Precision Neuroscience 的 4096 通道 μECoG 用于术中中央沟定位

Objective. Intraoperative localization of the central sulcus (CS) using somatosensory evoked-potential (SSEP) phase reversal (PR) is routinely performed with sparse subdural strip electrodes. We evaluated whether a multi-array, 4096-channel surface micro-electrocorticography (μECoG) configuration could provide dense two-dimensional sampling of stimulus-dependent sensorimotor responses within a standard neurosurgical workflow. Approach. Four Precision Neuroscience Layer 7 μECoG arrays (1024 platinum electrodes per array on a flexible polyimide substrate, 400 μm electrode pitch, ∼1.5 cm2 active area per array) were deployed during resection of a right parafalcine meningioma in a 56-year-old male. Two arrays were placed horizontally on the exposed precentral gyrus; two were inserted under the intact dura overlying the postcentral gyrus, using a custom flexible stylet. Somatosensory evoked potentials (SSEPs) were recorded across five contralateral stimulation conditions (median nerve, ulnar nerve, index finger, middle finger, ring finger). Per-electrode SSEP responses were classified without anatomical labels, and spatial organization was quantified across stimulation conditions. Stimulus-locked high-gamma activity, digit-response maps, and signal quality across micro- and macroelectrodes were also evaluated. Main results. Aggregate channel yield was 91.3% (3739/4096) at a 2 MΩ impedance criterion (per-array range 86.7%–97.0%). Per-electrode amplitude maps resolved continuous phase-reversal contours with stimulus-specific spatial structure; reversal latencies were 19, 21, 25, 26, and 25 ms for median, ulnar, index, middle, and ring stimulation respectively. PR occurred within a single array in eight of ten sensory-array recordings, with the phase-reversal pattern varying across stimulation conditions. Digit responses showed measurable spatial differentiation within substantially overlapping response fields. Automated classification of the phase-reversal responses recovered the expected motor and sensory organization, consistent with the standard-of-care intraoperative localization performed in the same case. Significance. The deployment shows that a four-array Layer 7 μECoG configuration can be used within a standard neurosurgical exposure and provides dense two-dimensional sampling of stimulus-dependent phase-reversal patterns across the peri-Rolandic recording field. The measured responses were correlated across approximately 3–4 mm of cortex, an order of magnitude coarser than the 400 μm electrode pitch, so the contribution demonstrated here is dense spatial sampling rather than submillimeter physiological resolution. Dense sampling rendered the polarity transition as a continuous two-dimensional boundary across the recording field rather than as a reversal between two adjacent contacts. The platform supports future evaluation of high-density surface μECoG for intraoperative mapping and chronic brain–computer interface applications.