現今台灣已是高齡社會，隨著高齡化趨勢發展，依國發會推估將於114年進入超高齡社會。隨著老年人口增加，偏癱發生人口在台灣也逐漸增加，其中又以中風最為大宗。而中風後產生的後遺症當中，約有88%的患者在中風後產生出不同輕重程度的偏癱症狀，而長期的復健療程可以來改善患者病況。藉由雙腳之間的步態差異可以來評估復健的成效。在後疫情時代，生活作息的轉變下，居家照護及自我復健的觀念日漸興起，但大多數的生理反饋分析系統只能在特定少數的生物力學實驗室或是相關醫療機構才能夠進行，且系統器材昂貴，所占空間大、所需時間長的原因並不適用於居家應用，因此開發一套簡便的復健評估系統遂成為臨床的重要議題。
本研究使用兩個慣性測量模組分別固定於雙腳之腳背上，並讀取出慣性元件中的加速度和角速度來做為步態量化的依據。由於加速度計與陀螺儀本身的限制，本研究透過感測器融合的方式，讓兩個模組進行互相校正，並整合研究中所使用的演算法來盡可能正確地顯示出步態數值。為了讓患者容易檢視及了解受測成果，本研究於Android系統平台上進行應用程式的開發，將慣性測量模組的原始資料進行預處理後透過藍牙4.0的技術，以藍牙傳輸的方式傳輸到智慧型手機，患者即可在手機的應用程式上即時觀看行走的步態圖像量化狀況，也可以在行走後經由計算後得知當前的差異分析。
本研究中設計進行了一套實驗，受測者會被要求分別模擬在雙腳正常及單腳異常情境下，採取自然速度的方式直線步行，並藉由智慧型手機觀看即時量化的特徵圖形，並在實驗結束後獲得相關的角度差異，藉此來評估受測者的復健之成效。由實驗結果可得知，在模擬雙腳正常的步態特徵中，仰俯角度差異約為10度以內。而模擬單腳異常的步態特徵，其角度差異則會大於10度，代表任一腳在某些步態特徵中會顯著不同於另一腳。由結果可得知，此系統可以反饋生理資訊的變化，辨別出正常及異常雙腳間的步態資訊。

Taiwan is now an aging society. With the development of the aging trend, it is estimated by the National Development and Reform Commission that it will enter a super-aged society in 114 years. With the increase of the elderly population, the incidence of hemiplegia is also increasing in Taiwan, among which stroke is the most common. Among the sequelae after stroke, about 88% of patients have hemiplegia symptoms of different severity after stroke. Long-term rehabilitation treatment can improve the patient's condition. The effectiveness of rehabilitation can be assessed by the difference in gait between the feet. In the post-epidemic era, with changes in daily life, the concept of home care and self-rehabilitation is rising day by day, but most physiological feedback analysis systems can only be performed in a few biomechanical laboratories or related medical institutions, and The system equipment is expensive, occupies a large space, and takes a long time, which is not suitable for home application. Therefore, the development of a simple rehabilitation evaluation system has become an important clinical issue.
Two inertial measurement modules were fixed on the insteps of the feet respectively in this study, and the acceleration and angular velocity in the inertial elements were read out as the basis for gait quantification. Due to the limitations of accelerometers and gyroscopes, this study uses sensor fusion to allow the two modules to calibrate each other, and integrate the algorithms used in the study to display gait values as accurately as possible. In order to make it easy for patients to view and understand the measured results, this study developed an application program on the Android system platform, preprocessed the original data of the inertial measurement module, and then transmitted it to the smartphone through Bluetooth 4.0 technology. The patient can view the gait image quantification status of walking on the mobile phone app in real-time and can also know the current difference analysis after walking through the calculation.
A set of experiments was designed and carried out in this study. The subjects were asked to simulate walking in a straight line at a natural speed under normal conditions of two feet and abnormal conditions of one foot respectively and watched the real-time quantitative characteristic patterns through a smartphone. And the relevant difference values were obtained after the end of the experiment, to evaluate the rehabilitation effect of the subjects. From the experimental results, it can be known that in the simulated normal gait characteristics of both feet, the pitch angle difference is within 10 degrees. For the gait characteristics that simulate the abnormality of one foot, the angle difference will be greater than 10 degrees, which means that one foot will be significantly different from the other foot in some gait characteristics. It can be seen from the results that the system can feed back changes in physiological information and identify the gait information between normal and abnormal feet.

[1] 葉伯壽, "台灣腦中風概況與急性腦梗塞的治療發展," 中國統計學報, vol. 55, no. 2, pp. 63-66, 2017.
[2] Neofect. What are the Different Types of Stroke? Available: https://www.neofect.com/us/blog/what-are-the-different-types-of-stroke
[3] C.L. Chen, F.T. Tang, H.C. Chen, C.Y. Chung, and M.K. Wong, "Brain lesion size and location: effects on motor recovery and functional outcome in stroke patients," Archives of physical medicine and rehabilitation, vol. 81, no. 4, pp. 447-452, 2000.
[4] National Stroke Association. (2014). Post-stroke conditions. Available: http://www.stroke.org/we-can-help/survivors/stroke-recovery/post-stroke-conditions
[5] B. Balaban and F. Tok, "Gait disturbances in patients with stroke," PM&R, vol. 6, no. 7, pp. 635-642, 2014.
