Xiaotian Wang
ABSTRACT
Vertical Ground Reaction Force (VGRF) signal obtained from foot-worn sensors, also known as plantar data, provides a highly informative and detailed representation of an individual's gait features. Existing methods, such as CNNs, LSTMs and Transformers, have revealed the efficiency of deep learning in Parkinson's Disease (PD) diagnosis using VGRF signal. However, the intrinsic topologic graph and pressure transmission characteristics of plantar data are overlooked in those approaches, which are essential features for gait analysis. In this paper, we propose to construct a plantar directed topologic graph to fully exploit the plantar topology in gait circles. It can facilitate the expression of gait information by representing sensors as nodes and pressure transmissions as directional edges. Accordingly, an Adaptive Spatio-Temporal Directed Graph Neural Network (AST-DGNN) is proposed to extract the connection features of the plantar directed topologic graph. Each AST-DGNN Unit includes an Adaptive Directed Graph Network (ADGN) block and a Temporal Convolutional Network (TCN) block. In order to capture both local and global spatial relationships among sensor nodes and pressure transmission edges, the ADGN block performs message passing on the plantar directed topologic graph in an adaptive manner. To capture the temporal features of sensor nodes and pressure transmission edges, the TCN block defines a temporal feature extraction process for each node and edge in the graph. Moreover, the data augmentation is introduced for plantar data to improve the generalization ability of the AST-DGNN. Experimental results on Ga, Ju, and Si datasets demonstrate that the proposed method outperforms the existing methods under both cross-dataset validation and mixed-data cross-validation. Especially in cross-dataset validation, there is an average improvement of 2.13%, 7.73%, and 12.27% in accuracy, F1 score, and G-mean, respectively.
- Enas Abdulhay, N. Arunkumar, Kumaravelu Narasimhan, Elamaran Vellaiappan, and V. Venkatraman. 2018. Gait and tremor investigation using machine learning techniques for the diagnosis of Parkinson disease. Future Generation Computer Systems, Vol. 83 (2018), 366--373.Google Scholar
Digital Library
- Abdullah S. Alharthi, Alexander J. Casson, and Krikor B. Ozanyan. 2021. Gait Spatiotemporal Signal Analysis for Parkinson's Disease Detection and Severity Rating. IEEE Sensors Journal, Vol. 21, 2 (2021), 1838--1848.Google ScholarCross Ref
- Rami Alkhatib. 2016. Human locomotion analysis, classification and modeling of normal and pathological vertical ground reaction force signals in elderly. Theses. Université de Lyon.Google Scholar
- Tunç Aşuroçlu, Koray Açıcı, Çağatay Berke Erdaş, Münire Kılınç Toprak, Hamit Erdem, and Hasan Oğul. 2018. Parkinson's disease monitoring from gait analysis via foot-worn sensors. Biocybernetics and Biomedical Engineering, Vol. 38, 3 (2018), 760--772.Google ScholarCross Ref
- Peter W. Battaglia, Jessica B. Hamrick, Victor Bapst, and et al. 2018. Relational inductive biases, deep learning, and graph networks. arxiv: 1806.01261 [cs.LG]Google Scholar
- Lonneke ML de Lau and Monique MB Breteler. 2006. Epidemiology of Parkinson's disease. The Lancet Neurology, Vol. 5, 6 (2006), 525--535.Google ScholarCross Ref
- Haodong Duan, Jiaqi Wang, Kai Chen, and Dahua Lin. 2022. Pyskl: Towards good practices for skeleton action recognition. In Proceedings of the 30th ACM International Conference on Multimedia. 7351--7354.Google ScholarDigital Library
