臺灣博碩士論文加值系統

輕度認知功能障礙(MCI)被認為在正常衰老和失智症之間的狀態，並非所有MCI都會變為失智症，但失智症是常見的退化性神經疾病。MCI患者的認知功能退化會影響行走的表現，所以進行一次神經心理測驗，使用可攜式步態分析系統獲得步態與跳躍特徵參數，進行直線行走、計時起步(TUG)和原地向上跳三種測驗，分隔半年後再次測驗步態，追蹤MCI患者的步態變化。本論文採用橫斷性(Cross-sectional)與縱貫性研究(Longitudinal Research)，橫斷性研究就第一次的步態測驗結果針對不同種類的MCI患者之步態建立分類模型。獲得第二次的步態測驗結果後，與第一次的差異進行縱貫性研究，透過神經心理測驗結果預測未來步態衰退與否。兩種研究方法皆利用機器學習來分析，若有遺漏值採用回歸型支持向量機(SVR)預測數據，再正規化(Normalization)處理數據，之後利用支持向量機(SVM)與主成分分析(PCA)來建立分類模型，最後透過接收者操作特徵曲線(ROC)之曲線下面積(AUC)驗證預測結果。橫斷性研究的結果對於PD-MCI與非PD-MCI患者的預測準確率達91.67%，能夠證實在臨床上醫師對MCI患者診斷。縱貫性研究的結果則顯示MCI患者演變成失智症的關鍵步態參數可能為走路的速度，期望能輔助醫生診斷MCI患者，使得步態中的速度成為診斷MCI的生物標記。

Mild cognitive impairment (MCI) refers to a transitional condition between normal aging and early dementia, but not all MCI become dementia. Dementia is a common neurodegenerative disorder. Degradation of the cognitive function of MCI affects the performance of walking. In this study, a neuropsychological test was executed and there were two stages of gait test. First, a portable gait analysis system was used to obtain gait features and did the neuropsychological test. After half a year, tested gait again. Cross-section and longitudinal research were performed in this paper. The cross-sectional research used the first stage gait result to establish a classification model for different types of MCI. The predictive results were 91.67% accurate for PD-MCI and non-PD-MCI patients, confirming the clinical diagnosis of MCI patients. Besides, the longitudinal research built a classification model for the difference between the first and second stage results that can be used to predict future gait declines. The result shown the most critical gait parameter for MCI patients becoming dementia may be the speed of walking. It is expected to assist the doctors in diagnosing MCI patients, making the speed in gait a biomarker for diagnosing MCI.

摘　要 i
ABSTRACT ii
誌　謝 iii
目　錄 iv
表目錄 vii
圖目錄 ix
第一章 緒論 1
1.1 研究動機 1
1.2 文獻回顧 3
1.3 研究方法 5
1.4 論文架構 7
第二章 實驗設計 8
2.1 參與研究人員條件設定 8
2.2 實驗方法設定 9
2.2.1 神經心理測驗 9
2.2.2 行走測驗 10
2.2.3 計時起步測驗(Time Up and Go, TUG) 11
2.2.4 原地向上跳測驗 11
第三章 實驗設備與系統架構 12
3.1 BTS G-Walk 12
3.1.1 G-Sensor 12
3.1.2 G-Studio 14
3.2 系統架構 19
第四章 機器學習 20
4.1 回歸型支持向量機(Support Vector Regression, SVR) 20
4.1.1 均方根誤差(Root Mean Square Error, RMSE) 25
4.1.2 平均絕對誤差(Mean Absolute Error, MAE) 25
4.2 正規化(Normalization) 25
4.3 主成份分析(Principal Component Analysis, PCA) 26
4.4 支持向量機(Support Vector Machine, SVM) 27
4.4.1 接收者操作特徵曲線(Receiver Operating Characteristic, ROC)的曲線下面積(Area Under the Curve, AUC) 28
第五章 結果與討論 30
5.1 橫斷性研究(Cross-Sectional Research) 30
5.1.1 人數統計與比較 31
5.1.2 遺漏值處理與預測 35
5.1.3 預測分類模型 42
5.2 縱貫性研究(Longitudinal Research) 45
5.2.1 神經心理測驗結果 46
5.2.2 遺漏值處理與預測 48
5.2.3 行走測驗之速度 52
5.2.4 TUG測驗之時間 54
5.2.5 原地向上跳測驗之高度 56
第六章 結論與未來展望 58
6.1 結論 58
6.2 未來展望 59
參考文獻 60
附　錄 67
附錄A 馬偕紀念醫院人體研究倫理審查委員會同意臨床試驗證明書 67
附錄B 人體研究對象(或受試者)同意書 69
附錄C 神經心理檢查 75

[1]P. H. Chen, S. J. Cheng, H. C. Lin, C. Y. Lee, and C. H. Chou, “Risk Factors for the Progression of Mild Cognitive Impairment in Different Types of Neurodegenerative Disorders,” Behavioural Neurology, vol. 2018, no. 1, 2018, pp. 1-8.
