Assessing Temporal Fine Structure Processing Indirectly: An AM/FM Interference Task
Nihaad Paraouty, Nicolas Wallaert, Daniel Pressnitzer & Christian Lorenzi
DEC - Ecole Normale Supérieure Paris, Laboratoire des Systèmes Perceptifs - CNRS UMR 8248
It is believed that auditory sensitivity to amplitude modulation (AM) reflects the use of temporal-envelope (E) cues, whereas sensitivity to slow frequency modulation (FM) primarily reflects the use of temporal-fine structure (TFS) cues for carrier frequencies up to 1 kHz. Several psycho-acoustical studies have demonstrated that hearing-impaired listeners show poorer-than-normal sensitivity to FM, and normal or even better-than-normal sensitivity to AM, consistent with the notion that cochlear damage alters neural phase locking to TFS cues while sparing (or even enhancing) processing of E cues. However, the perceptual deficit observed for FM detection may partly result from reduced “processing (central) efficiency”, rather than from alterations in neural phase-locking to TFS cues. We developed an interference paradigm in order to address this issue. Importantly, the paradigm was designed such that poor sensitivity to TFS should actually lead to better perceptual performance. This paradigm was based on a pilot study showing that AM detection worsens when the tonal carrier is frequency modulated at the same rate as the AM. Detection thresholds were measured for a 5-Hz sine AM applied to a 0.5-kHz tone carrier. The latter was either unmodulated or frequency-modulated at 5 Hz at increasing FM depths. It was reasoned that listeners with poor TFS sensitivity (and thus, poor FM sensitivity) should show better AM detection thresholds than listeners with good TFS sensitivity when the pure-tone carrier is modulated in frequency. For each listener, psychoacoustic measures of TFS sensitivity (i.e., thresholds for detecting a change in interaural phase, IPD) and frequency selectivity (i.e., thresholds for detecting a pure tone in notched noise using 2 notch widths: 0 and 150 Hz) were also obtained at 0.5 kHz. It was reasoned that the magnitude of the interference effect between AM and FM should be predicted by IPD scores only, if this interference effect reflects the use of TFS (neural phase-locking) cues rather than the conversion of FM into AM (E cues) at the output of cochlear filters. Data collected on young and elderly normal-hearing listeners will be presented and discussed. Preliminary results obtained in young normal-hearing listeners indicate that as expected, AM detection thresholds increase (become poorer) systematically when the carrier is frequency modulated, and the magnitude of this interference effect increases systematically as a function of FM modulation depth. We plan to extend the study with data from hearing-impaired listeners, with a particular focus on patients suffering from acoustic trauma.
This work was supported by a grant from ENTENDRE-SAS and ANR HEARFIN.