臺灣博碩士論文加值系統

醫院與照護中心的受傷通報事件中，經常是跌倒所引起的，而這些跌倒事件裡，大多是患者或老人剛起床時，嘗試獨自上下床，他們錯估自己的身體狀況，進行離床而發生跌倒，再加上照護或醫護人員的人力不足，無法及時發現並且上前提供協助，最終導致憾事發生。
本研究提出一個離床行為辨識與預測系統，能及時提醒醫護或照護人員協助患者或老人進行下床動作，防範跌倒事件的發生。本系統包含四個荷重元感測器(load cell sensor)，透過轉接座將其安裝在床角的下方，偵測患者或老人在床上動作所產生的壓力變化，並藉此檢測患者或老人當下的行為狀態與離床意圖。本系統使用k-means++演算法將離床動作分群；隱馬爾可夫模型(hidden Markov model, HMM)來建立動作與動作之間的關聯，以執行離床行為辨識與預警。
實驗結果顯示，平躺、坐起、側坐與離床動作的辨識精確率(precision)分別達 87.76%、90.22%、90.93% 與 99.44%。此外，我們在坐起至側坐之間發現了「扭腰」的關鍵動作，相較於側坐在床邊準備下床時，提出離床警示的系統，本研究在扭腰時即提出預警可使醫護或照護人員有更多時間上前幫忙，能夠減少跌倒的發生。離床預警的precision與召回率(recall)分別為98.8%與95.58%，本系統具備預警的功效。

Fall is one of the most common injury incident reported in hospitals and nursing home. Most of falls occur when the patients or elderly try to get off the bed after they wake up. Because of the shortage of staff and resources, nurses cannot always provide the assistance in time. As a result, it is common for patients and elderly to fall off the bed due to being alone in these places.
In this research, we present a bed exit behavior recognition and prediction system. The system comprises four load cell sensors installed under the legs of bed with stands. Load cell itself can detect the pressure changes generated by the patient or the elderly in the bed. The system uses k-means++ algorithm to cluster bed exit actions and the hidden Markov model (HMM) to establish the association between different actions to perform bed exit behavior recognition and prediction.
The experimental results showed the precision of four different posture: lying on bed, sitting on bed, sitting on edge of bed, and leaving the bed with 87%, 90%, 90%, and 99%, respectively. Furthermore, we use load cell sensor to distinguish a new action: turning waist on bed. Compares to the posture of sitting at the edge of bed, this new action can achieve earlier recognition of patient leaving the bed than using the action of sitting at the edge of bed. As a result, turning waist on bed, can help the system to send earlier alarm to nurses. Nurses can prevent patients fall. The recall and precision of bed exit prediction was 98.59% and 94.24% respectively. These results show that the proposed system is capable of providing early alarm for medical personnel to prevent patients from falls.

