DESIGNING NOVEL STIMULI FOR SELECTIVE ACTIVATION AND FUNCTIONAL CLASSIFICATION OF RETINAL GANGLION CELL TYPES

Retinal ganglion cells (RGCs) are the sole output of the retina, carrying all visual information from retinal circuits to the rest of the brain. Decades of work have shown that RGCs comprise distinct functional populations that tile the visual field in a mosaic like pattern. Yet we still lack a reliable way to identify the functional subtype of each RGC in large-scale recordings. Standard functional typing relies mainly on responses to the chirp stimulus—a full-field modulation of light intensity over time. However, full-field stimulation disregards spatial computations that are known to happen in the retina (e.g., center–surround interactions) and often fails to distinguish similarly responding RGC types within and across experiments. To address this issue, we developed a barcode stimulus: a moving pattern of black and white stripes containing multiple spatial frequencies and applied it to ex vivo mouse retinas recorded with multi-electrode array (MEA). Compared with the chirp, the barcode stimulus elicited stronger and more reliable responses across stimulus repeats and its spatial richness reveals consistent differences between otherwise indistinguishable cells. We analyzed barcode response features with t-SNE (Stochastic Neighbor Embedding) and clustered cells in the resulting low-dimensional space. Preliminary results suggest that the t-SNE map organizes cells into broad neighborhoods matching canonical response classes (ON/OFF; transient/sustained-like), and within each neighborhood it revealed finer clusters whose receptive-field locations formed mosaic patterns. Together, our results suggest that the barcode - based clustering outperforms chirp based approaches for resolving functional RGC groups, improving the reliability of cell type labeling.