Beneficiary: HörSys, Germany
In order to better understand the reasons for different performance after cochlear implantation, we will investigate both technological factors (i.e. the dimensions and exact placement of the cochlear implant, its stimulation parameters, …) and biological factors (i.e. inter-individual variation of cochlear size and shape), and then analyse the impact of both on hearing outcomes in adults with cochlear implants.
Due to the necessity of analysing modern cochlear imaging (digital volume tomography, cone-beam CT) for the implantation, as well as the complexity of manually segmenting the relevant anatomical structures in CT scans. We first planned to create a deep learning based 3D automated classification and segmentation algorithm in order to automatically segment the cochlea-related regions when they appear on CT. In addition, computational approaches are combined to localise the cochlear implant electrode.
Finally by combining and correlating the above quantitative results with audiological test data, we plan to investigate potential correlations between them. This could be used to support the need of individualised cochlear implants and the need of cochlear imaging before implantation.
ESR 2: Yifan Wang