摘　要 i
ABSTRACT ii
誌 謝 iv
目 錄 v
表目錄 viii
圖目錄 ix
第一章 緒論 1
1.1 前言 1
1.2 研究背景與動機 1
1.3 論文架構 3
第二章 相關研究 4
2.1 離床相關研究 4
2.1.1 離床偵測 4
2.1.2 離床預警 4
2.2 離床相關感測器研究 5
2.2.1 壓力感測墊 5
2.2.2 穿戴式裝置 6
2.2.3 攝影機 7
2.2.4 荷重元感測器 7
2.3 離床動作分析 7
2.4 機器學習方法 9
2.4.1 監督式學習 9
2.4.2 非監督式學習 10
2.4.3 半監督式學習 11
第三章 系統架構與研究方法 12
3.1 系統架構 12
3.2 資料處理流程 13
3.3 離床動作 14
3.4 床體壓力特徵計算 16
3.5 離床行為辨識方法 19
3.5.1 k-means++ 19
3.5.2 隱馬爾可夫模型 20
第四章 實驗結果與分析 22
4.1 實驗環境與設置 22
4.1.1 實驗平台 22
4.1.2 實驗環境 25
4.1.3 量測方式 28
4.1.4 感測特性 29
4.1.5 離床動作辨識演算法比較 29
4.2 系統評量指標 31
4.2.1 評量指標 31
4.2.2 離床預測與離床行為辨識 32
4.3 實驗結果分析與比較 32
4.3.1 決策樹的離床行為辨識結果 32
4.3.2 k-means++搭配最近鄰居法的離床行為辨識結果 35
4.3.3 HMM搭配k-means++的離床行為辨識結果 39
4.3.4 離床預測結果 46
第五章 結論與未來研究方向 50
5.1 結論 50
5.2 未來研究方向 50
參考文獻 52
附錄 57
A. 荷重元感測器周邊PCB電路設計 57
B. 發表之論文 58

[1] 衛生福利部台灣病人安全資訊網，台灣病人安全通報系統2016年TPR年 報，取自: http://www.patientsafety.mohw.gov.tw/Content/Downloads/List01.aspx?SiteID=1&MmmID=621273303702500244
[2] H. C. Hanger, M. C. Ball, and L. A. Wood, “An analysis of falls in the hospital: can we do without bedrails?,” Journal of the American Geriatrics Society, 1999.
[3] M. Vassallo, R. A. Amersey, J. C. Sharma, and S. C. Allen, “Falls on integrated medical wards,” Gerontology, 2000.
[4] D. Oliver, A. Hopper, and P. Seed, “Do hospital fall prevention programs work? A systematic review,” Journal America Geriatrics Society, 2000.
[5] K. Hill, M. Vu, and W. Walsh, “Falls in the acute hospital settingimpact on resource utilisation,” Australian Health Review, 2007.
[6] S. Heinrich, K. Rapp, U. Rissmann, C. Becker, H.H. König, “Cost of falls in old age: A systematic review,” Osteoporosis International 2010, Volume 21, Issue 6, pp. 891–902.
[7] 國家發展委員會，中華民國人口推估（105至150年），取自: https://www.ndc.gov.tw/Content_List.aspx?n=84223C65B6F94D72
[8] 台灣失智症協會，台灣失智症人口推估，取自: http://www.tada2002.org.tw/About/IsntDementia
[9] 社團法人台灣長期照護專業協會（無日期），長期照護歷史軌跡，民106年1 月6日，取自：
http://www.ltcpa.org.tw/main/index.php?func=introduce&nID=8
[10] 全國法規資料庫（民104年6月3日），制定長期照顧服務法，民106年1月6日，取自：
http://law.moj.gov.tw/News/news_detail.aspx?id=114581
[11] 中華民國行政院（民104年6月4日），行政院會通過「長期照顧保險法」草案，民106年1月6日，取自： http://www.ey.gov.tw/news_Content2.aspx?n=F8BAEBE9491FC830&s=86739C171F4C3CCD
[12] 衛生福利部（民105年8月3日），長照十年計畫2.0，民106年1月6日，取自：
http://www.mohw.gov.tw/MOHW_Upload/doc/105年8月3日溝通說明會簡報_0055618003.pdf
[13] 失能老人及身心障礙者補助使用居家服務計畫 - 內政部，取自： www.moi.gov.tw/files/news_file/101身心障礙者補助使用居家服務計畫.doc
[14] 李宗育、陸鳳屏、詹鼎正，「老年人跌倒之危險因子、評估、及預防」，內科學誌，第二十五卷，第三期，2014，137-142頁，取自: http://www.tsim.org.tw/journal/jour25-3/02.PDF
[15] D. Oliver, “Prevention of falls in hospital inpatients: agendas for research and practice,” Age Ageing, 2004.
[16] E. Capezuti, B. L. Brush, S. Lane, H. U. Rabinowitz, and M. Secic, “Bed-exit alarm effectiveness,” Archives of Gerontology and Geriatrics, 2009.
