PID Controller Design for FES Applied to Ankle Muscles in Neuroprosthesis for Standing Balance

• Published in: Frontiers in neuroscience Vol. 11; p. 347
• Main Authors: Rouhani, Hossein, Same, Michael, Masani, Kei, Li, Ya Qi, Popovic, Milos R
• Format: Journal Article
• Language: English
• Published: Switzerland Frontiers Research Foundation 20.06.2017; Frontiers Media S.A
• Subjects: Ankle; Arm; Balance; balance control; Controllers; Electrical stimuli; Engineering; Feasibility studies; functional electrical stimulation; inverted pendulum; Laboratories; Muscle fatigue; Muscles; neuromuscular modeling; neuroprosthesis; Neuroscience; Physiology; Posture; Prosthetics; Quality of life; Spinal cord injuries; Vestibular system; Canada; United States > US; functional electrical stimulation; neuroprosthesis; neuromuscular modeling; inverted pendulum; balance control
• ISSN: 1662-4548; 1662-453X; 1662-453X
• DOI: 10.3389/fnins.2017.00347

Closed-loop controlled functional electrical stimulation (FES) applied to the lower limb muscles can be used as a neuroprosthesis for standing balance in neurologically impaired individuals. The objective of this study was to propose a methodology for designing a proportional-integral-derivative (PID) controller for FES applied to the ankle muscles toward maintaining standing balance for several minutes and in the presence of perturbations. First, a model of the physiological control strategy for standing balance was developed. Second, the parameters of a PID controller that mimicked the physiological balance control strategy were determined to stabilize the human body when modeled as an inverted pendulum. Third, this PID controller was implemented using a custom-made Inverted Pendulum Standing Apparatus that eliminated the effect of visual and vestibular sensory information on voluntary balance control. Using this setup, the individual-specific FES controllers were tested in able-bodied individuals and compared with disrupted voluntary control conditions in four experimental paradigms: (i) quiet-standing; (ii) sudden change of targeted pendulum angle (step response); (iii) balance perturbations that simulate arm movements; and (iv) sudden change of targeted angle of a pendulum with individual-specific body-weight (step response). In paradigms (i) to (iii), a standard 39.5-kg pendulum was used, and 12 subjects were involved. In paradigm (iv) 9 subjects were involved. Across the different experimental paradigms and subjects, the FES-controlled and disrupted voluntarily-controlled pendulum angle showed root mean square errors of <1.2 and 2.3 deg, respectively. The root mean square error (all paradigms), rise time, settle time, and overshoot [paradigms (ii) and (iv)] in FES-controlled balance were significantly smaller or tended to be smaller than those observed with voluntarily-controlled balance, implying improved steady-state and transient responses of FES-controlled balance. At the same time, the FES-controlled balance required similar torque levels (no significant difference) as voluntarily-controlled balance. The implemented PID parameters were to some extent consistent among subjects for standard weight conditions and did not require prolonged individual-specific tuning. The proposed methodology can be used to design FES controllers for closed-loop controlled neuroprostheses for standing balance. Further investigation of the clinical implementation of this approach for neurologically impaired individuals is needed.