Threshold estimation and speech perception in HLS (Roman et al., 2023)

Purpose: Hearing loss simulation (HLS) has been recommended for clinical teaching and counseling of patients and their families, so that they can experience hearing impairment. However, few validated procedures for simulating hearing loss are available to instructors and practicing clinicians. The aim of this study was to assess the accuracy of the Immersive Hearing Loss and Prosthesis Simulator (I-HeLPS) on reducing hearing sensitivity and word recognition to determine its adequacy for educational and clinical use.

Method: Thirty-seven young adults with normal hearing completed hearing threshold and word recognition testing under normal and simulated hearing losses. The accuracy of the nominal hearing threshold settings within the IHeLPS software was assessed with behavioral detection of frequency-modulated pure tones presented in a calibrated sound field, while listeners wore I-HeLPS headphones. The impact of the HLSs on speech perception was measured using the California Consonant Test. Hearing thresholds, word identification accuracy, and sound confusions were compared across listening conditions.

Results: Hearing thresholds increased systematically with worse simulated hearing loss. Performance on the California Consonant Test worsened, and the number of phoneme confusions increased with simulated hearing loss severity. Most of the confusions were place confusions with near neighbors and manner confusions increased as a function of increasing severity of simulated hearing loss.

Conclusions: The I-HeLPS accurately elevated hearing thresholds with increasing HLS severity and impacted word recognition in a manner consistent with sensorineural hearing loss. The simulations were considered reasonable for educational and clinical purposes.

Supplemental Materials

The Supplemental Materials include 10 tables containing confusion matrices for the California Consonant Test and from which the summary consonant error data for the manuscript were derived. The matrices present the confusions found for all five hearing loss simulation conditions and are presented as both counts and percentages.

When reviewing the matrixes, keep in mind that the California Consonant Test has a high-frequency emphasis and includes CVC word stimuli that are problematic for listeners with high-frequency hearing loss. The English speech sounds in the test are not equally represented, nor is the location within the word stimuli. As indicated in the manuscript, 80 of the 100 target consonants were voiceless and 64 were located at the end of the words.

Supplemental Material S1. Confusion matrix for the Normal Hearing condition based on number of responses. The shading reflects proportion of responses.

Supplemental Material S2. Confusion matrix for the Normal Hearing condition based on proportion of responses. The shading reflects proportion of responses.

Supplemental Material S3. Confusion matrix for the Mild Hearing Loss Simulation condition based on number of responses. The shading reflects proportion of responses.

Supplemental Material S4. Confusion matrix for the Mild Hearing Loss Simulation condition based on proportion of responses. The shading reflects proportion of responses.

Supplemental Material S5. Confusion matrix for the Moderate Hearing Loss Simulation condition based on number of responses. The shading reflects proportion of responses.

Supplemental Material S6. Confusion matrix for the Moderate Hearing Loss Simulation condition based on proportion of responses. The shading reflects proportion of responses.

Supplemental Material S7. Confusion matrix for the Severe Hearing Loss Simulation condition based on number of responses. The shading reflects proportion of responses.

Supplemental Material S8. Confusion matrix for the Severe Hearing Loss Simulation condition based on proportion of responses. The shading reflects proportion of responses.

Supplemental Material S9. Confusion matrix for the Maximum Hearing Loss Simulation condition based on number of responses. The shading reflects proportion of responses.

Supplemental Material S10. Confusion matrix for the Maximum Hearing Loss Simulation condition based on proportion of responses. The shading reflects proportion of responses.

Roman, A. M., Pratt, S. R., & Zhen, L. Q. (2023). Threshold estimation and speech perception under hearing loss simulation: Examination of the immersive hearing loss and prosthesis simulator. American Journal of Audiology, 33(1), 275–282. https://doi.org/10.1044/2023_AJA-23-00155