Our hands and arms are incredibly dexterous and capable appendages. Controlling our upper limbs requires the integration of complex descending motor control signals with rich streams of sensory information returned from the limb. Advanced bionic prostheses are rapidly advancing, however the ability to intuitively control the device and receive sensory information remains a challenge. In response to these current barriers to intuitive control and feedback, nerve-machine interfaces have emerged that can decode a user’s motor intent directly from the human nervous system. However, many of these nerve-machine interfaces are experimental and may involve surgically implanted hardware, thus making this solution inaccessible to many. Currently at UC Davis Health, many individuals with upper limb loss have been receiving targeted muscle reinnervation for the prevention of phantom and neuroma pain (N-TMR). As there is a need for pain management, N-TMR is a rapidly emerging clinical treatment for pain management done at the time of amputation. In this work, we demonstrate that N-TMR not only offers prophylactic benefits but also provides opportunities for intuitive prosthetic control and feedback. Although N-TMR is more widely accessible, it faces many challenges with current advanced prostheses. With N-TMR, there is no consideration for the depth of the muscle being reinnervated nor electrical crosstalk which can impact surface measurement techniques used for prosthetic control such as surface electromyography. In this thesis work, we examined a novel control technique and sensory feedback method to address these challenges by employing ultrasound and machine learning approaches. Sonomyography (SMG) is an ultrasound imaging technique that applies image processing and pattern recognition algorithms to detect user’s intentions from muscle deformations when a missing hand movement is attempted. We found that four participants with N-TMR surgery could enact 4-10 hand grasps with 83.33-99.44% prediction accuracy. We also investigated how the vibration of these same N-TMR muscles could be used for movement feedback. We designed and benchtop tested a device that elicits the Kinesthetic Illusion. This illusion is the external stimulation of muscles with vibration between 70-110 Hz to activate muscle spindles. Activating the muscle spindles results in a sensation of the muscles stretching or elongating, thus participants report feeling their limb moving at the joint the muscle acts upon. We then demonstrated with five able-bodied participants and one N-TMR participant that this device can readily be used in experimental settings. Taken together, this work demonstrated that non-invasive accessible sensorimotor techniques for bionic upper limb control may be created, leveraged, applied, and adapted to the unique population of N-TMR patients.