巴金森病(Parkinson's Disease)為全球第二常見神經退化疾病，患者常伴隨步態障礙、起身困難與步態凍僵(Freezing of Gait, FoG)等運動症狀，且病況隨藥效改變。然現行統一巴金森病評估量表(Unified Parkinson's Disease Rating Scale, UPDRS)高度仰賴定期回診之短暫觀察，難以反映患者於居家日常生活中之真實運動狀態。
本研究以Wi-Fi通道狀態資訊(Channel State Information, CSI)感測技術為基礎，延伸既有少樣本卷積神經網路辨識架構與雙通道特徵訊號圖生成方式，提出一套以可量化行走片段為核心之居家步態監測方法，使模型於每類僅需少樣本訓練即可完成個人化動作辨識任務。為適應連續監測情境，本研究進一步導入滑動視窗逐半秒判斷機制與2.5秒持續時間篩選機制，以排除過渡性動作所造成之短暫誤判。
為驗證架構之可行性，本研究設計四項實驗：步行方向對可量化行走辨識之影響實驗、UPDRS相關起身行走測試(Timed Up and Go, TUG)模擬實驗、UPDRS相關 FoG 模擬實驗，以及長時間日常活動監測實驗。結果顯示，於步行方向與視距(Line-of-Sight, LoS)中垂線夾角介於90°至60°之適用範圍內，可量化行走之平均F1分數達91.62%，辨識擷取區段之步態量化速度平均絕對百分比誤差與人工標註基準之差距控制於1%以內。TUG模擬實驗中行走類別之跨情境整合F1分數達91.14%，FoG模擬實驗中行走與FoG類別之F1分數分別達95.36%與79.25%。長時間日常活動監測實驗中，低複雜度單人場域之可量化行走F1分數達98.61%，高複雜度多人共處場域亦維持於88.30%。
綜合而言，本研究初步證實所延伸之CSI辨識架構於長時居家監測環境中對可量化行走片段之擷取能力，可作為後續步態速度估測之前處理步驟，並為巴金森病居家步態功能之長期客觀追蹤提供可行之技術基礎。

Parkinson's disease (PD) often causes gait impairment, sit-to-stand difficulty, and freezing of gait (FoG), yet conventional clinical scales like the UPDRS capture only brief snapshots that may not reflect real-world motor function. This study extends a Wi-Fi Channel State Information (CSI)-based few-shot Convolutional Neural Network (CNN) framework for Human Activity Recognition (HAR) to identify quantifiable gait segments within continuous daily activities, enabling long-term home monitoring. The Few-shot Learning strategy allows personalized recognition from limited training samples per class. A 0.5-second sliding window performs classification, while a 2.5-second duration filter suppresses transient misclassifications from non-gait movements; detected segments are then used for CSI-based gait quantification and speed estimation. Four experiments evaluated the framework: walking-direction tests, UPDRS-related Timed Up and Go (TUG) simulations, FoG simulations, and long-term daily monitoring. At transmitter–receiver angles of 90°–60°, gait recognition achieved an average F1 score of 91.62%, with gait-speed MAPE differing by under 1% from manual annotation. The TUG simulation yielded an integrated F1 score of 91.14%, and the FoG simulation achieved 95.36% (gait) and 79.25% (FoG). Long-term monitoring reached F1 scores of 98.61% in a low-complexity environment and 88.30% in a high-complexity multi-person setting. These results confirm the feasibility of extracting gait segments for speed estimation from continuous indoor activity via CSI-based HAR, offering a technical basis for objective, non-intrusive, long-term gait monitoring in Parkinson's disease patients.

[1] D. Su et al., "Projections for prevalence of Parkinson's disease and its driving factors in 195 countries and territories to 2050: modelling study of global burden of disease study," BMJ, vol. 388, Mar. 2025, Art. no. e080952.
[2] L. V. Kalia and A. E. Lang, "Parkinson's disease," Lancet, vol. 386, no. 9996, pp. 896–912, Aug. 2015.
[3] D. M. Radhakrishnan and V. Goyal, "Parkinson's disease: a review," Neurol. India, vol. 66, no. 1, pp. 26–35, Mar. 2018.
[4] K. R. Chaudhuri and A. H. V. Schapira, "Non-motor symptoms of Parkinson's disease: dopaminergic pathophysiology and treatment," Lancet Neurol., vol. 8, no. 5, pp. 464–474, May 2009.