[17] J. Hilbe, E. Schulc, B. Linder, and C. Them, “Development and alarm threshold evaluation of a side rail integrated sensor technology for the prevention of falls,” International Journal of Medical Informatics, 2010.
[18] K. Imaizumi, Y. Iwakami, and K. Yamashita, “Availability of Monitoring System for Supporting Healthcare of Elderly People,” Japanese Journal of Applied IT Healthcare, 2010.
[19] M. Bruyneel, W. Libert, and V. Ninane, “Detection of bed-exit events using a new wireless bed monitoring assistance,” International Journal of Medical Informatics, 2011.
[20] http://www.ncbi.nlm.nih.gov/pubmed/15532858, “Bed exit alarms,” Health Devices, 2004.
[21] R. I. Shorr, A. M. Chandler, L. C. Mion, T. M. Waters, M. Liu, M. J. Daniels, L. A. Kessler, and S. T. Miller, “Effects of an intervention to increase bed alarm use to prevent falls in hospitalized patients: a cluster randomized trial,” Annals of Internal Medicine, 2012.
[22] R. Tideiksaar, C. F. Feiner, and J. Maby, “Falls prevention: the efficacy of a bed alarm system in an acute-care setting,” The Mount Sinai Journal of Medicine New York, 1993.
[23] A. Härmä, W. ten Kate, and J. Espina, “Bed Exit Prediction Based on Movement and Posture Data,” Proceedings of International Conference on Biomedical and Health Informatics (BHI), 2014, pp. 165-168
[24] T. L. Bozano, G. Camus, J. C. Basilio, J. M. Basilio, and J. L. Viard, U. S. Patent No. 2011,006,893,2A1 (14 November, 2016)
[25] R.L.S. Torres, Q. Shi, A.V. D. Hengel, and D. C. Ranasinghe, “A hierarchical model for recognizing alarming states in a batteryless sensor alarm intervention for preventing falls in older people,” Pervasive and Mobile Computing, vol 40, pp. 1-16, 2017.
[26] W. O. de Morais, J. Lundström, and N. Wickström, “Active In-Database Processing to Support Ambient Assisted Living Systems,” IEEE Sensors Journal, vol. 14, no. 8, pp.14765-14785, 2014.
[27] W. K. Lee, H. Yoon, C. Han, K. M. Joo, and K. S. Park, “Physiological Signal Monitoring Bed for Infants Based on Load-Cell Sensors,” IEEE Sensors Journal, vol. 16, no. 3, pp. 409, 2016.
[28] A. M. Adami, M.Pavel, T. L. Hayes, A. G. Adami, and C. Singer, “A Method for Classification of Movements in Bed,” Proceedings of 2011 Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Boston, Massachusetts, 2011, pp. 7881-7884.
[29] T. X. Chen, R. S. Hsiao, C. H. Kao, H. P. Lin, S. S. Jeng, and D. B. Lin, “Vision-Assisted Human Motion Analysis for Bed Exit Prediction Model Construction,” Proceedings of the 2017 International Conference on Innovation, Communication and Engineering (ICICE2017), Kunming, Yunnan Province, P. R. China, 2017.
[30] H. Madokoro, N. Shimoi, and K. Sato, “Unrestrained Multiple-Sensor System for Bed-Leaving Detection and Prediction,” Nursing and Health, vol. 3, no. 3, pp. 58-68, 2015.
[31] H. Madokoro , N. Shimoi and K. Sato, “Unrestrained multi-sensor systems for real-time prediction of bed-leaving behavior patterns,” Proceedings of the SICE Annual Conference (SICE), Sapporo, Japan, 2014, pp. 1946-1953.
[32] H. Madokoroh, K. Kakuta, R. Fujisawa, N. Shimoi, K. Sato and L. Xu, “Bed-Leaving Behavior Detection and Recognition Based on Time-Series Learning Using Elman-Type Counter Propagation Networks,” Proceedings of　the 2014 14th International Conference on Control, Automation and Systems (ICCAS 2014), Seoul, South Korea, 2014, pp. 540-545.
