Abstract
Understanding how sensory information is transformed along cortical hierarchies requires reliable measurement of sensory representations across areas. In vivo two-photon calcium imaging through chronic cranial windows has become an essential approach for characterizing cortical sensory representations with cellular resolution and spatial registration across experiments. However, cortical computations, particularly in higher-order cortices, depend on intricate intracortical circuitry and can be sensitive to subtle tissue perturbations introduced during surgical preparation. Here, we report that implantation of large cranial windows can mechanically compress the intrinsically curved cortical surface, producing systematic distortion that is usually not recognized as overt tissue damage. By comparing sensory maps in the mouse auditory cortex before and after window implantation, we show that larger windows commonly used across laboratories are associated with distorted sensory maps away from the center of the window. Furthermore, both macroscopic and cellular-level imaging reveal a deterioration of feature-selective responses in the secondary auditory cortex (A2) following such distortion. Together, these findings identify an overlooked but controllable surgical factor that can bias measurements of sensory representations and highlight an important consideration for enhancing reproducibility across studies.