Fig. 6. Validation of cell-type-selective responses to PRF in a constant duty cycle.
a The temporal profiles of the specific tFUS waveforms employed in this study. Stimulation parameters are listed in detail in Table 2. Two examples of PRF levels are shown. Number of cycles per pulse for PRF 1500 Hz and 3000 Hz are 200 and 100, respectively, with associated TBD of 0.4 and 0.2 ms. f0 = 500 kHz, UD = 67 ms, ISoI = 2.5 s with 10% jittering. b, c The comparison for identified 417 single units across different PRF conditions under tFUS conditions with isoflurane anesthesia. 174 RSUs and 243 FSUs were separated. During tFUS condition, RSUs significantly differed their responses to different tFUS conditions (Kruskal–Wallis chi-squared = 70.61, p = 1.69 × 10−14). Data are shown as the mean ± s.e.m. of normalized firing rates, with statistical comparisons made through Kruskal–Wallis one-way ANOVA on ranks and post hoc one-tail two-sample Wilcoxon tests with Bonferroni correction for multiple comparisons (PRF 30 Hz vs. PRF 300 Hz: p = 4.52 × 10−5; PRF 30 Hz vs. PRF 1500 Hz: p = 5.76 × 10−11; PRF 30 Hz vs. PRF 3000 Hz: p = 4.18 × 10−9; PRF 30 Hz vs. PRF 4500 Hz: p = 2.86 × 10−8; PRF 300 Hz vs. PRF 4500 Hz: p = 1.05 × 10−3; PRF 1500 Hz vs. PRF 3000 Hz: p = 9.30 × 10−5; PRF 1500 Hz vs. PRF 4500 Hz: p = 7.92 × 10−7). *p < 0.005, ***p < 0.0001. NS, not significant. s.e.m., standard error of mean. d In a two-way ANOVA analysis, a statistically significant interaction (p = 4.51 × 10−10) between the neuronal type, i.e., RSU/FSU, and the PRF levels was found. Data are shown as the mean±s.d. of normalized firing rates. ***p < 0.001. s.d., standard deviation. e, f RSUs and FSUs showed homogeneous lack of response to the SSKF conditions. 120 RSUs, 128 FSUs. No significant effect of the ultrasound conditions was observed through statistics by Kruskal–Wallis two-sided H-test. Data are shown as the mean ± s.e.m. NS, not significant. s.e.m., standard error of mean. Source data are provided in the Source Data file.