帕金森氏症診斷目前依賴神經科專科醫師的臨床評估，隨著全球疾病負擔增加，此種診斷方式造成顯著的成本障礙，且評估作業僅限於醫院環境中進行。穿戴式感測器已成為客觀評估帕金森氏症的前景工具，具有多種應用包括跌倒檢測、步態分析、凍結步態評估、震顫評估和動作遲緩評估等。值得注意的是，MARG（磁力、角速度、重力）感測器已達到良好性能表現，在步態動作檢測上的百分比誤差可低至2%。

儘管具有良好的性能指標，MARG感測器仍面臨噪聲和漂移等固有限制，若要在實際居家監測應用中可靠地測量帕金森氏症運動症狀，必須透過先進的信號處理技術來克服這些限制。其中一個關鍵的初始障礙是如何在進行有意義的臨床分析之前，準確識別並定位感測器數據流中的特定運動模式。

本論文提出一個針對MARG感測器角速度數據最佳化的處理管線，包含數據收集、預處理、特徵選擇和運動檢測等步驟，採用系統化方法於帕金森氏症監測中進行運動分類。本研究方法實施留一法交叉驗證框架，使用專門設計的隨機森林分類器來識別統一帕金森氏症評定量表（UPDRS）中的四種不同運動模式：手指敲擊、靜止震顫、手部動作和姿勢震顫。透過遞歸特徵消除法，我們辨識出一組涵蓋時域和頻域的判別特徵集。實驗進一步探討資料配置的選擇、隨機森林超參數調整、特徵選擇和運動檢測性能之影響。

本研究框架在12名受試者的測試中，使用集成模型達到了0.886的F1分數。這些令人鼓舞的結果促進了基於特徵方法在帕金森氏症運動檢測領域的進一步研究。本研究為更複雜的運動品質和症狀嚴重程度分析奠定基礎，有助於推動開發經濟實惠的居家帕金森氏症進展監測工具。

Parkinson's disease (PD) diagnosis currently relies on clinical evaluations by specialized neurologists, creating significant cost barriers and limiting assessment to hospital settings as the global disease burden increases. Wearable sensors have emerged as promising tools for objective PD assessment, with various applications such as fall detection, gait analysis, freezing of gait, tremor assessment, and bradykinesia assessment. Notably, MARG (Magnetic, Angular Rate, and Gravity) sensors have achieved good performance, including percentage errors as low as 2% for gait movement detection.

Despite their promising performance metrics, MARG sensors face inherent limitations from noise and drift that must be overcome through advanced signal processing techniques before they can reliably measure PD motor symptoms in practical home monitoring applications. A critical initial barrier is accurately identifying and localizing specific movement patterns within sensor streams before meaningful clinical analysis can proceed.

This paper presents a pipeline for data collection, preprocessing, feature selection, and movement detection optimized for angular rate data from MARG sensors, employing a systematic approach to movement classification in PD monitoring. Our methodology implements a leave-one-out cross-validation framework with specialized Random Forest classifiers for four distinct PD movement patterns from the Unified Parkinson's Disease Rating Scale (UPDRS): finger tapping, rest tremor, hand movements, and postural tremor. Through recursive feature elimination, we identified a set of discriminative features spanning both time and frequency domains. Experiments further explore the choice of data configuration, Random Forest hyperparameter tuning, feature selection, and movement detection performance.

Our framework achieved F1 scores of 0.886 using ensembled models across 12 subjects. These encouraging results motivate further study into feature-based approaches for PD movement detection. This research establishes a foundation for more sophisticated analyses of movement quality and symptom severity, advancing the development of affordable, at-home monitoring tools for Parkinson's disease progression.

Adams, J. L., Kangarloo, T., Tracey, B., O’Donnell, P., Volfson, D., Latzman, R. D., Zach, N., Alexander, R., Bergethon, P., Cosman, J., et al. (2023). Using a smartwatch and smartphone to assess early parkinson’s disease in the watch-pd study. npj Parkinson’s Disease, 9(1):64.

Antonini, A., Reichmann, H., Gentile, G., Garon, M., Tedesco, C., Frank, A., Falkenburger, B., Konitsiotis, S., Tsamis, K., Rigas, G., et al. (2023). Toward objective monitoring of parkinson’s disease motor symptoms using a wearable device: wearability and performance evaluation of pdmonitor®. Frontiers in neurology, 14:1080752.

Arami, A., Poulakakis-Daktylidis, A., Tai, Y. F., and Burdet, E. (2019). Prediction of gait freezing in parkinsonian patients: A binary classification augmented with time series prediction. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 27(9): 1909–1919.