[5] J. Jankovic, "Motor fluctuations and dyskinesias in Parkinson's disease: clinical manifestations," Mov. Disord., vol. 20, no. S11, pp. S11–S16, May 2005.
[6] S. H. Fox et al., "The movement disorder society evidence‐based medicine review update: treatments for the motor symptoms of Parkinson's disease," Mov. Disord., vol. 26, no. S3, pp. S2–S41, Oct. 2011.
[7] E. Tolosa, G. Wenning, and W. Poewe, "The diagnosis of Parkinson's disease," Lancet Neurol., vol. 5, no. 1, pp. 75–86, Jan. 2006.
[8] S. Fahn, "Unified Parkinson's disease rating scale," in Recent Developments in Parkinson's Disease, S. Fahn, C. D. Marsden, D. B. Calne, and M. Goldstein, Eds. Florham Park, NJ, USA: Macmillan Healthcare Information, 1987, pp. 153–163.
[9] C. G. Goetz et al., "Movement disorder society‐sponsored revision of the unified Parkinson's disease rating scale (MDS‐UPDRS): scale presentation and clinimetric testing results," Mov. Disord., vol. 23, no. 15, pp. 2129–2170, Nov. 2008.
[10] D. Podsiadlo and S. Richardson, "The timed 'up & go': a test of basic functional mobility for frail elderly persons," J. Am. Geriatr. Soc., vol. 39, no. 2, pp. 142–148, Feb. 1991.
[11] S. Tan, Y. Ren, J. Yang, and Y. Chen, "Commodity WiFi sensing in ten years: status, challenges, and opportunities," IEEE Internet Things J., vol. 9, no. 18, pp. 17832–17854, Sep. 2022.
[12] Z. Yang et al., "From RSSI to CSI: Indoor localization via channel response," ACM Comput. Surv., vol. 46, no. 2, pp. 1–32, Nov. 2013.
[13] W. Zhu et al., "A computer vision-based system for stride length estimation using a mobile phone camera," in Proc. 18th Int. ACM SIGACCESS Conf. Comput. Accessibility, 2016, pp. 121–130.
[14] B. Jin et al., “Diagnosing Parkinson disease through facial expression recognition: Video analysis,” J. Med. Internet Res., vol. 22, no. 7, 2020, Art. no. e18697.
[15] A. Procházka, O. Vyšata, M. Vališ, O. Ťupa, and M. Schätz, “Bayesian classification and analysis of gait disorders using image and depth sensors of Microsoft Kinect,” Digit. Signal Process., vol. 47, pp. 169–177, Dec. 2015.
[16] A. Salarian, H. Russmann, C. Wider, P. R. Burkhard, F. J. G. Vingerhoets, and K. Aminian, "Quantification of tremor and bradykinesia in Parkinson's disease using a novel ambulatory monitoring system," IEEE Trans. Biomed. Eng., vol. 54, no. 2, pp. 313–322, Feb. 2007.
[17] S. Patel et al., "Monitoring motor fluctuations in patients with Parkinson's disease using wearable sensors," IEEE Trans. Inf. Technol. Biomed., vol. 13, no. 6, pp. 864–873, Nov. 2009.
[18] R. LeMoyne et al., “Implementation of an iPhone for characterizing Parkinson’s disease tremor through a wireless accelerometer application,” in Proc. Annu. Int. Conf. IEEE Eng. Med. Biol. Soc. (EMBC), 2010, pp. 4954–4958.
[19] K. Niazmand et al., “Quantitative evaluation of Parkinson’s disease using sensor based smart glove,” in Proc. 24th Int. Symp. Comput.-Based Med. Syst. (CBMS), 2011, pp. 1–8.
[20] B. Mariani et al., “On-shoe wearable sensors for gait and turning assessment of patients with Parkinson’s disease,” IEEE Trans. Biomed. Eng., vol. 60, no. 1, pp. 155–158, 2013.
[21] A. Zijlstra, M. Mancini, U. Lindemann, L. Chiari, and W. Zijlstra, "Sit-stand and stand-sit transitions in older adults and patients with Parkinson's disease: event detection based on motion sensors versus force plates," J. NeuroEng. Rehabil., vol. 9, no. 1, Nov. 2012, Art. no. 75.
[22] A. I. Meigal et al., "Novel parameters of surface EMG in patients with Parkinson's disease and healthy young and old controls," J. Electromyogr. Kinesiol., vol. 19, no. 3, pp. e206–e213, Jun. 2009.
[23] L. Bao and S. S. Intille, “Activity recognition from user-annotated acceleration data,” in Proc. Int. Conf. Pervasive Comput., 2004, pp. 1–17.
