An audibility based approach to model the head-shadow effect and the speech intelligibility in quiet and noise with a Percutaneous Bone Conduction Device in Single-sided Deaf Subjects
Guido Cattani (1), Bsc., M.SEN, Anne W. Wendrich (1,2), MD, Koenraad S. Rhebergen (1,2), PhD, Adriana L. Smit (1,2), MD, PhD,
- Department of Otorhinolaryngology, Head and Neck Surgery, University Medical Center Utrecht, Utrecht, The Netherlands
- Brain Center Rudolf Magnus, University Medical Center Utrecht, Utrecht, The Netherlands
New affiliation Guido Cattani: Royal Dutch Kentalis – Kentalis Academy, Sint Michielsgestel, The Netherlands; Koenraad S. Rhebergen https://orcid.org/0000-0002-4389-1267 ; Adriana L. Smit https://orcid.org/0000-0001-9126-9969 ;
Adriana L. Smit, Department of Otorhinolaryngology and Head & Neck Surgery, University Medical Center Utrecht, House Postal Number G05.129, PO BOX 85500, 3508 GA, Utrecht, The Netherlands, Tel: +31 88 75 56644,
To assess the ability of a percutaneous Bone Conduction Device (BCD) to optimally compensate for the head-shadow effect in patients with Single-Sided Deafness (SSD).
A retrospective analysis was performed using a database comprising patient measurements, fitting parameters, and acoustic properties of BCDs. The analysis compared sensation levels between bone-conduction (BC) and air-conduction (AC) sound paths in three spatial conditions with speech in quiet. Additionally, the Speech Intelligibility Index was employed to compute phoneme-scores in both quiet and noisy environments. Modelled data were compared with measured speech-in-noise data.
Twenty SSD adults using a percutaneous BCD.
Sixty percent of participants achieved optimal compensation for the head-shadow effect with the signal presented at the BCD side, and twenty percent with the signal in front. Phoneme-scores in quiet conditions showed no improvement over the unaided state. Noise conditions revealed mixed results, with slight to moderate improvement or slight worsening, depending on the specific scenario. The model validation was unsuccessful, potentially due to the speech-in-noise test used and model design.
According to an audibility model, percutaneous BCDs may not consistently provide optimal head-shadow effect compensation for SSD patients. Future investigations should concentrate on enhancing BCD fittings for this group, through specialized fitting algorithms and verification tools.
The code is provided in 2 different flavors: .ipynb and .html. If you want just to consult the code, it is best to use the html format. The html files provide also results, tables, and graphs. If you want to run the code, you have to use the ipynb files in the jupyter notebook (https://jupyter.org). Beside python and R, you will need different libraries. See the folder "Packages_py_version". For this purpose, you may consider the installation of the anaconda platform (https://www.anaconda.com). To get the code working properly, you have to set the input and output file paths according to your operating system and map tree.