Relationship between spectrotemporal modulation detection and music perception in normal-hearing, hearing-impaired, and cochlear implant listeners

Abstract

The objective of this study was to examine the relationship between spectrotemporal modulation (STM) sensitivity and the ability to perceive music. Ten normal-nearing (NH) listeners, ten hearing aid (HA) users with moderate hearing loss, and ten cochlear Implant (CI) users participated in this study. Three different types of psychoacoustic tests including spectral modulation detection (SMD), temporal modulation detection (TMD), and STM were administered. Performances on these psychoacoustic tests were compared to music perception abilities. In addition, psychoacoustic mechanisms involved in the improvement of music perception through HA were evaluated. Music perception abilities in unaided and aided conditions were measured for HA users. After that, HA benefit for music perception was correlated with aided psychoacoustic performance. STM detection study showed that a combination of spectral and temporal modulation cues were more strongly correlated with music perception abilities than spectral or temporal modulation cues measured separately. No correlation was found between music perception performance and SMD threshold or TMD threshold in each group. Also, HA benefits for melody and timbre identification were significantly correlated with a combination of spectral and temporal envelope cues though HA.

Conflict of interest statement

The authors declare that they have no competing interests.

Figures

Figure 1

Psychoacoutic performances of each subject. Results of detection thresholds (dB) for each stimuli condition for NH listeners, HA users, and CI users are shown in circle (•), square (■), and triangle (▲), respectively. Spectral modulation detection (SMD) threshold (A) and temporal modulation detection (TMD) threshold (B) are shown in the upper row. Spectrotemoral modulation (STM) detection thresholds (C) are shown in the lower row. STM detection thresholds for spectral rates of 0.5, 1.0, and 2.0 c/o are shown in the left, middle, and right columns, respectively. Bars and error bars represent mean detection thresholds and standard deviation. Asterisk (*) indicates significant difference between two groups in post-hoc analysis (adjusted P-value was 0.05/3 based on Bonferroni correction).

Figure 2

Music perception abilities for each subject. Results of music perception abilities for NH listeners, HA users, and CI users are shown in circle (•), square (■), and triangle (▲), respectively. Bars and error bars represent mean abilities and standard deviation. Asterisk (*) indicates significant difference between two groups in post-hoc analysis (adjusted P-value was 0.05/3 based on Bonferroni correction).

Figure 3

Scatter plots of mean spectrotemporal modulation (STM) detection thresholds and music abilities. X-axis represents music perception abilities. Y-axis represents mean STM detection test. Mean STM detection thresholds defines averaged thresholds across six different stimuli conditions. Results of scatter plots for NH listeners, HA users, and CI users are shown in green circle (•), blue squared (■), and red triangle (▲), respectively. Panel A indicates pitch discrimination scores. Panel B indicates melody identification scores. Panel C indicates timbre identification sores.

Figure 4

Scatter plots of difference in music perception performances between aided and unaided conditions and aided psychoacoustic performances. X-axis represents psychoacoustic thresholds. Y-axis represents difference in music perception between unaided and aided conditions. If music perception was improved after wearing HA, values of Y-axis are positive. Values of X-axis going to the right side means better psychoacoustic performance. Red circle indicates subjects with better unaided scores than limit of 95% confidence interval (1.35 semitones for pitch-direction discrimination, 78% for melody identification scores, and 55% for timbre identification scores).

Figure 5

Audiograms for normal hearing (NH) listeners, hearing aid (HA) users, and cochlear implant (CI) users. Panel A shows audiograms for NH listeners. Pure tone thresholds are shown in circle (⚪) for tested ear and in square (□) for non-tested ear. NH listeners had pure tone thresholds better than or equal to 25 dB HL at all frequencies in both ears. Panel B shows audiograms for HA users. Aided pure tone thresholds are shown in black circle (•) for both ears. Unaided pure tone thresholds are shown in circle (⚪) for the tested ear and in square (□) for the non-tested ear. HA users had pure tone average worse than or equal to 40 dB HL for 500 Hz, 1,000 Hz, 2,000 Hz, and 3,000 Hz in both ears. Panel C shows audiograms for CI users. Aided pure tone thresholds are shown in black circle (•) for the tested ear.