- Manish Dubey, AK Wadhwani, and S Wadhwani. 2013. Gait based vertical ground reaction force analysis for Parkinson's disease diagnosis using self organizing map. Int. J. Adv. Biol. Biomed. Res, Vol. 1, 6 (2013), 624--636.Google Scholar
- Balaji E, Brindha D, Vinodh Kumar Elumalai, and Umesh K. 2021. Data-driven gait analysis for diagnosis and severity rating of Parkinson's disease. Medical Engineering & Physics, Vol. 91 (2021), 54--64.Google ScholarCross Ref
- Imanne El Maachi, Guillaume-Alexandre Bilodeau, and Wassim Bouachir. 2020. Deep 1D-Convnet for accurate Parkinson disease detection and severity prediction from gait. Expert Systems with Applications, Vol. 143 (2020), 113075.Google ScholarDigital Library
- Wenqi Fan, Yao Ma, Qing Li, Yuan He, Eric Zhao, Jiliang Tang, and Dawei Yin. 2019. Graph Neural Networks for Social Recommendation. In The World Wide Web Conference (WWW '19). Association for Computing Machinery, 417--426.Google ScholarDigital Library
- Evan N. Feinberg, Debnil Sur, Zhenqin Wu, Brooke E. Husic, Huanghao Mai, Yang Li, Saisai Sun, Jianyi Yang, Bharath Ramsundar, and Vijay S. Pande. 2018. PotentialNet for Molecular Property Prediction. ACS Central Science, Vol. 4, 11 (2018), 1520--1530.Google ScholarCross Ref
- Silvi Frenkel-Toledo, Nir Giladi, Chava Peretz, Talia Herman, Leor Gruendlinger, and Jeffrey M. Hausdorff. 2005. Treadmill walking as an external pacemaker to improve gait rhythm and stability in Parkinson's disease. Movement Disorders, Vol. 20, 9 (2005), 1109--1114.Google ScholarCross Ref
- Virginia L Giddings, Gary S Beaupre, Robert T Whalen, and Dennis R Carter. 2000. Calcaneal loading during walking and running. Medicine & Science in Sports & Exercise, Vol. 32, 3 (2000), 627--634.Google ScholarCross Ref
- Justin Gilmer, Samuel S. Schoenholz, Patrick F. Riley, Oriol Vinyals, and George E. Dahl. 2017. Neural Message Passing for Quantum Chemistry. In Proceedings of the 34th International Conference on Machine Learning (Proceedings of Machine Learning Research, Vol. 70). PMLR, 1263--1272.Google Scholar
- Rui Guo, Jie Sun, Chencheng Zhang, and Xiaohua Qian. 2022. A Contrastive Graph Convolutional Network for Toe-Tapping Assessment in Parkinson's Disease. IEEE Transactions on Circuits and Systems for Video Technology, Vol. 32, 12 (2022), 8864--8874.Google ScholarCross Ref
- Jeffrey M. Hausdorff, Justine Lowenthal, Talia Herman, Leor Gruendlinger, Chava Peretz, and Nir Giladi. 2007. Rhythmic auditory stimulation modulates gait variability in Parkinson's disease. European Journal of Neuroscience, Vol. 26, 8 (2007), 2369--2375.Google ScholarCross Ref
- Brian Kenji Iwana and Seiichi Uchida. 2021. An empirical survey of data augmentation for time series classification with neural networks. PLOS ONE, Vol. 16, 7 (07 2021), 1--32.Google ScholarCross Ref
- Thomas N. Kipf and Max Welling. 2017. Semi-Supervised Classification with Graph Convolutional Networks. arxiv: 1609.02907 [cs.LG]Google Scholar
- Soojin Lee, Ramy Hussein, and Martin J. McKeown. 2019. A Deep Convolutional-Recurrent Neural Network Architecture for Parkinson's Disease EEG Classification. In 2019 IEEE Global Conference on Signal and Information Processing (GlobalSIP). 1--4.Google Scholar
- Xu Liu, Wang Li, Zheng Liu, Feixiang Du, and Qiang Zou. 2021. A dual-branch model for diagnosis of Parkinson's disease based on the independent and joint features of the left and right gait. Applied Intelligence, Vol. 51 (2021), 7221--7232.Google ScholarDigital Library
- Xiaokai Liu, Zhaoyang You, Yuxiang He, Sheng Bi, and Jie Wang. 2022. Symmetry-Driven hyper feature GCN for skeleton-based gait recognition. Pattern Recognition, Vol. 125 (2022), 108520.Google ScholarDigital Library
