Abstract
Rapid ear canal growth in infancy and childhood poses a challenge for maintaining accurate hearing aid fitting during critical developmental stages. A method to predict subject-specific ear canal growth has the potential to optimize paediatric ear mould preparation, enabling timely refitting while reducing the need for frequent visits to specialised audiological clinics. We present a framework for modelling ear canal growth using implicit neural representations. Our approach builds on unsigned distance fields and uses a latent vector for capturing subjectspecific growth trajectories. The contributions of this work are threefold: (i) A novel use of implicit neural representations for growth modelling, benefiting from their inherent continuity across both space and time; (ii) a robust scheme for handling variable-length data sequences and irregularly sampled time points; and (iii) an approach that accommodates open surface meshes, enabling the modelling of open anatomical structures such as the ear canal without the need to artificially close the geometry. We evaluate the method on a longitudinal dataset of 912 3D ear canal scans from children aged 1 month to 10.5 years, and show that it robustly models anatomical growth, achieving a mean Chamfer distance of 0.65 mm in predictions of ear canal shapes at future time points.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026-sat/paper/PedAItrics_009.pdf
SharedIt Link: Not yet available
SpringerLink (DOI): Not yet available
Supplementary Material: Not Submitted
Link to Open Review
BibTex
@InProceedings{GulKar_Growth_MICCAISAT2026,
author = { Guldhammer-Skov, Karen AND Sørensen, Kristine Aavild AND López Diez, Paula AND Palama, Christian AND Kjærbøl, Erik F. AND Encina-Llamas, Gerard AND Paulsen, Rasmus R. AND González Ballester, Miguel A.},
title = { { Growth modelling of children’s ear canals using implicit neural distance representations } },
booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026 Workshops and Challenges},
year = {2026},
publisher = {Springer Nature Switzerland},
volume = {LNCS 17257},
month = {pending},
page = {pending}
}