Feeling Without Touching: Cortical and Perceptual Hand Representations through Intracortical Microstimulation of the Human Somatosensory Cortex

Recent advancements in neurotechnology allow the use of implantable mechatronic systems to record and stimulate the human nervous system. Thanks to this intimate contact with neural structures (e.g., brain, spinal cord), such technology can both restore functions lost due to injury or disease (e.g., motor control, vision, speech) and study brain mechanisms with unprecedented resolution. One example is the human somatosensation, shaped by principles of somatotopic and hierarchical organization of the primary somatosensory cortex (S1). By measuring tactile perception, we learn not only about the organization of the somatic sensory system, but also about the formation of coherent mental body representations. Extensive evidence demonstrated large and systematic perceptual distortions of tactile space. However, emerging psychophysical phenomena have been studied mostly with natural touch, which undergoes extensive processing along the tactile system, and it is unclear at which stages distortions arise. Cortical interfacing with implantable electrodes has been used to target specific brain areas (e.g., motor, visual, speech), enabling both neural recording and stimulation. This technique plays a central role in brain-computer interfaces and neuroprosthetics, helping people with neurological conditions. Intracortical microstimulation (ICMS) of S1 can directly evoke vivid touch sensations, whose properties can be systematically manipulated by varying stimulation parameters. In this work, we used ICMS of human S1 in participants implanted with microelectrode arrays to define cortical-body maps of the hand and link them to mental-body maps. This neuroprosthetic tool enables exploration of hand representations with unprecedented granularity. To assess the resolution and reliability of ICMS-evoked sensations, we performed the Tactile Estimation Task in a participant with preserved contralateral sensation, allowing direct comparison between ICMS and natural touch on matched skin sites. ICMS was used to evoke percepts by directly activating cortical neurons projecting on specific patches of skin (i.e., projected fields). These locations were marked and then stimulated mechanically, enabling comparisons across modalities. Spatial discrimination threshold under ICMS (7.02mm) was slightly higher than of natural touch at the fingertip (4.03 mm), but comparable to palm-level threshold. Then, we compared distances between these points on skin and cortex (via inter-electrode spacing) to create detailed cortical-body maps, revealing somatotopy and magnification factors typical of the somatosensory homunculus. By collecting both physical and perceived distances, we defined the relationship between cortical-body and mental-body spaces. Perceptual biases emerged under both types of stimulation, rejecting the hypothesis that they originate from peripheral processing. These biases correlated with cortical representations, suggesting that S1's spatial layout contributes to perceptual distortions. By modeling perceived distances based on anatomical, neural and skin parameters, we confirmed that mental-body representations are shaped by multiple spatial factors. Our findings show that ICMS, delivered through chronic implants, provides a powerful tool to investigate brain coding in humans and potentially expand our understanding of cortical processing. Finally, this study contributes to bionic research by identifying neurotechnological needs and informing implant and robotic hand design for sensory restoration.