EUVIMEDEUVIMED entdecken
Beobachtungsstudie

Using Machine Learning to Predict Cochlear Implant Outcomes

Madeleine Anthonisen, Diane S Lazard, Alexandre Lehmann

Audiology and Neurotology · 2026

DOI: 10.1159/000550216 ↗

Kurz gefasst

Beobachtungsstudie in der Zeitschrift Audiology and Neurotology, 2026. Verfasst von Madeleine Anthonisen und 2 weiteren Autor:innen. Beteiligte Einrichtung: Faculté de médecine et des sciences de la santé, Université de Sherbrooke, Sherbrooke, Québec, Canada, antm5380@usherbrooke.ca. Hauptthemen laut MEDLINE: Cochlear Implantation, Cochlear Implants, Hearing Loss.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Introduction

Cochlear implant outcomes vary widely and are difficult to predict, with traditional methods explaining 18 years) with post-lingual hearing loss (onset >15 years) across fifteen centers in Australia, Europe, and North America. Data were collected between 2003 and 2011, with follow-up at 6 months and 2 years post-implantation. Linear Regression was compared against seven other machine learning models: Extreme Gradient Boosting (XGBoost), Random Forest, Categorical Boosting (CatBoost), Light Gradient Boosting Machine (LightGBM), Support Vector Regression, Ridge Regression, and Lasso Regression. Models were optimized using grid search with 5-fold cross-validation on an 80/20 training-test split. The primary outcome was prediction of percentile-ranked postoperative speech recognition scores in quiet, assessed using mean squared error (MSE) and coefficient of determination (R²). SHapley Additive exPlanations (SHAP) values identified feature importance.

Results

XGBoost achieved the best performance with a modest but significant 4.11% reduction in prediction error compared to linear regression (mean squared error: 739.67±19.27 vs 771.41±21.51, P = 0.003; R²: 0.114±0.011 vs 0.076±0.012, P < 0.001). All ensemble methods significantly outperformed linear regression. Duration of cochlear implant use, age at implantation, duration of severe/profound hearing loss, and preoperative hearing scores emerged as the most influential predictors across all models.

Conclusions

Machine learning models modestly improve cochlear implant outcome prediction, though substantial variance remains unexplained (>80%). Critical determinants of cochlear implant performance likely extend beyond variables routinely measured in clinical practice, highlighting the need for novel predictive factors.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Cochlear ImplantationCochlear ImplantsHearing LossAdultAgedFemaleHumansMaleMiddle AgedBoosting Machine Learning AlgorithmsLinear ModelsPredictive Learning ModelsRandom ForestRetrospective StudiesSpeech PerceptionTreatment Outcome

Stichwörter der Autor:innen

Clinical predictorsCochlear implant outcomesEnsemble methodsMachine learning

Autor:innen und Institutionen

  • Faculté de médecine et des sciences de la santé, Université de Sherbrooke, Sherbrooke, Québec, Canada, antm5380@usherbrooke.ca.
  • ENT Department, Princess Grace Hospital, Monaco, Monaco.
  • Université Paris Cité, Institut Pasteur, AP-HP, INSERM, CNRS, Fondation Pour l'Audition, Institut de l'Audition, IHU reConnect, Paris, France.
  • Otolaryngology - Head and Neck Surgery, McGill University, Montreal, Québec, Canada.

Zitieren

APA

Anthonisen, M., Lazard, D. S., & Lehmann, A. (2026). Using Machine Learning to Predict Cochlear Implant Outcomes. Audiology and Neurotology. https://doi.org/10.1159/000550216

Vancouver

Anthonisen M, Lazard DS, Lehmann A. Using Machine Learning to Predict Cochlear Implant Outcomes. Audiology and Neurotology. 2026. doi:10.1159/000550216

BibTeX
@article{Anthonisen2026,
  author = {Anthonisen, Madeleine and Lazard, Diane S and Lehmann, Alexandre},
  title = {Using Machine Learning to Predict Cochlear Implant Outcomes},
  journal = {Audiology and Neurotology},
  year = {2026},
  doi = {10.1159/000550216}
}
Kostenlos starten →