Powered Prostheses: Design, Control, and Clinical Applications

Cover
Title
Copyright
Contributors
Chapter 1 - Control of transhumeral prostheses based on electromyography pattern recognition: from amputees to deep learning
	1 - Introduction
	2 - Unmet needs for transhumeral amputees
	3 - Available upper limb prostheses
	4 - Myoelectric prosthesis, how does it work?
	5 - EMG pattern recognition
	6 - Deep learning
	7 - Usability instead of accuracy
	8 - Latest trend for transhumeral prosthesis
	9 - Discussion
	10 - Conclusion
	References
Chapter 2 - The 2-DOF mechanical impedance of the human ankle during poses of the stance phase
	1 - Introduction
	2 - Methods
		2.1 - Subjects
		2.2 - Experimental setup
		2.3 - Experimental protocol
		2.4 - Dynamic simulation with Simulink multibody
		2.5 - Analysis methods
			2.5.1 - Apparatus dynamics
			2.5.2 - Ankle impedance estimation
	3 - Results and discussion
		3.1 - Range of ankle torque and angle
		3.2 - Dynamics of the experimental apparatus
		3.3 - Mechanical impedance of the human ankle
		3.4 - Validation using the simulation
	4 - Conclusions
	References
Chapter 3 - Task-dependent modulation of multi-dimensional human ankle stiffness
	1 - Introduction
		1.1 Importance of human ankle stiffness
		1.2 Quasi-stiffness of the human ankle
		1.3 - Quantification of single-dimensional human ankle stiffness
	2 - Task-dependent modulation of multi-dimensional human ankle stiffness
		2.1 - Importance of multi-dimensional human ankle stiffness
		2.2 - Quantification of 2-dimensional human ankle stiffness during seated tasks
		2.3 - Quantification of 2-dimensional human ankle stiffness during walking tasks
		2.4 - Quantification of 2-dimensional human ankle stiffness during standing tasks
	3 - Summary
	References
Chapter 4 - Kriging for prosthesis control
	1 - Related work
	2 - Locomotion envelopes
		2.1 - Noise modeling
			2.1.1 - Complex noise models
		2.2 - Gait cycle stride-time regression
		2.3 - Analysis of human ambulation
		2.4 - Experimental data
	3 - Simulation setup
		3.1 - Impedance control
	4 - Prosthesis impedance controller
	5 - Empirical evaluation
		5.1 - Kernel design
		5.2 - Accelerating and decelerating
		5.3 - Torque-angle relationship at test points
			5.3.1 - Nearest neighbor interpolation
			5.3.2 - Linear regression
			5.3.3 - Piecewise cubic curvature-minimizing interpolation
		5.4 - Torque-angle relationship for held-out observations
		5.5 - Hardware experiments
			5.5.1 - Preprocessing and single-cycle extraction
			5.5.2 - Results
	6 - Discussion and conclusion
		6.1 - Conclusion
	References
Chapter 5 - Disturbance observer applications in rehabilitation robotics: an overview
	1 - Introduction
	2 - Background on disturbance observers
	3 - Disturbance observers in upper extremity robotic rehabilitation
		3.1 - Upper arm/forearm rehabilitation
		3.2 - Hand rehabilitation
	4 - Disturbance observers in lower extremity robotic rehabilitation
		4.1 - Ankle rehabilitation
		4.2 - Knee and knee-ankle rehabilitation
	5 - Disturbance observers in robotic telerehabilitation
	6 - Conclusion and further remarks
	References
Chapter 6 - Reduction in the metabolic cost of human walking gaits using quasi-passive upper body exoskeleton
	1 - Introduction
	2 - Methodology
		2.1 - Testing apparatus and procedure
			2.1.1 - Upper body passive exoskeleton fabrication and experimental results
			2.1.2 Upper body quasi-passive exoskeleton fabrication and experimental results
	3 - Discussion and conclusion
	References
Chapter 7 - Neural control in prostheses and exoskeletons
	1 - Overview
	2 - Sensory signals for control
	2.1 Physical signals
		2.2 - Physiological signals
		2.3 - Neural signals
	3 Neural control as part of the control architecture
		3.1 - Background
		3.2 - Neural control
			3.2.1 - Central pattern generators
		3.3 - Neuromuscular model for control
		3.4 - Application of neural control for gait assistance
	4 - Template-based neural control
		4.1 - Why template-based control?
