Unlocking Robust Segmentation Across All Age Groups via Continual Learning

Chih-Ying Liu, Jeya Maria Jose Valanarasu, Camila Gonzalez, Curtis Langlotz, Andrew Y. Ng, Sergios Gatidis

Keywords: Image Segmentation, Age bias, Continual Learning

Abstract: Most deep learning models in medical imaging are trained on adult data with unclear per- formance on pediatric images. In this work, we aim to address this challenge in the context of automated anatomy segmentation in whole-body Computed Tomography (CT). We eval- uate the performance of CT organ segmentation algorithms trained on adult data when applied to pediatric CT volumes and identify substantial age-dependent underperformance. We subsequently propose and evaluate strategies, including data augmentation and con- tinual learning approaches, to achieve good segmentation accuracy across all age groups. Our best-performing model, trained using continual learning, achieves high segmentation accuracy on both adult and pediatric data (Dice scores of 0.90 and 0.84 respectively).

Submission Number: 98