Project Summary Every 30s, someone in the world has a limb amputated, with approximately 2.1 million people living with amputations in the U.S.A. Upper limb amputees traditionally use passive, body-pow- ered, or electrically powered prostheses that use surface Electromyographic (EMG) signals from intact muscles in the residual limb for movement and rely on visual and/or surrogate sensory input. Advanced peripheral nervous (PN) system interfaces have been proposed as a viable mechanism to improve the control by amputees by delivering naturalistic sensory feedback from sensorized robotic prosthetics. Unfortunately, current neural interfaces suffer from common chal- lenges such as electrode failure, signal deterioration over time, and unstable or problematic sen- sory percepts (“stinging and tingling”) that remain a challenge. Our Hypothesis is that we can obtain superior control over the somatosensory thalamus and cortex (S1/VPL), which will lead to more controlled sensory percepts by using molecularly guided regenerative peripheral nerve con- duits that entice cutaneous and proprioceptive axons down their respective molecularly cued arms of a Y-shaped Regenerative Ultra-thin Multi-Electrode Interface (Y-MG-RUMEI) implanted at me- dian nerve transections. As the Ultra-thin MEIs have much smaller electrode surface areas it is hoped they will offer more local control and decrease excitation of large, possible unrelated, axons in the peripheral nerve. Additionally, we will utilize optimized microstimulation (opMiSt) through UMEIs embedded in the 2 branches of the Y with their molecular cues for touch and propriocep- tion, respectively. Neural recordings will be made from the S1/VPL towards the optimization of the peripheral MiSt. Three specific aims are included: SA1. Achieve selective MiSt from Y-MG-REMEIs with Proprioceptive and Cutaneous conduits in a rat model. SA2. Determine controllability of VPL/S1 via MiSt in Y-MG-RUMEI sensory conduits in a rat forepaw (hand) amputee model. SA3. Deter- mine if Y-MG-RUMEIs allow better bidirectional Brain-Nerve-Machine Interface (biBNMI) control, as compared to control-RUMIs, without molecular guidance, based biBNMI performance in a rat amputee model.