Comparison of Artificial Neural Networks with other Machine Learning Methods in Foot Movement Classification

Yıl 2023, , 153 - 171, 15.03.2023

Selin Aydın Fandaklı , Halil Okumuş

https://doi.org/10.31466/kfbd.1214950

Cited By: 1

Öz

Modern prostheses can be controlled by using gait analysis data from Inertial Measurement Units compared to traditional prostheses. This article aims to classify foot movements for the robotic ankle system in lower limb prostheses to recognize motion intent and adapt to abnormal walking conditions. The statistical features are extracted from IMU data from 11 volunteers aged 20-34 and then the features are classified using machine learning. In this study, the classification accuracies of Naïve Bayes Classifier, Linear Discriminant Analysis, K-Nearest Neighbour Classifier and Support Vector Machines and Artificial Neural Networks in classifying foot movements are examined separately for the raw data and the processed data such as Euler angles and quaternions which estimate with Madwick Filter. Gait analysis data were obtained by using the Inemo inertial module LSM9DS1 work on an NRF52 including 9 DOF, triaxial gyroscope, triaxial accelerometer, and triaxial magnetometer in the Biomechanics Laboratory of the Department of Mechanical Engineering, Middle East Technical University from eleven subjects and achieved an highest classification accuracy rate of 90.9% on test data, 97.3% for training data.

Anahtar Kelimeler

K-Nearest Neighbor, Support Vector Machines, Accelerometers, Gyroscope

Ayak Hareketleri Sınıflandırmasında Yapay Sinir Ağlarının diğer Makine Öğrenme Yöntemleri ile Karşılaştırılması

Öz

Modern protezler, geleneksel protezlere kıyasla Atalet Ölçüm Birimlerinden (IMU'lar) alınan yürüyüş analizi verileri kullanılarak kontrol edilebilir. Bu makale, hareket niyetini tanımak ve anormal yürüme koşullarına uyum sağlamak için alt ekstremite protezlerinde robotik ayak bileği sistemi için ayak hareketlerini sınıflandırmayı amaçlamaktadır. 20-34 yaşları arasındaki 11 gönüllüden toplanan IMU verilerinden istatistiksel özellikler çıkarılmış ve daha sonra öznitelikler makine öğrenmesi kullanılarak sınıflandırılmıştır. Bu çalışmada, Naive Bayes Sınıflandırıcısı, Doğrusal Ayırım Analizi, K-En Yakın Komşu Sınıflandırıcısı ve Destek Vektör Makineleri ve Yapay Sinir Ağları ayak hareketlerini sınıflandırmadaki sınıflandırma doğrulukları ham veriler için ve Madwick Filtresi ile tahmin edilen Euler açıları ve kuaterniyonlar gibi işlenmiş veriler için ayrı ayrı incelenmiştir. Yürüyüş analizi verileri, Orta Doğu Teknik Üniversitesi Makine Mühendisliği Bölümü Biyomekanik Laboratuvarı'nda on bir denekten üç eksenli jiroskop, üç eksenli ivmeölçer ve üç eksenli manyetometre içeren 9 serbestlik dereceli bir NRF52 üzerinde Inemo atalet modülü LSM9DS1 çalışması kullanılarak alındı ve test verilerinde % 90.90, eğitim verilerinde %97.3 en yüksek sınıflandırma doğruluk oranı elde edildi.

