Sensory Feedback Restoration in Prosthetics Requires Direct Cortical Stimulation—Peripheral Nerve Interfaces Fail at Fine Discrimination

Current neural prosthetics restore motor control but leave patients with no sensation. The brain sends commands to robotic limbs, but receives no tactile or proprioceptive feedback in return. This one-way communication limits dexterity and prevents natural use.

Peripheral nerve interfaces attempt to restore sensation by stimulating residual nerves in the stump. These work for coarse sensations like pressure or vibration, but fail at fine discrimination—texture, temperature, exact grip force. The peripheral nerve has lost its sensory end organs; stimulating it electrically produces unnatural, imprecise percepts.

Direct cortical stimulation of somatosensory cortex bypasses the peripheral bottleneck. The brain already contains the neural architecture for sensation; it just needs the right input signals. Early trials show patients can distinguish between multiple virtual textures and adjust grip force appropriately when cortical stimulation provides feedback.

The question is whether we can encode naturalistic sensory information—slip detection, texture, limb position—into stimulation patterns the brain learns to interpret. This requires bidirectional brain-machine interfaces that read motor intention and write sensory experience simultaneously.