Fig. 5: An expanded parameter search for tFUS continues to find no source localizable activity in resting-state humans.
A follow-up study was conducted in 10 subjects (Supplementary Table T2) to test for possible changes in activity from varying the duty cycle or PRF from the original study parameters. A EEG data were projected to the cortex using the dSPM source imaging method. Subject data were averaged over the 10–63 ms post-stimulation window and source morphed to FreeSurfer’s FSAverage common brain model and averaged together for visualization. The cut-off level of activity for brain sources were defined based on Otsu’s threshold. Visualizations continue to show strong deep brain and temporal lobe activations, but nothing concordant with the ROI. B No tFUS-only parameter combination yielded significantly higher ipsilateral than contralateral ROI source activity (one-tailed permutation t-test with FDR multiple comparison correction; all padj > 0.05). Each subplot contains a violin plot highlighting the distribution of the data. A box plot is embedded within the violin plot, marking the data median (center line), 1st and 3rd quartiles (whiskers = 1.5 * quartile ranges) and means (white circle). The individual datapoints are presented in color-matched circles. C Comparisons across the subjects’ tFUS-only ipsilateral ROI responses were compared for constant PRFs and a constant duty cycle. Their 30 Hz and 3 kHz PRF with 30% DC data from the initial session were also included in this analysis. No significant changes were detected with a permutation one-way ANOVA test (all padj > 0.05). Beyond the lack of statistic differences, Bayes Factor analysis indicates the resting-state tFUS responses are statistically similar across the parameter spaces. The p-values for all tests are listed in the Source Data file. The box plots mark the data median (center line), and 1st and 3rd quartiles (whiskers = 1.5 * quartile ranges). The individual datapoints are presented in color-matched circles.
