Peri-Infarct Cortical Network Remapping During Stroke Recovery via a Soft Liquid Metal Neural Interface

Stroke causes substantial reorganization of neural activity and connectivity in the peri-infarct cortex during functional recovery. Tracking these changes at high resolution, however, remains difficult. Mechanical mismatch of conventional rigid neural probes is exacerbated in the fragile peri-infarct region, destabilizing the tissue-electrode interface and limiting long-term recording. Here, we present a fully soft, liquid metal-based multichannel neural interface that enables stable electrophysiological recording in freely moving mice. Using this platform, we performed longitudinal recordings before and after ischemic stroke, enabling within-animal comparisons. We applied complementary analyses of population-level latent dynamics and functional connectivity, capturing network-level coordination and pairwise interactions. Both analyses revealed acute suppression followed by partial recovery without restoration of the pre-stroke state. Instead, neural networks reorganized into a distinct and reproducible functional configuration. These findings establish soft liquid metal neural interfaces combined with population-level analysis as a useful approach for resolving neural reorganization during stroke recovery.