- Francesca Miraglia, Fabrizio Vecchio, Chiara Pappalettera, Lorenzo Nucci, Maria Cotelli, Elda Judica, Florinda Ferreri, and Paolo Maria Rossini. 2022. Brain Connectivity and Graph Theory Analysis in Alzheimer's and Parkinson's Disease: The Contribution of Electrophysiological Techniques. Brain Sciences, Vol. 12, 3 (2022).Google ScholarCross Ref
- Meg E Morris, Robert Iansek, Thomas A Matyas, and Jeffery J Summers. 1994. Ability to modulate walking cadence remains intact in Parkinson's disease. Journal of neurology, neurosurgery & psychiatry, Vol. 57, 12 (1994), 1532--1534.Google ScholarCross Ref
- Duc Minh Dimitri Nguyen, Mehdi Miah, Guillaume-Alexandre Bilodeau, and Wassim Bouachir. 2022. Transformers for 1D signals in Parkinson's disease detection from gait. In 2022 26th International Conference on Pattern Recognition (ICPR). 5089--5095.Google ScholarCross Ref
- World Health Organization et al. 2022. Parkinson disease: a public health approach: technical brief. (2022).Google Scholar
- Xiaomin Pei, Huijie Fan, and Yandong Tang. 2021. Temporal pyramid attention-based spatiotemporal fusion model for Parkinson's disease diagnosis from gait data. IET Signal Processing, Vol. 15, 2 (2021), 80--87.Google ScholarCross Ref
- Bart Post, Lieneke Van Den Heuvel, Teije Van Prooije, Xander Van Ruissen, Bart Van De Warrenburg, and Jorik Nonnekes. 2020. Young onset Parkinson's disease: a modern and tailored approach. Journal of Parkinson's Disease, Vol. 10, s1 (2020), S29--S36.Google ScholarCross Ref
- Changqin Quan, Kang Ren, and Zhiwei Luo. 2021. A Deep Learning Based Method for Parkinson's Disease Detection Using Dynamic Features of Speech. IEEE Access , Vol. 9 (2021), 10239--10252.Google ScholarCross Ref
- Suraj Rajan, Bonnie Kaas, and Emile Moukheiber. 2019. Movement disorders emergencies. In Seminars in neurology, Vol. 39. Thieme Medical Publishers, 125--136.Google Scholar
- Abdelrahman Ramadan, Ahmed Elbery, Nizar Zorba, and Hossam S. Hassanein. 2020. Traffic Forecasting using Temporal Line Graph Convolutional Network: Case Study. In ICC 2020 - 2020 IEEE International Conference on Communications (ICC). 1--6.Google Scholar
- Carlos Salort Sánchez, Alexander Wieder, Paolo Sottovia, Stefano Bortoli, Jan Baumbach, and Cristian Axenie. 2020. GANNSTER: Graph-Augmented Neural Network Spatio-Temporal Reasoner for Traffic Forecasting. In Advanced Analytics and Learning on Temporal Data, Vincent Lemaire, Simon Malinowski, Anthony Bagnall, Thomas Guyet, Romain Tavenard, and Georgiana Ifrim (Eds.). Springer International Publishing, 63--76.Google Scholar
- Lei Shi, Yifan Zhang, Jian Cheng, and Hanqing Lu. 2018. Non-local graph convolutional networks for skeleton-based action recognition. arXiv preprint arXiv:1805.07694, Vol. 1, 2 (2018), 3--11.Google Scholar
- Lei Shi, Yifan Zhang, Jian Cheng, and Hanqing Lu. 2019. Two-Stream Adaptive Graph Convolutional Networks for Skeleton-Based Action Recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).Google ScholarCross Ref
- Krista G Sibley, Christine Girges, Ehsan Hoque, and Thomas Foltynie. 2021. Video-based analyses of Parkinson's disease severity: A brief review. Journal of Parkinson's disease, Vol. 11, s1 (2021), S83--S93.Google ScholarCross Ref
- Jehan Suleiman, Nadia Hamwi, and Ayman W. El-Hattab. 2018. ATP13A2 novel mutations causing a rare form of juvenile-onset Parkinson disease. Brain and Development, Vol. 40, 9 (2018), 824--826.Google ScholarCross Ref
