Inducing self-selected human engagement in robotic locomotion training

Area of Science:

• Neuroscience
• Robotics
• Rehabilitation Medicine

Background:

• Stroke causes significant mobility impairments, with manual therapy improving outcomes but limited by cost.
• Current robotic gait training is less effective than manual therapy, potentially due to patient disengagement with corrective assistance.
• Subconscious energy-minimizing drives may hinder active patient participation in robotic rehabilitation.

Purpose of the Study:

• To develop a new robotic gait rehabilitation method that leverages least-effort tendencies.
• To make therapeutic goals, like increased paretic limb use, easier to achieve through selective robotic assistance.
• To investigate if adaptive robotic assistance can guide patients toward desirable gait patterns.

Main Methods:

• A pilot study used a tethered robotic ankle-foot orthosis to alter self-selected stride length in a healthy subject.
• The system provided selective assistance, decreasing metabolic energy cost during treadmill walking.
• An adaptive protocol continuously monitored stride length and offered assistance to achieve a target increase.

Main Results:

• Voluntarily increasing stride length by 5% increased energy cost by 4%.
• Robotic assistance reduced energy cost by 6-8% at various stride lengths.
• Adaptive assistance led to a 4% increase in stride length and a 5% reduction in metabolic energy consumption.

Conclusions:

• Selective-assistance protocols can guide patients toward self-selected desirable gait patterns.
• This approach may enhance patient adherence and improve functional outcomes in robotic gait training.
• Harnessing least-effort tendencies offers a promising new direction for stroke rehabilitation robotics.