People with upper-limb amputations are often disatisfied with the performance of prosthetic limbs. Myoelectric prostheses, which rely on electrodes attached to the skin surface, suffer from poor and unreliable performance, and a high rate of abandonment. Body-powered prostheses typically only provide one degree-of-freedom control (usually hand opening and closing), severely limiting their functionality. Both types of prosthetic limbs lack a means of providing sensory feedback to the user, requiring the user to rely on visual feedback when controlling the device. Our lab is working to develop solutions to improve prosthetic control and provide naturalistic sensory feedback to users of upper-limb prosthetics.
For transradial amputees, the extrinsic muscles that control most motions of the hand often remain intact after amputation. By recording electrical signals from electrodes directly implanted in these muscles, we can improve signal quality, stability, and selectivity over what can be achieved with electrodes on the surface of the skin. Additionally, by sending those signals into a realistic biomechanical model of the muscle and bone of the hand and arm, we hypothesize that we can improve the control of prosthetic limbs over what can be achieved by currently available machine learning or pattern recognition techniques. Ongoing studies in the lab involve temporary placement of fine-wire electrodes in the forearms of individuals with transradial amputation and intact limbs to determine their ability to accurately control a prosthesis using these realistic biomechanical models.
Sensory feedback is crucial for many of the dexterous tasks we perform every day, such as buttoning a shirt or tying shoes. Additionally, sensory feedback contributes to the sense of embodiment, in which we identify our limbs as part of our body, and a lack of sensory feedback causes people with amputation to view their prostheses as extracorporeal tools. There is also evidence that a lack of sensory feedback contributes to phantom limb pain, a condition that afflicts as many as 85% of amputees. By stimulating to dorsal roots, which are sensory nerves near the spinal cord, we aim to generate sensations that appear to emanate from the amputated limb and feel like touch, pressure, and limb movement. Signals recorded from sensors on a prosthetic limb can be used to modulate these sensations to provide functionally relevant feedback to the user. In this study, we are measuring the types and locations of sensations evoked by stimulation as well as the effects of sensation on motor control and phantom limb pain.