Abstract
Restoring sensory function lost by amputation remains a major challenge in prosthetics research. Peripheral nerve stimulation and targeted reinnervation techniques may partially restore somatotopic sensory feedback but their widespread adoption is hindered by the need for surgical interventions. In a study involving 17 able-bodied participants and five individuals with upper-limb amputation, we show that transcutaneous spinal cord stimulation (tSCS) can evoke proprioceptive sensations—such as movement, touch, and tapping—that appear to originate from the missing limb. Notably, these perceptions were primarily isolated to the phantom limb and were largely absent in the stump. Additionally, amputee participants experienced an increased level of comfort during tSCS. We further characterized how various stimulation parameters, such as electrode placement, carrier frequency, and burst frequency, modulated the quality and type of perceived sensations. Additionally, we show that tSCS maintained force proprioception necessary for effective prosthesis control. These results underscore the potential use of tSCS to provide sensory feedback for upper-limb prosthetics.
Introduction
For individuals with upper-limb amputations, the absence of sensory feedback transforms advanced prosthetics into mere tools rather than integrated limbs. This lack of "feeling" contributes significantly to high device abandonment rates. While invasive methods like peripheral nerve stimulation (PNS) or epidural spinal cord stimulation (eSCS) offer solutions, they involve surgical risks and high costs. This study investigates transcutaneous spinal cord stimulation (tSCS)—a non-invasive alternative—to determine its efficacy in restoring naturalistic sensations in the phantom limbs of amputees without interfering with motor control.
Methods
The study recruited 17 able-bodied (AB) participants and 5 upper-limb amputees (ULA). The experimental setup utilized a 5x5 cm cathode electrode placed over the cervical spine (C5-C7 or C6-T1) and two larger anode electrodes on the clavicles.
Stimulation consisted of kilohertz burst-modulated alternating current. Researchers tested carrier frequencies (CF) of 6 kHz and 10 kHz, and burst frequencies (BF) of 50 Hz, 100 Hz, and 150 Hz. Crucially, intensity was maintained at 90% of the discomfort or motor threshold to ensure purely sensory feedback. To validate the spinal origin of these sensations, tSCS was compared against Brachial Plexus Stimulation (BPS). Furthermore, an EMG modulation task was employed to confirm that tSCS did not disrupt the motor signals required for prosthetic control.
Results and Discussion
A primary finding was the high success rate of tSCS, which elicited sensations in the phantom hands of all amputees with prior phantom perception and in the fingertips of 93% of able-bodied participants.
Somatotopic Accuracy and Sensation Quality
Amputees reported more "natural" sensations, such as movement and tapping, compared to the "parasthetic" (tingling) sensations reported by able-bodied participants. This difference may be explained by "signal collision theory": in able-bodied individuals, artificial signals conflict with natural sensory input, whereas in amputees, the artificial signal reaches the brain without interference. Notably, tSCS successfully targeted the phantom hand while bypassing the residual limb (stump), a critical requirement for intuitive feedback.
Parameter Optimization and Comparative Efficacy
The 6 kHz carrier frequency proved more efficient (requiring lower current), while 10 kHz was rated as more comfortable. The C5-C7 electrode placement was generally more effective for distal hand targeting.
As shown in the comparative analysis, tSCS demonstrated a superior ability to evoke sensations in distal extremities (fingertips and hand) compared to brachial plexus stimulation, which more frequently targeted proximal segments like the upper arm.
Motor Control and Comfort
The EMG modulation task confirmed that participants could accurately control force levels during stimulation, indicating that tSCS is compatible with myoelectric prosthesis control. Unexpectedly, amputee participants reported an overall increase in comfort during the stimulation sessions, suggesting potential therapeutic benefits beyond sensory restoration.
Conclusion and Future Directions
This study demonstrates that sub-motor threshold tSCS is a viable, non-invasive method for providing somatotopic sensory feedback. While the results in the small amputee cohort are promising, future research should expand to include individuals with more proximal amputations (e.g., shoulder disarticulations). The ultimate goal remains the integration of tSCS into a closed-loop prosthetic system, where real-time sensor data from the prosthesis translates into immediate sensory feedback for the user.
Summary updated