[2]R. C. Petersen, “Mild cognitive impairment as a diagnostic entity,” Journal of Internal Medicine, vol. 256, no. 3, 2004, pp. 183-194.
[3]圖片來源：台灣失智症協會，http://www.tada2002.org.tw/About/IsntDementia。
[4]O. V. Forlenza, B. S. Diniz, W. F. Gattaz, “Diagnosis and biomarkers of predementia in Alzheimer's disease,” BMC Medcine, vol. 8, no. 1, 2010, pp. 89.
[5]R. Morris, S. Lord, J. Bunce, D. Burn, L. Rochester, “Gait and cognition: Mapping the global and discrete relationships in ageing and neurodegenerative disease,” Neurosci Biobehav Rev, vol. 64, 2016, pp. 326–345.
[6]M. Montero-Odasso, J. Verghese, O. Beauchet, J. M. Hausdorff, “Gait and cognition: A complementary approach to understanding brain function and the risk of falling,” J. Am Geriatr Soc, vol. 60, no. 11, 2012, pp. 2127–2136.
[7]D. Rodríguez-Martín, A. Samà, C. Pérez-López, A. Català, J. M. M. Arostegui, J. Cabestany, À. Bayés, S. Alcaine, B. Mestre, A.Prats, M. C. Crespo, T. J. Counihan, P. Browne, L. R. Quinlan, G. ÓLaighin, D. Sweeney, H. Lewy, J. Azuri, G. Vainstein, R. Annicchiarico, A. Costa, A. Rodríguez-Molinero, “Home detection of freezing of gait using support vector machines through a single waist-worn triaxial accelerometer,” PLoS One, vol. 12, no. 2, 2017, pp. e0171764.
[8]H. Gonçalves, R. Moreira, A. Rodrigues, C. Santos, “Finding parameters around the abdomen for a vibrotactile system: healthy and patients with Parkinson’s disease,” J Med Syst, vol. 42, 2018, pp. 232.
[9]Y. J. Lai, P.H. Chen, J. Shaw, “Development of Gait Analysis System for Parkinson’s Disease Patients in Circular Walking,” Journal of the Chinese Society of Mechanical Engineers, vol.36, no.1, 2015, pp.1-9.
[10]P.H. Chen, D.J. Liou, K.C. Liou, C.L. Liang, S.J. Cheng, J. Shaw, “Walking Turns in Parkinson’s Disease Patients with Freezing of Gait: The Short-term Effects of Different Cueing Strategies,” International Journal of Gerontology, vol. 10, no. 2, 2016, pp. 71-75.
[11]C. Curtze,J. G. Nutt, P. Carlson-Kuhta, M. Mancini, F.B. Horak, “Levodopa Is a Double-Edged Sword for Balance and Gait in People With Parkinson's Disease,” Mov. Disord., vol. 30, no. 10, 2015, pp. 1361-1370.
[12]M. Nergui, C. Murai, Y. Koike et al., “Probabilistic Information Structure of Human Walking,” J. Med. Syst., vol. 35, no.5, 2011, pp. 835-844.
[13]V. J. Verlinden, J. N. van der Geest, Y. Y. Hoogendam, A. Hofman, M. M. Breteler, M. A. Ikram, “Gait patterns in a community-dwelling population aged 50 years and older,” Gait Posture, vol. 37, no. 4, 2013, pp. 500-505.
[14]U. K. Pun, H. Gu, Z. Dong, N. Sertac Artan, “Classification and visualization tool for gait analysis of Parkinson's disease,” International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), USA, 16-20 Aug. 2016.
[15]A. Vienne, R. P. Barrois, S. Buffat, D. Ricard, P. P. Vidal, “Inertial Sensors to Assess Gait Quality in Patients with Neurological Disorders: A Systematic Review of Technical and Analytical Challenges,” Front. Psychol., vol. 8, 2017, pp. 817.
[16]R. Morris, S. Lord, R. A. Lawson, S. Coleman, B. Galna, G. W. Duncan, T. K. Khoo, A. J. Yarnall, D. J. Burn, L. Rochester, “Gait Rather Than Cognition Predicts Decline in Specific Cognitive Domains in Early Parkinson's Disease,” J. Gerontol. A Biol. Sci. Med. Sci., vol. 72, no. 12, 2017, pp.1656-1662.
[17]D. Aarsland, K. Andersen, J. P. Larsen, A. Lolk, P. Kragh-Sorensen, "Prevalence and characteristics of dementia in Parkinson disease: an 8-year prospective study," Arch Neurol, vol. 60, no. 3, 2003, pp.387-392.
