RESUMEN

This pilot study reports the development of a novel closed-loop (CL) FES-gait control system, which employed a finite-state controller that processed kinematic feedback from four miniaturized motion sensors. This strategy automated the control of knee extension via quadriceps and gluteus stimulation during the stance phase of gait on the supporting leg, and managed the stimulation delivered to the common peroneal nerve (CPN) during swing-phase on the contra-lateral limb. The control system was assessed against a traditional open-loop (OL) system on two sensorimotor 'complete' paraplegic subjects. A biomechanical analysis revealed that the closed-loop control of leg swing was efficient, but without major advantages compared to OL. CL automated the control of knee extension during the stance phase of gait and for this reason was the method of preference by the subjects. For the first time, a feedback control system with a simplified configuration of four miniaturized sensors allowed the addition of instruments to collect the data of multiple physiological and biomechanical variables during FES-evoked gait. In this pilot study of two sensorimotor complete paraplegic individuals, CL ameliorated certain drawbacks of current OL systems - it required less user intervention and accounted for the inter-subject differences in their stimulation requirements.

Asunto(s)

RESUMEN

This is a case series study with the objective of comparing two motion sensor automated strategies to avert knee buckle during functional electrical stimulation (FES)-standing against a conventional hand-controlled (HC) FES approach. The research was conducted in a clinical exercise laboratory gymnasium at the University of Sydney, Australia. The automated strategies, Aut-A and Aut-B, applied fixed and variable changes of neurostimulation, respectively, in quadriceps amplitude to precisely control knee extension during standing. HC was an "on-demand" increase of stimulation amplitude to maintain stance. Finally, maximal FES amplitude (MA) was used as a control condition, whereby knee buckle was prevented by maximal isometric muscle recruitment. Four AIS-A paraplegics undertook 4 days of testing each, and each assessment day comprised three FES standing trials using the same strategy. Cardiorespiratory responses were recorded, and quadriceps muscle oxygenation was quantified using near-infrared spectroscopy. For all subjects, the longest standing times were observed during Aut-A, followed by Aut-B, and then HC and MA. The standing times of the automated strategies were superior to HC by 9-64%. Apart from a lower heart rates during standing (P = 0.034), the automation of knee extension did not promote different cardiorespiratory responses compared with HC. The standing times during MA were significantly shorter than during the automated or "on-demand" strategies (by 80-250%). In fact, the higher isometric-evoked quadriceps contraction during MA resulted in a greater oxygen demand (P < 0.0001) and wider arteriovenous oxygen extraction (P = 0.08) when compared with the other strategies. In conclusion, even though increased standing times were demonstrated using automated control of knee extension, physiological benefits compared with HC were not evident.

Asunto(s)

RESUMEN

Objectives. To investigate the different approaches in the field of functional electrical stimulation (FES) control of gait and address fundamental perquisites to enable FES walking systems to become safer, more practical, and therefore clinically efficacious. Design. Systematic review was conducted from electronic data bases up to March 2008. Studies with innovative control strategies were highlighted for analysis, but all relevant literatures were described to deliver a broad viewpoint. Study Selection. FES studies applying 1) open and closed-loop controllers; 2) control algorithm techniques; or 3) feedback information to the control unit of neuromuscular stimulators via biological signals or artificial sensors. These studies were mostly associated to FES gait. Results. By far, more spinal cord-injured users have benefited from open-loop FES walking systems because they have had an easier and faster setup. However, because of their limitations over the control of knee extension, closed-loop control of gait may be a superior approach. The use of electromyogram to quantify quadriceps fatigue was not considered sufficiently appropriate to predict knee-buckle events; instead, the use of motion sensors for such purposes is recommended. Finite state controllers based on a set of deterministic rules to process feedback signals seemed more suitable to provide accurate command-and-control compared with dynamic or neural network controllers. Conclusions. Progress in the development of closed-loop FES walking systems has been impeded by their lack of practicality. In the near future, this obstacle could be overcome via implanted systems, especially if using controllers based on deterministic rule sets derived from motion sensor feedback.