Ashyani, A., Lin, C.-L., Roman, E., Yeh, T., Kuo, T., Tsai, W.-F., Lin, Y., Tu, R., Su, A., Wang, C.-C., Tan, C.-H., and Nordling, T. E. M. (2022). Digitization of updrs upper limb motor examinations towards automated quantification of symptoms of parkinson’s disease [unpublished manuscript]. Manuscript in preparation.

Bailo, G., Saibene, F. L., Bandini, V., Arcuri, P., Salvatore, A., Meloni, M., Castagna, A., Navarro, J., Lencioni, T., Ferrarin, M., et al. (2024). Characterization of walking in mild parkinson’s disease: Reliability, validity and discriminant ability of the six-minute walk test instrumented with a single inertial sensor. Sensors, 24(2):662.

Beach, T. G. and Adler, C. H. (2018). Importance of low diagnostic accuracy for early parkinson’s disease. Movement Disorders, 33(10):1551–1554.

Bloem, B. R., Okun, M. S., and Klein, C. (2021). Parkinson’s disease. The Lancet, 397(10291):2284–2303.

Bremm, R. P., Pavelka, L., Garcia, M. M., Mombaerts, L., Krüger, R., and Hertel, F. (2024). Sensor-based quantification of mds-updrs iii subitems in parkinson＇s disease using machine learning. Sensors, 24(7):2195.

Esling, P. and Agon, C. (2012). Time-series data mining. ACM Computing Surveys (CSUR), 45(1):1–34.

Cutler, A., Cutler, D. R., and Stevens, J. R. (2012). Random forests. In Ensemble machine learning, pages 157–175. Springer.

Fakhrazari, A. and Vakilzadian, H. (2017). A survey on time series data mining. In 2017 IEEE International Conference on Electro Information Technology (EIT), pages 476–481. IEEE.

Fridolfsson, J., Börjesson, M., Buck, C., Ekblom, Ö., Ekblom-Bak, E., Hunsberger, M., Lissner, L., and Arvidsson, D. (2019). Effects of frequency filtering on intensity and noise in accelerometer-based physical activity measurements. Sensors, 19(9):2186.

Fu, T.-c. (2011). A review on time series data mining. Engineering Applications of Artificial Intelligence, 24(1):164–181.

Garcia, M. (2020). Attitude and Heading Reference Systems in Python.

Goetz, C. G., Tilley, B. C., Shaftman, S. R., Stebbins, G. T., Fahn, S., Martinez-Martin, P., Poewe, W., Sampaio, C., Stern, M. B., Dodel, R., et al. (2008). Movement disorder society-sponsored revision of the unified parkinson’s disease rating scale (mds-updrs): scale presentation and clinimetric testing results. Movement disorders: official journal of the Movement Disorder Society, 23(15):2129–2170.

Hammoud, M., Shcherbak, A., Istrakova, O., Shindryaeva, N., Bril, E., Passerone, R., and Somov, A. (2025). Wrist-worn sensors and machine learning for parkinson＇s disease detection: Investigation of binary and multi-classification problem. IEEE Transactions on Instrumentation and Measurement.

Han, S., Williamson, B. D., and Fong, Y. (2021). Improving random forest predictions in small datasets from two-phase sampling designs. BMC medical informatics and decision making, 21(1):322.

Hur, S., Zhang, J., Kim, M.-H., and Chung, T.-M. (2025). Real-world parkinson’s hand tremor detection using ensemble learning techniques. IEEE Access.

Jeon, H., Lee, W., Park, H., Lee, H. J., Kim, S. K., Kim, H. B., Jeon, B., and Park, K. S. (2017). Automatic classification of tremor severity in parkinson＇s disease using a wearable device. Sensors, 17(9):2067.

Kleanthous, N., Hussain, A. J., Khan, W., and Liatsis, P. (2020). A new machine learning based approach to predict freezing of gait. Pattern Recognition Letters, 140:119–126.

Lim, S.-Y. and Tan, A. H. (2018). Historical perspective: the pros and cons of conventional outcome measures in parkinson’s disease. Parkinsonism & Related Disorders, 46:S47– S52.

Lima, F. T. and Souza, V. M. (2023). A large comparison of normalization methods on time series. Big Data Research, 34:100407.

Madgwick, S. O., Harrison, A. J., and Vaidyanathan, R. (2011). Estimation of imu and marg orientation using a gradient descent algorithm. In 2011 IEEE international conference on rehabilitation robotics, pages 1–7. IEEE.

Mazomenos, E. B., Biswas, D., Cranny, A., Rajan, A., Maharatna, K., Achner, J., Klemke, J., Jöbges, M., Ortmann, S., and Langendö rfer, P. (2015). Detecting elementary arm movements by tracking upper limb joint angles with marg sensors. IEEE journal of biomedical and health informatics, 20(4):1088–1099.