[18]K. Pigott, J. Rick, S. X. Xie, H. Hurtig, A. Chen-Plotkin, J. E. Duda, J. F. Morley, L. M. Chahine, N. Dahodwala, R. S. Akhtar, A. Siderowf, J. Q. Trojanowski, D. Weintraub, "Longitudinal study of normal cognition in Parkinson disease," Neurology, vol. 85, no. 15, 2015, pp. 1276-1282.
[19]I. Litvan, J. G. Goldman, A. I. Troster, B. A. Schmand, D. Weintraub, R. C. Petersen, B. Mollenhauer, C. H. Adler, K. Marder, C. H. Williams-Gray, D. Aarsland, J. Kulisevsky, M. C. Rodriguez-Oroz, D. J. Burn, R. A. Barker, M. Emre, "Diagnostic criteria for mild cognitive impairment in Parkinson's disease: Movement Disorder Society Task Force guidelines," Mov. Disord., vol. 27, no. 3, 2012, pp. 349-356.
[20]G. Santangelo, C. Vitale, M. Picillo, M. Moccia, S. Cuoco, K. Longo, D. Pezzella, A. di Grazia, R. Erro, M. T. Pellecchia, M. Amboni, L. Trojano and P. Barone, "Mild Cognitive Impairment in newly diagnosed Parkinson's disease: A longitudinal prospective study," Parkinsonism Relat. Disord., vol. 21, no. 10, 2015, pp. 1219-1226.
[21]K. F. Pedersen, J. P. Larsen, O. B. Tysnes and G. Alves, "Natural course of mild cognitive impairment in Parkinson disease: A 5-year population-based study," Neurology, vol. 88, no. 8, 2017, pp. 767-774.
[22]B. Bratić, V. Kurbalija, M. Ivanović, I. Oder, Z. Bosnić, “Machine Learning for Predicting Cognitive Diseases: Methods, Data Sources and Risk Factors,” J. Med. Syst., vol. 42, no. 12, 2018, pp. 243.
[23]J. Howcroft, E. D. Lemaire, J. Kofman, “Wearable-Sensor-Based Classification Models of Faller Status in Older Adults,” PLoS One, vol.11, no. 4, 2016, pp. e0153240.
[24]C. Tunvirachaisakul, T. Supasitthumrong, S. Tangwongchai, S. Hemrunroj, P. Chuchuen, I. Tawankanjanachot, Y. Likitchareon, K. Phanthumchinda, S. Sriswasdi, M. Maes, “Characteristics of Mild Cognitive Impairment Using the Thai Version of the Consortium to Establish a Registry for Alzheimer's Disease Tests: A Multivariate and Machine Learning Study,” Dement. Geriatr. Cogn. Disord., vol. 45, no. 1-2, 2018, pp.38-48.
[25]A. J. C. C. Lins, M. T. C. Muniz, A. N. M. Garcia, A.V. Gomes, R. M. Cabral, C.J.A. Bastos-Filho, “Using artificial neural networks to select the parameters for the prognostic of mild cognitive impairment and dementia in elderly individuals,” Comput. Methods Programs Biomed., vol. 152, 2017, pp. 93-104.
[26]R. C. Petersen, “Mild Cognitive Impairment,” Continuum (Minneap. Minn.), vol. 22, no. 2 Dementia, 2016, pp. 404-418.
[27]I. Litvan, J. G. Goldman, A. I. Tröster, B. A. Schmand, D. Weintraub, R. C. Petersen, B. Mollenhauer, C. H. Adler, K. Marder, C. H. Williams-Gray, D. Aarsland, J. Kulisevsky, M. C. Rodriguez-Oroz, D. J. Burn, R. A. Barker, M. Emre, “Diagnostic Criteria for Mild Cognitive Impairment in Parkinson’s Disease: Movement Disorder Society Task Force Guidelines,” Mov. Disord., vol. 27, no. 3, 2012, pp.349-356.
[28]Y. K. Chakravarthy, M. S. Sultana, A. Srinath, G. R. S. Naga Kumar, “A Mathematical Model to Determine the Torque for a Prosthetic Leg-Lagrangian Equation,” International Journal of Pure and Applied Mathematics, vol. 116, no. 5, 2017, pp.51-56.
[29]M. R. Nobile, P. Perego, L. Piccinini, E. D. Mani, A. Rossi, M. O. Bellina, M. Molteni, “Further evidence of complex motor dysfunction in drug naive children with autism using automatic motion analysis of gait,” Autism : the international journal of research and practice, vol. 15, no. 3, 2011, pp. 263-83 .
