Figure 4 | Scientific Reports

Figure 4

From: Pupillometry as a reliable metric of auditory detection and discrimination across diverse stimulus paradigms in animal models

Figure 4The alternative text for this image may have been generated using AI.

Pupillometric estimates of call-in-noise detection thresholds. (a) Spectrogram of a guinea pig purr call. (b) Purr call at − 3 dB SNR, obtained by adding white noise to the ‘clean’ purr call in (A). (c) Structure of call-in-noise detection paradigm. Standard stimuli were white noise bursts; deviants were calls embedded in white noise at different SNR levels. (d) Examples of PD traces from one subject for detecting chut calls in noise. Blue line and shading correspond to pupil responses to standard stimuli, and black line and shading correspond to mean ± 1 s.e.m. of pupil responses to deviants. Color gradient corresponds to stimulus SNR. Green line and shading denote pupil trace at threshold (− 1.5 dB SNR). Red line signifies stimulus onset, and teal dashed lines demarcate the GCA window. (e) Pooled and averaged pupil responses across 5 subjects. (f) GCA weight estimates for the detection of chuts, purrs, and harmonic tones in noise (bar colors as earlier). Weight estimates for deviants were significantly different from the standard for most SNRs above − 1.5 dB SNR. Asterisks correspond to p < 0.01 (hypothesis test for linear model coefficients, exact p-values in Supplementary Tables 3–5). Note that the + 6 dB SNR data point was not sampled for purr calls. (g) Psychometric functions fit to the percent of trials that evoked significant PD changes as a function of SNR for purr-in-noise (yellow), chut-in-noise (black), and harmonic tone-in-noise (orange). Psychometric functions were largely similar, reaching 50% of maximum at about − 1.5 dB SNR.

Back to article page