Orphanidou, N. K., Hussain, A., Keight, R., Lishoa, P., Hind, J., and Al-Askar, H. (2018). Predicting freezing of gait in parkinsons disease patients using machine learning. In 2018 IEEE Congress on Evolutionary Computation (CEC), pages 1–8. IEEE.

Ou, Z., Pan, J., Tang, S., Duan, D., Yu, D., Nong, H., and Wang, Z. (2021). Global trends in the incidence, prevalence, and years lived with disability of parkinson’s disease in 204 countries/territories from 1990 to 2019. Frontiers in public health, 9:776847.

Parati, M., Gallotta, M., Muletti, M., Pirola, A., Bellafà, A., De Maria, B., and Ferrante, S. (2022). Validation of pressure-sensing insoles in patients with parkinson＇s disease during overground walking in single and cognitive dual-task conditions. Sensors, 22(17):6392.

Patel, S., Lorincz, K., Hughes, R., Huggins, N., Growdon, J., Standaert, D., Akay, M., Dy, J., Welsh, M., and Bonato, P. (2009). Monitoring motor fluctuations in patients with parkinson’s disease using wearable sensors. IEEE transactions on information technology in biomedicine, 13(6):864–873.

Pierleoni, P., Belli, A., Bazgir, O., Maurizi, L., Paniccia, M., and Palma, L. (2019). A smart inertial system for 24h monitoring and classification of tremor and freezing of gait in parkinson＇s disease. IEEE Sensors Journal, 19(23):11612–11623.

Rastegari, E., Ali, H., and Marmelat, V. (2022). Detection of parkinson’s disease using wrist accelerometer data and passive monitoring. Sensors, 22(23):9122.

Rizek, P., Kumar, N., and Jog, M. S. (2016). An update on the diagnosis and treatment of parkinson disease. Cmaj, 188(16):1157–1165.

Safarpour, D., Dale, M. L., Shah, V. V., Talman, L., Carlson-Kuhta, P., Horak, F. B., and Mancini, M. (2022). Surrogates for rigidity and pigd mds-updrs subscores using wearable sensors. Gait & posture, 91:186–191.

Santoyo-Ramón, J. A., Casilari, E., and Cano-García, J. M. (2022). A study of the influence of the sensor sampling frequency on the performance of wearable fall detectors. Measurement, 193:110945.

Sharma, S. and Pandey, S. (2016). Approach to a tremor patient. Annals of Indian Academy of Neurology, 19(4):433.

Shawen, N., O’Brien, M. K., Venkatesan, S., Lonini, L., Simuni, T., Hamilton, J. L., Ghaffari, R., Rogers, J. A., and Jayaraman, A. (2020). Role of data measurement characteristics in the accurate detection of parkinson＇s disease symptoms using wearable sensors. Journal of neuroengineering and rehabilitation, 17:1–14.

Shin, J., Matsumoto, M., Maniruzzaman, M., Hasan, M. A. M., Hirooka, K., Hagihara, Y., Kotsuki, N., Inomata-Terada, S., Terao, Y., and Kobayashi, S. (2024). Classification of hand-movement disabilities in parkinson’s disease using a motion-capture device and machine learning. IEEE Access.

Shourijeh, M. S., Razavian, R. S., and McPhee, J. (2017). Estimation of maximum muscle contraction frequency in a finger tapping motion using forward musculoskeletal dynamic simulations. ASME J. Comput. Nonlinear Dyn.

Stamatakis, J., Ambroise, J., Crémers, J., Sharei, H., Delvaux, V., Macq, B., and Garraux, G. (2013). Finger tapping clinimetric score prediction in parkinson’s disease using low-cost accelerometers. Computational intelligence and neuroscience, 2013:1.

Tolosa, E., Garrido, A., Scholz, S. W., and Poewe, W. (2021). Challenges in the diagnosis of parkinson’s disease. The Lancet Neurology, 20(5):385–397.

Webber, M. and Rojas, R. F. (2021). Human activity recognition with accelerometer and gyroscope: A data fusion approach. IEEE Sensors Journal, 21(15):16979–16989.

WHO (2023). Parkinson’s disease.

Williams, H. J., Holton, M. D., Shepard, E. L., Largey, N., Norman, B., Ryan, P. G., Duriez, O., Scantlebury, M., Quintana, F., Magowan, E. A., et al. (2017). Identification of animal movement patterns using tri-axial magnetometry. Movement ecology, 5:1–14.

Zhao, H., Cao, J., Xie, J., Liao, W.-H., Lei, Y., Cao, H., Qu, Q., and Bowen, C. (2023). Wearable sensors and features for diagnosis of neurodegenerative diseases: A systematic review. Digital Health, 9:20552076231173569.