[24] M. Jaén-Vargas et al., “Effects of sliding window variation in the performance of acceleration-based human activity recognition using deep learning models,” PeerJ Comput. Sci., vol. 8, 2022, Art. no. e1052.
[25] S. Ren, K. He, R. Girshick, and J. Sun, “Faster R-CNN: Towards real-time object detection with region proposal networks,” in Adv. Neural Inf. Process. Syst., vol. 28, 2015.
[26] H. Xu, A. Das, and K. Saenko, “R-C3D: Region convolutional 3D network for temporal activity detection,” in Proc. IEEE Int. Conf. Comput. Vis. (ICCV), 2017, pp. 5783–5792.
[27] B. Tan, Q. Chen, K. Chetty, K. Woodbridge, W. Li, and R. Piechocki, “Exploiting WiFi channel state information for residential healthcare informatics,” IEEE Commun. Mag., vol. 56, no. 5, pp. 130–137, May 2018.
[28] Y. Ge et al., “Contactless WiFi sensing and monitoring for future healthcare—Emerging trends, challenges, and opportunities,” IEEE Rev. Biomed. Eng., vol. 16, pp. 171–191, 2023.
[29] J. Yang, H. Zou, H. Jiang, and L. Xie, “Device-free occupant activity sensing using WiFi-enabled IoT devices for smart homes,” IEEE Internet Things J., vol. 5, no. 5, pp. 3991–4002, Oct. 2018.
[30] H. Zou, Y. Zhou, J. Yang, H. Jiang, L. Xie, and C. J. Spanos, “DeepSense: Device-free human activity recognition via autoencoder long-term recurrent convolutional network,” in Proc. IEEE Int. Conf. Commun. (ICC), Kansas City, MO, USA, 2018, pp. 1–6.
[31] Y. Wang, K. Wu, and L. M. Ni, “WiFall: Device-free fall detection by wireless networks,” IEEE Trans. Mobile Comput., vol. 16, no. 2, pp. 581–594, Feb. 2017.
[32] Z. Yang, Y. Zhang, and Q. Zhang, “Rethinking fall detection with Wi-Fi,” IEEE Trans. Mobile Comput., vol. 22, no. 10, pp. 6126–6143, Oct. 2023.
[33] Z. Zheng, J. Zhu, S. Zhang, and Y. Xiao, “PreFall: Early detection of consecutive fall events with commercial Wi-Fi devices,” in Proc. IEEE 22nd Int. Conf. Mobile Ad-Hoc Smart Syst. (MASS), Chicago, IL, USA, 2025, pp. 456–464.
[34] J. Ding, Y. Wang, H. Si, S. Gao, and J. Xing, “Multimodal fusion-AdaBoost based activity recognition for smart home on WiFi platform,” IEEE Sensors J., vol. 22, no. 5, pp. 4661–4674, Mar. 2022.
[35] E. Lopez-Hernandez, F. F. Gonzalez-Navarro, B. L. Flores-Rios, and J. Caro-Gutierrez, “WIFIALR WiFi alerts software for human movements using machine learning algorithms,” in Proc. Mexican Int. Conf. Comput. Sci. (ENC), Morelia, Mexico, 2021, pp. 1–7.
[36] Z. Li, X. Lin, and Y. Lin, “Research on WiFi-CSI based behavior monitoring and analysis system in smart homes,” in Proc. Int. Conf. Electron. Devices Comput. Sci. (ICEDCS), Marseille, France, 2024, pp. 90–95.
[37] A. K. Sahoo, V. Kompally, and S. K. Udgata, “Wi-Fi sensing based real-time activity detection in smart home environment,” in Proc. IEEE Appl. Sens. Conf. (APSCON), Bengaluru, India, 2023, pp. 1–3.
[38] B. Dong et al., “Monitoring of atopic dermatitis using leaky coaxial cable,” Healthc. Technol. Lett., vol. 4, no. 6, pp. 244–248, 2017.
[39] J. Luo, K. Liu, Y. Wang, L. Li, and Z. Tian, “Respiratory monitoring using millimeter-wave base stations based on OFDM signals,” in Proc. IEEE Int. Instrum. Meas. Technol. Conf. (I2MTC), Glasgow, U.K., 2024, pp. 1–5.
[40] Y. Ma, Y. Zeng, and S. Sun, “A software defined radio based multi-function radar for IoT applications,” in Proc. 24th Asia-Pacific Conf. Commun. (APCC), Ningbo, China, 2018, pp. 239–244.
