Temporal modulation transfer functions derived from envelope following responses: what can they tell us about...

This study investigates how the properties of neural temporal modulation transfer functions (TMTFs) derived from envelope following responses (EFRs) to amplitude-modulated sounds relate to the hypothetical tuning of individual neurons in the midbrain. We followed a joint modeling and empirical approach. We measured EFRs for young adults with normal hearing (n = 15) using rectangular amplitude-modulated (RAM) tones with modulation frequencies varying between 70 and 160 Hz, to target the most sensitive region of brainstem/midbrain neurons. These empirical data portrayed a large variability across individuals, both in terms of TMTF shapes and gains; at the individual level, most individuals exhibited TMTFs band-pass or low-pass in shape, but at the group level, there was no significant trend. We also conducted simulations using computational models of the auditory periphery and midbrain to examine how simulated, EFR-derived TMTFs vary depending on hypothesized models of inferior-colliculus (IC) neurons, their parameters, and distributions. When considering a population of band-pass-tuned IC cells with varying best modulation frequencies, simulations suggest that the magnitude of the TMTF, rather than its shape, might actually better reflect a change in tuning. These experimental and simulation results are discussed in relation to previous works, along with additional simulations showing that the type of stimulus envelope (sinusoidal vs. rectangular modulation) or subtle threshold variations among individuals with normal hearing have only a limited impact on these trends. From these results, we derive several considerations for the interpretation of EEG-based TMTFs and the potential information they provide regarding auditory midbrain tuning.NEW & NOTEWORTHY What do EEG-derived temporal modulation transfer functions (TMTFs) tell us about human auditory midbrain tuning? We found substantial individual variability in both the shapes and gains of empirically derived TMTFs. Computational model simulations suggest that the TMTF's magnitude, rather than its shape, might better reflect underlying neural tuning changes. These results provide new perspectives for modeling individual differences and for developing more sensitive clinical tests of subcortical temporal processing.