Machine Learning for Active Gait Support with a Powered Ankle Prosthesis

In recent years active prosthetic devices emerged as research area. In comparison to passive devices, active ones mimic the motion of our intact limbs more perfectly. Hence, they overcome several drawbacks a user of a passive devices is confronted with. Examples for benefits caused by switching from a passive to an active prosthesis are a more normal gait and reduced metabolic costs of transport. Even though active devices can improve amputee’s every day life, there are still challenges to be solved. Prosthesis control is besides design one of those challenges. The controller needs to decide when and how to adapt active support. Regarding, for instance, the push off phase, the active push off must be aligned with the timing and strength intended by the user. This work introduces recent supervised machine learning methods such as Gaussian process regression and support vector machines for control of active prosthetic devices. The machine learning methods are used to form a supervisory controller that infers the user’s intent. As input to the supervisory controller, we use the data obtained by an inertial measurement unit mounted at the shank of the considered active ankle prosthesis. The output or rather the users intent is given by gait, speed and gait percent predictions. If the intent is known, the desired nut position can be determined by a lookup. To enforce the desired trajectory a slave controller, here a PD controller, is applied. The supervisory controller is designed, implemented and tested based on walking and running data recorded on a treadmill. At first, noise free motion capturing data is used to demonstrate the applicability of supervised machine learning methods in context of active ankle control. Afterwards, sensor data obtained with the prosthesis’s intertial-measuremt unit is used to proof real-world applicability of the introduced supervisory controller. The obtained supervisory controller is fast and moreover accurate. For gait percent prediction the error is bounded to ±5%. In case of speed and gait prediction, accuracies close to 100% and 95% are achieved, respectively.

[1]  S. Miyazaki,et al.  Long-term unrestrained measurement of stride length and walking velocity utilizing a piezoelectric gyroscope , 1997, IEEE Transactions on Biomedical Engineering.

[2]  J. Platt Sequential Minimal Optimization : A Fast Algorithm for Training Support Vector Machines , 1998 .

[3]  John Platt,et al.  Probabilistic Outputs for Support vector Machines and Comparisons to Regularized Likelihood Methods , 1999 .

[4]  Roger Woodard,et al.  Interpolation of Spatial Data: Some Theory for Kriging , 1999, Technometrics.

[5]  Kamiar Aminian,et al.  Spatio-temporal parameters of gait measured by an ambulatory system using miniature gyroscopes. , 2002, Journal of biomechanics.

[6]  Stefan Schaal,et al.  Statistical Learning for Humanoid Robots , 2002, Auton. Robots.

[7]  Bernhard Schölkopf,et al.  A Primer on Kernel Methods , 2004 .

[8]  Jeffrey M. Hausdorff Gait variability: methods, modeling and meaning , 2005, Journal of NeuroEngineering and Rehabilitation.

[9]  P. Bonato,et al.  An EMG-position controlled system for an active ankle-foot prosthesis: an initial experimental study , 2005, 9th International Conference on Rehabilitation Robotics, 2005. ICORR 2005..

[10]  Radford M. Neal Pattern Recognition and Machine Learning , 2007, Technometrics.

[11]  H.A. Varol,et al.  Real-time Intent Recognition for a Powered Knee and Ankle Transfemoral Prosthesis , 2007, 2007 IEEE 10th International Conference on Rehabilitation Robotics.

[12]  D. Alvarez,et al.  Multisensor Approach to Walking Distance Estimation with Foot Inertial Sensing , 2007, 2007 29th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[13]  S.K. Au,et al.  Powered Ankle-Foot Prosthesis for the Improvement of Amputee Ambulation , 2007, 2007 29th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[14]  H.A. Varol,et al.  Decomposition-Based Control for a Powered Knee and Ankle Transfemoral Prosthesis , 2007, 2007 IEEE 10th International Conference on Rehabilitation Robotics.

[15]  S.K. Au,et al.  Biomechanical Design of a Powered Ankle-Foot Prosthesis , 2007, 2007 IEEE 10th International Conference on Rehabilitation Robotics.

[16]  Doheon Lee,et al.  Speed Estimation From a Tri-axial Accelerometer Using Neural Networks , 2007, 2007 29th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[17]  Jan Peters,et al.  Local Gaussian process regression for real-time model-based robot control , 2008, 2008 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[18]  Hugh M. Herr,et al.  Powered Ankle--Foot Prosthesis Improves Walking Metabolic Economy , 2009, IEEE Transactions on Robotics.

[19]  Carl E. Rasmussen,et al.  Gaussian processes for machine learning , 2005, Adaptive computation and machine learning.

[20]  He Huang,et al.  A Strategy for Identifying Locomotion Modes Using Surface Electromyography , 2009, IEEE Transactions on Biomedical Engineering.

[21]  Thomas G. Sugar,et al.  A novel control algorithm for wearable robotics using phase plane invariants , 2009, 2009 IEEE International Conference on Robotics and Automation.

[22]  Hartmut Geyer,et al.  Control of a Powered Ankle–Foot Prosthesis Based on a Neuromuscular Model , 2010, IEEE Transactions on Neural Systems and Rehabilitation Engineering.

[23]  Michael Goldfarb,et al.  Multiclass Real-Time Intent Recognition of a Powered Lower Limb Prosthesis , 2010, IEEE Transactions on Biomedical Engineering.

[24]  J. Shaw,et al.  Global estimates of the prevalence of diabetes for 2010 and 2030. , 2010, Diabetes research and clinical practice.

[25]  Michael Goldfarb,et al.  Standing Stability Enhancement With an Intelligent Powered Transfemoral Prosthesis , 2011, IEEE Transactions on Biomedical Engineering.

[26]  Michael Goldfarb,et al.  Upslope Walking With a Powered Knee and Ankle Prosthesis: Initial Results With an Amputee Subject , 2011, IEEE Transactions on Neural Systems and Rehabilitation Engineering.

[27]  C. Mayer,et al.  100 Krankheitsbilder in der Physiotherapie , 2011 .

[28]  Ken Endo,et al.  Speed adaptation in a powered transtibial prosthesis controlled with a neuromuscular model , 2011, Philosophical Transactions of the Royal Society B: Biological Sciences.

[29]  Herbert K. H. Lee,et al.  Gaussian Processes , 2011, International Encyclopedia of Statistical Science.

[30]  Qingguo Li,et al.  Inertial Sensor-Based Methods in Walking Speed Estimation: A Systematic Review , 2012, Sensors.

[31]  Alena M. Grabowski,et al.  Effects of a powered ankle-foot prosthesis on kinetic loading of the unaffected leg during level-ground walking , 2013, Journal of NeuroEngineering and Rehabilitation.

[32]  Stephan Rinderknecht,et al.  Does it pay to have a damper in a powered ankle prosthesis? A power-energy perspective , 2013, 2013 IEEE 13th International Conference on Rehabilitation Robotics (ICORR).