Fig. 2: Evaluating the Brain Affective Arousal Signature (BAAS).
a and b depict predicted arousal experience compared to the actual level of arousal for the cross-validated discovery cohort (study 1, n = 60) and the independent validation cohort (study 2, n = 36), respectively. The colored lines show individual linear regression fits comparing predicted versus actual ratings, while the black line shows the regression fit for the complete dataset. The raincloud plots on the right show the distribution of within-participant predictions. c Distribution of within-subject trial-wise prediction-outcome Pearson correlation coefficient for each participant in the discovery and validation cohorts, respectively. Yellow color represents correlation coefficients in the discovery cohort (all P ≤ 3.8 × 10−3), while green color shows coefficients from the validation cohort, where all subjects except four showed significant correlations (P ≤ 0.04; four subjects had P > 0.05). Notably, the boxplot component of the raincloud plots in (a)–(c) represents the interquartile range with the box boundaries at the 25th and 75th percentiles, the median line at the 50th percentile, and the whiskers extending to the minimum and maximum values within 1.5 times the interquartile range from the quartiles. d BAAS could accurately predict (two-alternative forced-choice test) or strongly respond (one-sample t-test) to high arousal experiences induced by stimuli across multiple modalities. Upper panel: BAAS could accurately classify or predict high arousing negative affect in studies 3−6 (total n = 210, prediction performance is shown as forced-choice classification accuracy (Cohen’s d) or prediction-outcome Pearson correlation (uncorrected), P values in forced-choice tests were based on two-sided binomial tests); Left lower panel: BAAS could (marginally) significantly respond to high arousing positive affect in studies 7–16 (total n = 224, two-sided one-sample t-test); Middle lower panel: BAAS could accurately classify high arousing negative and positive affect in study 17 (total n = 150, prediction performance is shown as forced-choice classification accuracy (Cohen’s d), two-sided binomial test). Right lower panel: BAAS could strongly respond to high arousing positive and negative imaged events in study 18 (total n = 26, two-sided one-sample t-test, uncorrected). The violin and box plots in (d) show the distributions of the signature response. The boxplot’s boundaries are defined by the first and third quartiles, while the whiskers extend to the maximum and minimum values within the range of the median ± 1.5 times the interquartile range. See Supplementary Table S1 for the details of each contrast. # marginally significant, *P < 0.05, **P < 0.01, ***P < 0.001. BAAS Brain affective arousal signature. The icons were sourced from Pixabay under the Pixabay License or created in PowerPoint 2016. All icons are free to use in both commercial and noncommercial print and digital media. Source data are provided as a Source Data file.