[41] G. Mattela, M. Tripathi, and C. Pal, “A novel approach in WiFi CSI-based fall detection,” SN Comput. Sci., vol. 3, no. 3, 2022, Art. no. 214.
[42] S. A. Rishi, G. Srinith, U. Dheeraj, and D. Das, “WiCare: Accidental fall detection for elderly care using passive Wi-Fi sensing,” in Proc. 9th Int. Conf. Signal Process. Commun. (ICSC), Noida, India, 2023, pp. 190–195.
[43] IEEE Standard for Information Technology—Telecommunications and Information Exchange Between Systems—Local and Metropolitan Area Networks—Specific Requirements—Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications, IEEE Std 802.11-2020, 2021.
[44] Y. Zeng, D. Wu, J. Xiong, E. Yi, R. Gao, and D. Zhang, “FarSense: Pushing the range limit of WiFi-based respiration sensing with CSI ratio of two antennas,” Proc. ACM Interact. Mobile Wearable Ubiquitous Technol., vol. 3, no. 3, Sep. 2019, Art. no. 121.
[45] D. Halperin, W. Hu, A. Sheth, and D. Wetherall, “Tool release: Gathering 802.11n traces with channel state information,” ACM SIGCOMM Comput. Commun. Rev., vol. 41, no. 1, p. 53, Jan. 2011.
[46] S.-Y. Chen and C.-L. Lin, "WiFi-based human activity recognition for continuous, whole-room monitoring of motor functions in Parkinson's disease," IEEE Open J. Antennas Propag., vol. 5, pp. 1127–1137, 2024.
[47] 黃冠瑄, "基於Wi-Fi訊號之少樣本學習與輕量化原型網路用於巴金森氏症相關動作識別," 國立成功大學機械工程學系學位論文, pp. 1–68, 2025.
[48] F. R. Hampel, “The influence curve and its role in robust estimation,” J. Amer. Stat. Assoc., vol. 69, no. 346, pp. 383–393, Jun. 1974.
[49] S. Butterworth, “On the theory of filter amplifiers,” Wireless Eng., vol. 7, no. 6, pp. 536–541, Oct. 1930.
[50] K. J. Zuiderveld, “Contrast limited adaptive histogram equalization,” in Graphics Gems IV, P. S. Heckbert, Ed. Boston, MA, USA: Academic Press, 1994, pp. 474–485.
[51] S. M. Pizer et al., "Adaptive histogram equalization and its variations," Comput. Vis. Graph. Image Process., vol. 39, no. 3, pp. 355–368, Sep. 1987.
[52] Y. Song, T. Wang, P. Cai, S. K. Mondal, and J. P. Sahoo, “A comprehensive survey of few-shot learning: Evolution, applications, challenges, and opportunities,” ACM Comput. Surv., vol. 55, no. 13s, pp. 1–40, Dec. 2023.
[53] J. Snell, K. Swersky, and R. Zemel, “Prototypical networks for few-shot learning,” in Adv. Neural Inf. Process. Syst., vol. 30, 2017, pp. 4077–4087.
[54] C.-L. Lin, W.-J. Chang, and G.-H. Tu, “Wi-Fi-based tracking of human walking for home health monitoring,” IEEE Internet Things J., vol. 9, no. 11, pp. 8935–8942, Jun. 2022.
[55] 翁子揚, "Wi-Fi步態感測—技術開發並結合IMU感測器以評估節奏聽覺刺激對帕金森病患者步態參數的影響," 國立成功大學機械工程學系學位論文, pp. 1–73, 2023.
[56] 江振愷, "結合都卜勒頻移校正之Wi-Fi通道狀態資訊穩健型步態量化方法：應用於巴金森病臨床步態評估," 國立成功大學機械工程學系學位論文, pp. 1–104, 2026.
[57] P. Fernández-González, A. Koutsou, A. Cuesta-Gómez, M. Carratalá-Tejada, J. C. Miangolarra-Page, and F. Molina-Rueda, “Reliability of Kinovea® software and agreement with a three-dimensional motion system for gait analysis in healthy subjects,” Sensors, vol. 20, no. 11, Jun. 2020, Art. no. 3154.
[58] A. Kharb et al., “A review of gait cycle and its parameters,” Int. J. Comput. Eng. Manage., vol. 13, no. 1, pp. 78–83, 2011.
[59] M. W. Whittle, Gait Analysis: An Introduction. Oxford, U.K.: Butterworth-Heinemann, 2014.