Following an upper limb amputation, both the functional and sensory capabilities of the hand are lost. Prior studies demonstrated the ability of neural stimulation to evoke sensations on the missing hand of persons with upper limb loss. However, the sensations have only been cursorily assessed in terms of their perceptual characteristics and psychosocial impacts. Further, we do not know to what extent mimicking natural firing patterns can improve the stimulation-evoked sensations.
Using classical psychophysical techniques, we characterized the perception of intensity of sensations evoked by electrical stimulation through Flat Interface Nerve Electrodes (FINEs) implanted in three humans with upper limb loss, and compared these characteristics to natural intensity perception. We found that intensity perception is equivalent for natural and artificial touch, including intensity sensitivity, dynamic range, and the time course and extent of adaptation resulting from prolonged stimulation. We demonstrated that population spike rate is the neural basis of perceived intensity.
To determine the impact of sensation on the holistic experience of having a hand, two persons with upper limb loss utilized a sensory restoration system in their homes and communities. We found that sensation improved the psychosocial experience of using the prosthesis, including confidence in abilities, prosthesis incorporation, perception of social interactions, and body image. Sensation improved prosthesis function and increased prosthesis usage. This study indicated that extraneural stimulation is a feasible method of long-term sensory restoration in community use. Participant perspectives revealed that sensation was critical for outcome acceptance. We developed a theoretical model of the impacts of sensation based on a qualitative analysis of participant experiences, which may provide a unified framework to study outcome acceptance following amputation.
To determine whether mimicking natural afferent activity would improve stimulation-evoked sensations, we developed a computational model to construct biomimetic stimulation patterns and predict the evoked firing patterns in the recruited afferent population. We tested the biomimetic patterns in three humans implanted with FINEs, and found subject-specific differences in the perceived naturalness of the biomimetic patterns. The perceived naturalness of the stimuli was influenced by subject-specific preferences, expectations for intensity, and possibly learning. The computational model demonstrated that biomimetic stimulation evokes more natural neural activation patterns than conventional stimulation. The model may be useful for generating hypotheses about the neural correlates of perception that can be tested in future studies.
This work demonstrates how neural stimulation can serve as a tool to study neural coding. The perceptual similarities between natural and artificial touch indicates that stimulation-induced sensation likely utilizes the same sensory processing pathways as natural sensation. For persons with limb loss, sensory restoration reshapes the holistic experience of the prosthesis into having a hand that can feel. Finally, stimulation patterning can evoke complex patterns of neural activity which modify the sensory percepts.