Investigating the spatial limits of somatotopic and depth-dependent sensory discrimination stimuli in rats via intracortical microstimulation.

Abstract

The somatosensory cortex can be electrically stimulated via intracortical microelectrode arrays (MEAs) to induce a range of vibrotactile sensations. While previous studies have employed multi-shank MEA configurations to map somatotopic relationships, the influence of cortical depth on sensory discrimination remains relatively unexplored. In this study, we introduce a novel approach for investigating the spatial limits of stimulation-evoked sensory discrimination based on cortical depth and somatotopic relationships in rodents. To achieve this, we implanted single-shank and four-shank 16-channel MEAs into the primary somatosensory cortex of male rats. Then, we defined distinct stimulation patterns for comparison, each consisting of four simultaneously stimulated electrode sites separated along the length of the single-shank device or between shanks for the four-shank device. Next, we utilized a nose-poking, two-choice sensory discrimination task to evaluate each rat's ability to accurately differentiate between these patterns. We demonstrate that the rats were able to reliably discriminate between the most superficial (450-750 μm) and deepest (1650-1950 μm) single-shank patterns with 90% accuracy, whereas discrimination between the most superficial and next adjacent pattern (650-950 μm) significantly dropped to 53% (p < 0.05). Similarly, in the four-shank group, discrimination accuracy was 88% for the furthest pattern pairs (375 μm difference) but significantly fell to 62% (p < 0.05) for the closest pairs (125 μm difference). Overall, the single-shank subjects could robustly differentiate between stimuli separated by 800 μm along a cortical column whereas, the multi-shank animals could robustly differentiate between stimuli delivered from shanks separated by 250 μm. Results showed that when spatial distances between stimuli patterns were decreased, the rats had reduced discriminable accuracy, suggesting greater difficulty when differentiating closely positioned stimuli. To better understand the single-shank results, we also utilized computational modeling to compare our in-vivo results against neuronal activation volumes presented in a biophysically realistic model of the somatosensory cortex. These simulations displayed overlapping volumes of activated neurons via antidromic propagation of axons for the closest pattern pair, potentially influencing discriminable limits. This work, which offers insight into how the physical separation of stimulating microelectrode sites maps to discernable percepts, informs the design considerations for future intracortical microstimulation arrays.