilary Dulin, Lisa F Berkman, and Orfeu M Buxton. Measuring sleep: accuracy, sensitivity, and specificity of wrist actigraphy compared to polysomnography. Sleep, 36(11):1747–1755, 2013.
[12] Sharon Schutte­Rodin, Lauren Broch, Daniel Buysse, Cynthia Dorsey, and Michael Sateia. Clinical guideline for the evaluation and management of chronic insomnia in adults. Journal of clinical sleep medicine, 4(5):487–504, 2008.
[13] Kevin So, PAT Buckley, T Michael Adamson, and Rosemary SC Horne. Actigraphy correctly predicts sleep behavior in infants who are younger than six months, when compared with polysomnography. Pediatric research, 58(4):761–765, 2005.