Anahtar Kelimeler

K-En yakın komşu, destek vektör makineleri, ivmeölçer, jiroskop

Kaynakça

• Hoile, R. (1996). Amputation: Surgical Practice and Patient Management. British Medical Journal, 312(7036), 984-985.
• Kerr, M., Barron, E., Chadwick, P., Evans, T., Kong, W. M., Rayman, G., Jeffcoate, W. J. (2019). The cost of diabetic foot ulcers and amputations to the National Health Service in England. Diabetic Medicine, 36(8): 995-1002.
• Parkka, J., Ermes, M., Korpipaa, P., Mantyjarvi, J., Peltola, J., Korhonen, I. (2006). Activity classification using realistic data from wearable sensors. In: IEEE Trans. Inf. Technol. Biomed., 119-128.
• Seel, T., Raisch, J., Schauer, T. (2014). IMU-based joint angle measurement for gait analysis. Sensors, 14(4), 6891-6909.
• Haoyu, L, Derrode, S., Pieczynski, W. (2019). An adaptive and on-line IMU-based locomotion activity classification method using a triplet Markov model, Neurocomputing, 362, pp. 94-105.
• San-Segundo, R., Montero, J. M., Barra-Chicote, R., Fernandez, F., Pardo, J. M. (2016). Feature extraction from smartphones inertial signals for human activity segmentation, Signal Processing; 120, 359-372.
• Fullerton, E., Heller, B., Munoz-Organero, M. (2017). Recognizing human activity in free-living using multiple body-worn accelerometers, IEEE Sensors Journal, 17(16), 5290-5297.
• Gao, F., Liu, G., Liang, F., Liao, W-H. (2020). IMU-based locomotion mode identification for transtibial prostheses, orthoses, and exoskeletons. IEEE Transactions on Neural Systems and Rehabilitation Engineering; 28(6), 1334-1343.
• Wen, J., Wang, Z., (2016). Sensor-based adaptive activity recognition with dynamically available sensors. Neurocomputing, 218, 307-317.
• Wu, D., Wang, Z., Chen, Y., Zhao, H. (2016). Mixed kernel based weighted extreme learning machine for inertial sensor based human activity recognition with imbalanced dataset. Neurocomputing, 190, 35-49.
• Khera, P., Kumar, N., Ahuja, P. (2020). Machine Learning based Electromyography Signal Classification Feature Selection for Foot Movements. Journal of Scientific& Industrial Research, vol.79, p. 1011-1016.
• Chaobankoh, N., Jumphoo, T., Uthansakul, M. Phapatanaburi, K., Sindhupakorn, B., Rooppakhun, S., Uthansakul, P.(2022). Lower Limb Motion Based Ankle Foot Movement Classification Using 2D-CNN. Computers, Materials &Continua, vol.73, p. 1269-1282.
• Bernal-Polo, P., Martínez-Barberá, H. (2019). Kalman filtering for attitude estimation with quaternions and concepts from manifold theory. Sensors, 19(1), 149.
• Bhargaviand, P., Jyothi, S. (2009). Applying naive bayes data mining technique for classification of agricultural land soil. International Journal of Computer Science and Network Security, 9(8), 117-122.
• Gohari, M., Eydi, A. M. (2020). Modelling of shaft unbalance: Modelling and multi discs rotors using K-Nearest Neighbor and Decision Tree Algorithms. Measurement, 151.
• Maragliulo, S., Lopes, P. F. A., Osório L. B., De Almeida A. T., Tavakoli, M. (2019). Foot gesture through dual channel wearable EMG system. IEEE Sens. J. 19, p. 10187–10197.
• Hooda, N., Kumar, N. (2021). Optimal Channel-set and Feature-set Assessment for Foot Movement Based EMG Pattern Recognition. Applied Artifical Intelligence; 35:15, p. 1685-1707.
• Friedman, N., Geigerand, D., Goldszmidt, M. (1997). Bayesian network classifiers. Machine Learning, 29(2-3), 131-161.
• Craig, J. J. (2005). Introduction to Robotics Mechanics and Control. Pearson Education International, 42-50.
• Welling, M. (2005). Fisher linear discriminant analysis. Department of Computer Science, University of Toronto, 3, 1–4.
• McLachlan, G., (2004). Discriminant analysis and statistical pattern recognition. 132-134, John Wiley & Sons.
• Fisher, R. A. (1936). Annals of Eugenics. Management Science, 7(2), 179-188.
• Han, J., Kamber, M., Pei, J. (2011). Data Mining Concept and Techniques. 3rd Edition, Morgan Kaufmann Publishers, USA, 394-397.
• Theodoridis, S., Koutroumbas S. (2006). Pattern Recognition. 3rd Edition, Elsevier, USA, pp.119-133.
• Pollard, N. S., Hodgins, J. K., Riley, M. J., Atkeson, C. G. (2002). Adapting human motion for the control of humanoid robot, In: IEEE International Conference on Robotics and Automation. pp. 1390-1397.
• Miura, K., Morisawa, M., Kanehiro, F., Kajita, S., Kaneko, K., Yokoi, K. (2011). Human-like walking with toe supporting for humanoids. IEEE/RSJ International Conference on Intelligent Robots and Systems; pp. 4428-4435.
• Hacker, S., Kalkbrenner, C., Algorri, M., Blechschmidt-Trapp, R. (2014). Gait Analysis with IMU - Gaining new orientation information of the lower leg. In: Proceedings of the International Conference on Biomedical Electronics and Devices, pp. 127-133.
• Aydin Fandakli, S., Okumus, H. I., Erdem, A. F. (2018). Design and Dynamic Modelling of an Ankle-Foot Prosthesis for Transfemoral Amputees, International Conference on Engineering Technologies (ICENTE'18), pp.409-413.
• Aydin Fandakli, S., Okumus, H. I., Aydemir, O. (2017). A fast and highly accurate EMG signal classification approach for multifunctional prosthetic fingers control. Telecommunications and Signal Processing (TSP), 40th International Conference on, Barcelona, Spain: 2017. pp. 395-398.
• URL-1: https://cdn.sparkfun.com/datasheets/Sensors/IMU/SparkFun-LSM9DS1-Breakout-v10.pdf. (Data Accessed.04.05.2021).
• Madwick, S. O. (2010). An efficient orientation filter for inertial and inertial/magnetic sensor arrays. Citado, 9-19.
• Tharwat, A., Gaber, T., Ibrahim, A., Hassanien, A. E. (2017). Linear discriminant analysis: A detailed tutorial. AI Communications; 30(2), 169-190.