[6] A. H Khandoker, Assessment Methods of Post-stroke Gait: A Scoping Review of Technology-Driven Approaches to Gait Characterization and Analysis, 2021.
[7] S. J. Cuccurullo, Physical medicine and rehabilitation board review. Demos Medical Publishing, 2014.
[8] A. R. Anwary, Hongnian Yu, Michael Vassallo, "Hemiparetic gait," An Automatic Gait Feature Extraction Method for Identifying Gait Asymmetry Using Wearable Sensors, 2018.
[9] W. Liu, C. Zhang, H. Ma, S. Li, Learning Efficient Spatial-Temporal Gait Features with Deep Learning for Human Identification, 2018.
[10] K. K. Patterson, W. H. Gage, D. Brooks, S. E. Black, and W. E. McIlroy, "Evaluation of gait symmetry after stroke: a comparison of current methods and recommendations for standardization," Gait & posture, vol. 31, no. 2, pp. 241-246, 2010.
[11] S. M. Woolley, "Characteristics of gait in hemiplegia," Topics in stroke rehabilitation, vol. 7, pp. 1-18, 2001.
[12] E. J. Roth, C. Merbitz, K. Mroczek, S. A. Dugan, and W. W. Suh, "Hemiplegic gait: Relationships between walking speed and other temporal parameters1," American journal of physical medicine & rehabilitation, vol. 76, pp. 128-133, 1997.
[13] D. Levine, J. Richards, and M. W. Whittle, Whittle's Gait Analysis-E-Book. Elsevier Health Sciences, 2012.
[14] C. K. Balasubramanian, R. R. Neptune, and S. A. Kautz, "Variability in spatiotemporal step characteristics and its relationship to walking performance post-stroke," Gait & posture, vol. 29, no. 3, pp. 408-414, 2009.
[15] W. H. Organization, International Classification of Functioning, Disability and Health: ICF. World Health Organization, 2001.
[16] B.W. Hwang, S. Kim, and S.W. Lee, "2D and 3D full-body gesture database for analyzing daily human gestures," Advances in Intelligent Computing, pp. 611-620, 2005.
[17] NOKOV. Making High-Performance Motion Capture more Cost-Effective, High Frequency and High Accuracy Motion Capture. Available: https://www.nokov.com/en/motion-capture-movement-analysis.html
[18] Y. Shih, C.S. Ho, and T.Y. Shiang, "Measuring kinematic changes of the foot using a gyro sensor during intense running," Journal of sports sciences, vol. 32, no. 6, pp. 550-556, 2014.
[19] W. W. Lee et al., "A smartphone-centric system for the range of motion assessment in stroke patients," IEEE journal of biomedical and health informatics, vol. 18, no. 6, pp. 1839-1847, 2014.
[20] K. Oyake et al., "Validity of gait asymmetry estimation by using an accelerometer in individuals with hemiparetic stroke," Journal of physical therapy science, vol. 29, no. 2, pp. 307-311, 2017.
[21] R. Caldas, M. Mundt, W. Potthast, F. B. de Lima Neto, and B. Markert, "A systematic review of gait analysis methods based on inertial sensors and adaptive algorithms," Gait & posture, vol. 57, pp. 204-210, 2017.
[22] W. Vaughn. (2016). Better Mattresses and Mattress Selection through Pressure Mapping. Available: http://www.beds.org/blog/better-mattresses-and-mattress-selection-through-pressure-mapping/
[23] Y. Zhu, G. Zhang, W. Xu, and J. Zhao, " Flexible Force-sensing System for Wearable Exoskeleton Using Liquid Pressure Detectio " 2017.
[24] 鄧正隆, 慣性技術. Hyweb Technology Co. Ltd., 2011.
[25] W. Tao, T. Liu, R. Zheng, and H. Feng, "Gait analysis using wearable sensors," Sensors, vol. 12, no. 2, pp. 2255-2283, 2012.
[26] Wiki. (2007). Yaw, Pitch and Roll in an aircraft. Available: https://en.wikipedia.org/wiki/Aircraft_principal_axes
[27] X. Wang and Y. Wang, "Gait analysis of children with spastic hemiplegic cerebral palsy," vol. 7, no. 20, p. 1578, 2012.
[28] E. Molteni, E. Beretta, D. Altomonte, F. Formica, and S. Strazzer, "Combined robotic-aided gait training and 3D gait analysis provide objective treatment and assessment of gait in children and adolescents with Acquired Hemiplegia," in Engineering in Medicine and Biology Society (EMBC), 2015 37th Annual International Conference of the IEEE, pp. 4566-4569, 2015.
[29] R. A. Clark, S. Vernon, B. F. Mentiplay, K. J. Miller, J. L. McGinley, Y. H. Pua, et al., "Instrumenting gait assessment using the Kinect in people living with stroke: reliability and association with balance tests," Journal of neuroengineering and rehabilitation, vol. 12, p. 15, 2015.
[30] R. Lemoyne and T. Mastroianni, "Use of smartphones and portable media devices for quantifying human movement characteristics of gait, tendon reflex response, and Parkinson’s disease hand tremor," Mobile Health Technologies: Methods and Protocols, pp. 335-358, 2015.
[31] A. M. Sabatini, C. Martelloni, S. Scapellato, and F. Cavallo, "Assessment of walking features from foot inertial sensing," vol. 52, no. 3, pp. 486-494, 2005.