[30]圖片來源：BTS Bioengineering官方網站，https://www.btsbioengineering.com/products/g-walk-inertial-motion-system/。
[31]A. J. Smola, B. Schölkopf, “A tutorial on support vector regression,” Statistics and Computing, vol.14, no. 3, 2004, pp. 199-222.
[32]Y. Chen, P. Xu, Y. Chu, W. Li, Y. Wu, L. Ni, Y. Bao, K. Wang, “Short-term electrical load forecasting using the Support Vector Regression (SVR) model to calculate the demand response baseline for office buildings,” Applied Energy, vol. 195, 2017, pp. 659-670.
[33]H. Nourali, M. Osanloo, “Mining capital cost estimation using Support Vector Regression (SVR),” Resources Policy, Available online 22 November 2018.
[34]高安、洪瑞鍾、黃志鵬，2014，「支援向量迴歸建立血壓預測模型改善血壓量測的不準確性」，The 31st Workshop on Combinatorial Mathematics and Computation Theory，台北，第321頁。
[35]黃子桓，「Support Vector Regression」，取自：http://www.cmlab.csie.ntu.edu.tw/~cyy/learning/tutorials/SVR.pdf。
[36]余尚武、劉憶瑩、李凌崴，「應用帄滑支撐向量迴歸與灰預測建構全球 ETF 投資組合策略之研究」，中華管理評論國際學報，第十六卷，第三期，2013。
[37]董呈煌、李春長、陳俊麟、吳韻玲，「SVR與OLS在住宅價格預測正確率的比較」，住宅學報，第二十五卷，第二期，2016，第31-51頁。
[38]C. W. Hsu, C. C. Chang, C. J. Lin, “A Practical Guide to Support Vector Classification,” National Taiwan University, Taipei, 2003.
[39]S. Raschka, “About Feature Scaling and Normalization–and the effect of standardization for machine learning algorithms,” 2014, 取自：http://sebastianraschka.com/Articles/2014_about_feature_scaling.html#about-min-max-scaling。
[40]A. A. Al-Kheraif, M. Hashem, M. S. S. Al Esawy, “Developing Charcot–Marie–Tooth Disease Recognition System Using Bacterial Foraging Optimization Algorithm Based Spiking Neural Network,” Journal of Medical Systems, vol. 42, 2018, pp. 192.
[41]B. Akdemir, B. Oran, S. Gunes, S. Karaaslan, “Prediction of Aortic Diameter Values in Healthy Turkish Infants, Children, and Adolescents by Using Artificial Neural Network,” Journal of Medical Systems, vol. 33, 2009, pp. 379.
[42]F. Song, Z. Guo, D. Mei, “Feature Selection Using Principal Component Analysis,” IEEE International Conference on System Science, Engineering Design and Manufacturing Informatization, China, 12-14 Nov. 2010.
[43]A. Afzali, F. B. Mofrad, M. Pouladian, “Inter-Patient Modelling of 2D Lung Variations from Chest X-Ray Imaging via Fourier Descriptors,” Journal of Medical Systems, vol. 42, 2018, pp.233.
[44]S. Luo, T. Chen, L. Jian, “ Using Principal Component Analysis and Least Squares Support Vector Machine to Predict the Silicon Content in Blast Furnace System,” iJOE, vol. 14, no. 4, 2018, pp.149-162.
[45]Y. Bai, Z. Sun,B. Zeng, J. Long, L. Li, J. Valente de Oliveira, C. Li, “A comparison of dimension reduction techniques for support vector machine modeling of multi-parameter manufacturing quality prediction,” Journal of Intelligent Manufacturing, vol. 30, no. 5, 2019, pp. 2245-2256.
[46]林宗勳，「Support Vector Machines簡介」，2010，取自：http://www.cmlab.csie.ntu.edu.tw/~cyy/learning/tutorials/SVM2.pdf。
[47]J. Wu, J. Wang, “PCA-Based SVM for Automatic Recognition of Gait Patterns,” Journal of applied biomechanics, vol. 24, no, 1, 2008, pp. 83-87.
[48]A. Rakotomamonjy, “Support Vector Machines and Area Under ROC curve,” 2004, https://pdfs.semanticscholar.org/ed0f/645d69f74fc9773652f85c28c23c10695ad2.pdf.
[49]P. Flach, J. Hern´andez-Orallo, C. Ferri, “A Coherent Interpretation of AUC as a Measure of Aggregated Classification Performance,” Proceedings of the 28th International Conference on Machine Learning (ICML-11), USA ,2011, pp. 657-664.
[50]李棟洲、武俊傑，「老年步態障礙」，台灣老年醫學暨老年學雜誌，第十卷，第一期，2015，第1-15頁。
[51]R. W. Bohannon, “Comfortable and maximum walking speed of adults aged 20-79 years: Reference values and determinants,” Age and Ageing, vol. 26, no. 1, 1997, pp. 15-19.