Figure 4.

Validation of ISI map retinotopy in V1 and HVAs with electrophysiology. (A) Example overlay of VFS and azimuth retinotopy. Note HVAs show distinct retinotopic coverage (e.g., LM versus PM). (B) Overlay of VFS (black) and azimuth retinotopic map contours (blue, 10° increments) aligned with vasculature. White circles: sites and average size of craniotomies after cranial window removal and alignment to vasculature. Location of craniotomies is used to determine expected azimuth RF location within V1 or HVAs. Data in (A,B) from Mouse 1 (Fig. S3). (C) Correlation between expected ISI azimuth coordinates (abscissa) versus observed RF location from local field potential (LFP) responses (ordinate) in superficial layers of V1 (n = 6 mice, 22 recording sessions). Error: 8.2° ± 7.6° (mean ± SD). Overall r² = 0.82, p = 3.8e⁻³¹; Black stimulus r² = 0.84, p = 5.9e⁻¹⁷; White stimulus r² = 0.80, p = 2.4e⁻¹⁵. no significant difference between expected and observed (p = 0.39) and white and black not significantly different (p = 0.46, Wilcoxon signed rank test). Blue circles indicate recordings from mouse and sites in (B). Data from mice 1,2,3,5,6,7 in Supplemental Fig. S3. (D) As (C), for deep V1 layers. Error: 13.4° ± 13.1° (mean ± SD). Overall r² = 0.51, p = 5.2e⁻¹⁴; Black stimulus r² = 0.61, p = 2.1e⁻⁹; White stimulus r² = 0.43, p = 3.0e⁻⁶. No significant difference between expected and observed (p = 0.07). (E,F) Like (C,D) for higher visual areas AL, LM, PM, and RL (n = 4 mice; 15 recording sessions). Superficial layers of HVAs show greater correlation to ISI coordinates (r² = 0.74, p = 2.6e⁻¹³) than deep layers (r² = 0.54, p = 2.9e⁻⁸). Data from mice 1, 2, 8, 9 in Supplemental Fig. S3 are shown here. (G–I) Regular spiking (RS) neuron temporal frequency, spatial frequency, and speed tuning in V1 (n = 82 neurons) and HVAs (LM: n = 181; AL: n = 46; RL: n = 108; AM: n = 67; PM: n = 125). Data from mice 1, 2, 9 (Fig. S3). No significant effect of area for TF tuning (p = 0.17). Significant main effect of area on SF tuning (p = 0.049), with LM preferring higher SFs than RL. Significant main effect of area on speed tuning (p = 0.0148), with RL preferring higher speed stimuli than V1 and PM. Kruskal–Wallis tests with Bonferroni correction for all comparisons. (J–L) Same sessions as (G–I), for fast spiking (FS) neurons in V1 (n = 27 neurons) and HVAs (LM: n = 19; AL: n = 4; RL: n = 28; AM: n = 13; PM: n = 34). No significant effect of area on FS tuning properties (TF: p = 0.615; SF: p = 0.057; Speed: p = 0.624).