- Rene L. Utianski, John N. Caviness, Elisabeth C.W. van Straaten, Thomas G. Beach, Brittany N. Dugger, Holly A. Shill, Erika D. Driver-Dunckley, Marwan N. Sabbagh, Shyamal Mehta, Charles H. Adler, and Joseph G. Hentz. 2016. Graph theory network function in Parkinson's disease assessed with electroencephalography. Clinical Neurophysiology, Vol. 127, 5 (2016), 2228--2236.Google ScholarCross Ref
- Petar Veličković, Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Liò, and Yoshua Bengio. 2018. Graph Attention Networks. arxiv: 1710.10903 [stat.ML]Google Scholar
- Xiaolong Wang, Ross Girshick, Abhinav Gupta, and Kaiming He. 2018. Non-Local Neural Networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Google ScholarCross Ref
- Yi Xia, ZhiMing Yao, Qiang Ye, and Nan Cheng. 2020. A Dual-Modal Attention-Enhanced Deep Learning Network for Quantification of Parkinson's Disease Characteristics. IEEE Transactions on Neural Systems and Rehabilitation Engineering, Vol. 28, 1 (2020), 42--51.Google ScholarCross Ref
- Keyulu Xu, Weihua Hu, Jure Leskovec, and Stefanie Jegelka. 2019. How Powerful are Graph Neural Networks?arxiv: 1810.00826 [cs.LG]Google Scholar
- Sijie Yan, Yuanjun Xiong, and Dahua Lin. 2018. Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognition. Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 32, 1.Google ScholarCross Ref
- Galit Yogev, Nir Giladi, Chava Peretz, Shmuel Springer, Ely S. Simon, and Jeffrey M. Hausdorff. 2005. Dual tasking, gait rhythmicity, and Parkinson's disease: Which aspects of gait are attention demanding? European Journal of Neuroscience, Vol. 22, 5 (2005), 1248--1256.Google ScholarCross Ref
- Aite Zhao, Junyu Dong, Jianbo Li, Lin Qi, and Huiyu Zhou. 2022. Associated Spatio-Temporal Capsule Network for Gait Recognition. IEEE Transactions on Multimedia, Vol. 24 (2022), 846--860.Google ScholarCross Ref
- Aite Zhao, Lin Qi, Jie Li, Junyu Dong, and Hui Yu. 2018. A hybrid spatio-temporal model for detection and severity rating of Parkinson's disease from gait data. Neurocomputing, Vol. 315 (2018), 1--8.Google ScholarCross Ref
- Cankun Zhong and Wing W. Y. Ng. 2022. A Robust Frequency-domain-based Graph Adaptive Network for Parkinson's Disease Detection from Gait Data. IEEE Transactions on Multimedia (2022), 1--14.Google Scholar
- Ömer Faruk Ertuğrul, Yılmaz Kaya, Ramazan Tekin, and Mehmet Nuri Almalı. 2016. Detection of Parkinson's disease by Shifted One Dimensional Local Binary Patterns from gait. Expert Systems with Applications, Vol. 56 (2016), 156--163.Google ScholarDigital Library
Index Terms
- Adaptive Spatio-Temporal Directed Graph Neural Network for Parkinson's Detection using Vertical Ground Reaction Force
Recommendations
A Robotic Cadaveric Gait Simulator With Fuzzy Logic Vertical Ground Reaction Force Control
Lower limb dynamic cadaveric gait simulators are useful to investigate the biomechanics of the foot and ankle, but many systems have several common limitations, which include simplified tendon forces, nonphysiologic tibial kinematics, greatly reduced ...
Ground reaction force and required friction during stair ascent and descent
A stair climbing experiment was performed using a four-stair instrumented stairway. A force platform was mounted in the first stair. Four floor surfaces, including the original force platform surface and three additional floor coverings on the platform, ...
Bayesian Analysis of Sub-plantar Ground Reaction Force with BSN
BSN '09: Proceedings of the 2009 Sixth International Workshop on Wearable and Implantable Body Sensor NetworksThe assessment of Ground Reaction Forces (GRF) is important for gait analysis for sports, pathological gaits and rehabilitation. To capture GRF, force plates and foot pressure insoles are commonly used. Due to cost and portability issues, such systems ...
Comments