[33] K. Bressler, R. E. Redfern, and M. Brown, “Elimination of position-change alarms in an Alzheimer’s and dementia long-term care facility,” American journal of Alzheimers disease and other dementias, vol. 26, no. 8, pp. 599-605, 2011.
[34] D. C. Ranasinghe, R. L. S. Torres, and A. Wickramasinghe, “Automated activity recognition and monitoring of elderly using wireless sensors: Research challenges,” in 5th IEEE International Workshop on Advances in Sensors and Interfaces (IWASI), 2013, pp. 224-227.
[35] K. H. Wolf, K. Hetzer, H. M. Schwabedissen, B. Wiese, and M. Marschollek, “Development and pilot study of a bed-exit alarm based on a body-worn accelerometer,” Zeitschrift für Gerontologie und Geriatrie, vol. 46, no. 8, pp. 727-733, 2013.
[36] A. Wickramasinghe, D. C. Ranasinghe, C. Fumeaux, K. D. Hill, and R. Visvanathan, “Sequence Learning with Passive RFID Sensors for Real Time Bed-egress Recognition in Older People,” IEEE Journal of Biomedical and Health Informatics, vol. 21, no. 4, pp. 917-929, 2017.
[37] R. L. S. Torres, D. C. Ranasinghe, Q. Shi, and A. P. Sample, “Sensor enabled wearable RFID technology for mitigating the risk of falls near beds,” Proceedings of 2013 IEEE International Conference on RFID (RFID), Penang, Malaysia, 2013, pp. 191-198.
[38] D. C. Ranasinghe, R. S. Torres, K. Hill, and R. Visvanathan, “Low cost and batteryless sensor-enabled radio frequency identification tag based approaches to identify patient bed entry and exit posture transitions,” Gait & Posture, vol. 39, no. 1, pp. 118-123, 2014.
[39] R. L. S. Torres, R. Visvanathan, S. Hoskins, A. Hengel, and D. C. Ranasinghe, “Effectiveness of a batteryless and wireless wearable sensor system for identifying bed and chair exits in healthy older people,” IEEE Sensors Journal, vol. 16, no. 4, pp. 546-562, 2016.
[40] G. Demiris, B. K. Hensel, M. Skubic, and M. Rantz, “Senior residents’ perceived need of and preferences for “smart home” sensor technologies,” International journal of technology assessment in health care, vol. 24, no. 1, pp. 120-124, 2008.
[41] M. Gövercin, Y. Költzsch, M. Meis, S. Wegel, M. Gietzelt, J. Spehr, S. Winkelbach, M. Marschollek, and E. Steinhagen-Thiessen, “Defining the user requirements for wearable and optical fall prediction and fall detection devices for home use,” Informatics for health and social care, vol. 35, no. 3-4, pp. 177-187, 2010.
[42] M.Yu, A. Rhuma, S. M. Naqvi, L. Wang, and J. Chamber, “A posture recognition-based fall detection system for monitoring an elderly person in a smart home environment,” IEEE Trans on information technology in biomedicine, vol. 16, no. 6, pp.1274-1286, 2012.
[43] Y. S. Delahoz, and M. A. Labrador, “Survey on fall detection and fall prevention using wearable and external sensors,” IEEE Sensors Journal, vol. 14, no. 10, pp. 19806-19842, 2014.
[44] A. M. Adami, A. G. Adami, G. Schwarz, Z. T. Beattie, and T. L. Hayes, “A Subject State Detection Approach to Determine Rest-Activity Patterns Using Load Cells,” Proceedings of 2010 Annual International Conference of the IEEE Engineering in Medicine and Biology, Buenos Aires, Argentina, 2010, pp. 204-207.
[45] Z. T. Beattie , C. C. Hagen, and T. L. Hayes, “Classification of lying position using load cells under the bed,” Proceedings of 2011 Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Boston, MA, USA, 2011, pp. 474-477.