		4.2 - Neuromechanical-template-based control
		4.3 - FMCA concept
			4.3.1 - Simulation
			4.3.2 - Human experiments
	5 - Summary
	References
Chapter 8 - Stair negotiation made easier using low-energy interactive stairs
	1 Introduction
	2 Materials and methods
		2.1 Energy-recycling assistive stairs
		2.2 Human experiment
			2.2.1 - Gait phase definition
			2.2.2 - Joint work calculation
			2.2.3 - Statistical analysis
	3 Results
		3.1 Operation of energy-recycling stairs
		3.2 Assessment of assistance provided during stair negotiation
			3.2.1 - Stair ascent
			3.2.2 Stair descent
	4 Discussion
		4.1 Implications of ERAS as an assistive device
		4.2 Experimental limitations and justifications
	5 Summary
	Acknowledgments
	References
Chapter 9 - Semi-active prostheses for low-power gait adaptation
	1 - Adaptability in lower-limb control
		1.1 - Current semi-active prostheses
		1.2 - Benefits of modulating properties in prostheses
	2 - The semi-active approach to biomechatronic design
	3 - The semi-active design process
		3.1 - Motivating a semi-active device
			3.1.1 - “What is the body doing?”—defining models of biomechanical behavior
				3.1.1.1 - Describing behavior
				3.1.1.2 - Conceptual modeling
				3.1.1.3 - Mathematical modeling
			3.1.2 - “What happens if the task is done differently?”—predicting the effects of device parameter changes
			3.1.3 - “How can we tell if the movement changes?”—developing metrics to quantify behavior
			3.1.4 - “Do existing prostheses produce a deficit?”—evaluating the potential value
		3.2 - Designing the semi-active mechanism
			3.2.1 - “How could a semi-active device accomplish the goal?”—design for semi-active actuation
			3.2.2 - “How and when should the prosthesis adapt?”—Controlling semi-active adaptation
		3.3 - Evaluating performance in semi-active devices
			3.3.1 - “Does it work?”
			3.3.2 - “Does it have the intended effect?”
			3.3.3 - “Do people like it?”
		3.4 - Iterate, iterate, iterate
	4 - Design principles for semi-active mechanisms
		4.1 - Exploiting the gait cycle
		4.2 - Clutches and latches
		4.3 - Non-backdriveability
			4.3.1 - Friction as an ally
			4.3.2 - Orthogonal and Inclined-plane actuation
			4.3.3 - Kinematic singularities
		4.4 - Springs to replace actuators and simplify control
		4.5 - High static forces, quick unloaded movements
		4.6 - Simplicity and durability
	5 - Example concepts and design implementations
		5.1 - The foot as a wheel
			5.1.1 - Description of behavior: rolling
			5.1.2 - Mathematical model: radius of a wheel
			5.1.3 - Variations: effects of different radii
			5.1.4 - Implementation: the Rock’n’Lock foot
				5.1.4.1 - Semi-active design features of the Rock’n’Lock
				5.1.4.2 - Performance and benefits of the Rock’n’Lock
				5.1.4.3 - Challenges of the Rock’n’Lock design
		5.2 - The ankle as a spring
			5.2.1 - Description of behavior: increasing ankle moment with increasing angle
			5.2.2 - Mathematical model: torque versus ankle angle relationship
			5.2.3 - Variations: effects of different stiffnesses
			5.2.4 - Implementation: the variable-stiffness foot (VSF)
				5.2.4.1 - Semi-active design features of the VSF
				5.2.4.2 - Performance and benefits of the VSF
				5.2.4.3 - Challenges of the VSF design
		5.3 - The ankle as an angle adapter
			5.3.1 - Description of behavior: ankle adaptation to slopes, stairs, speeds, and turns
			5.3.2 - Mathematical model: controlling the neutral angle of a spring or shape
			5.3.3 - Variations: effects of angle changes
			5.3.4 - Implementation: The Two-Axis ‘Daptable Ankle (TADA)
				5.3.4.1 - Semi-active design features of the TADA
				5.3.4.2 - Performance of the TADA
				5.3.4.3 - Challenges of the TADA design
	6 - Challenges of semi-active prostheses
	7 - Conclusion
	References
Index
Back Cover