[46] D. Austin, Z. T. Beattie , T. Riley, A. M. Adami, C. C. Hagen, and T. L. Hayes, “Unobtrusive classification of sleep and wakefulness using load cells under the bed,” Proceedings of 2012 Annual International Conference of the IEEE Engineering in Medicine and Biology Society, San Diego, CA, USA, 2012, pp.5254-5257.
[47] K. Hirasawa, N. Matsumura, M. Motegi, T Yamada, S. Muto, N. Kanamaru, K. Shimokura, M. Abe, Y. Morita, and K. Kasai, “Analyzing Rising Patterns of　Patients to Prevent Bed-related Falls. -First Report-,”日本医療マネジメント学会雑誌, vol. 11, no. 1, pp. 31-36, 2010.
[48] M. Motegi, N. Matsumura, T. Yamada, S. Muto, N. Kanamaru, K. Shimokura, M. Abe, Y. Morita, and K. Kasai,” Analyzing Rising Patterns of　Patients to Prevent Bed-related Falls. -Second Report-,”日本医療マネジメント学会雑誌, vol. 12, no. 1, pp. 25-29, 2011.
[49] H. Madokoro, N. Shimoi, and K. Sato, “Unrestrained Multiple-Sensor System for Bed-Leaving Detection and Prediction,” Nursing and Health, vol. 3, no. 3, 2015, pp. 56-68.
[50] H. Madokoro, N. Shimoi, and K. Sato, “Unrestrained multi-sensor systems for real-time prediction of bed-leaving behavior patterns,” Proceedings of the SICE Annual Conference (SICE), Sapporo, Japan, 2014, pp. 1946-1953.
[51] H. Madokoro, K. Kakuta, R. Fujisawa, N. Shimoi, K. Sato, and L. Xu, ” Bed-leaving behavior detection and recognition based on time-series learning using Elman-Type Counter Propagation Networks Sign In or Purchase,” Proceedings of the 14th International Conference on Control, Automation and Systems (ICCAS 2014),” Seoul, South Korea, 2014, pp. 540-545.
[52] P. Kulkarni, Reinforcement and Systemic Machine Learning for Decision Making: N.J., Hoboken, Canada, 2012.
[53] J. MacQueen, “Some methods for classification and analysis of multivariate observations,” Proceedings of the fifth Berkeley symposium on mathematical statistics and probability, vol. 1, 1967, pp.281-296.
[54] A. David, and S. Vassilvitskii, “k-means++: The Advantages of Careful Seeding,” SODA ‘07: Proceedings of the Eighteenth Annual ACM-SIAM Symposium on Discrete Algorithms, 2007, pp. 1027–1035.
[55] L. R. Rabiner and B. H. Juang, “An introduction to hidden Markov models,” IEEE ASSP Magazine, vol. 3, pp. 4-16, January, 1986.
[56] HT Sensor, “TAS606 Button type compression load cell,” [Online]. Available: http://www.htc-sensor.com/products/151.html
[57] Texas Instruments, “INA12x Precision, Low-Power Instrumentation Amplifiers,” [Online]. Available: http://www.ti.com/lit/ds/symlink/ina129.pdf
[58] T. Sato, S. Nukaya, H. Tanaka, and T. Hiroyasu, “Patient behavior estimating system on the bed using a combination waveform with the piezoelectric ceramic sensors,”人工知能学会全国大会論文集, vol. 29, pp. 1-3, 2015.
[59] D. M. W. Powers, “Evaluation: from precision,recall and f-measure to roc, informedness, markedness and correlation,” Journal of Machine Learning Technologies, vol. 2, pp.37-63, 2011.
[60] I. H. Witten, E. Frank, and M.A. Hall, Data Mining - Practical Machine Learning Tools and Techniques, 3rd ed. Burlington, MA: Morgan Kaufmann, 2011.
[61] V. Labatut and H. Cherifi, "Accuracy Measures for the Comparison of Classifiers,” ArXiv Preprint arXiv: 1207.3790, 2012.