Investigating Gender Bias in Lymph-node Segmentation with Anatomical Priors
It addresses fairness issues in medical imaging for radiotherapy planning, though it is incremental as it builds on existing deep learning methods with prior information.
This study tackled gender bias in lymph-node segmentation for radiotherapy by using anatomical priors from simpler structures, finding that this approach improved segmentation quality in female patients and reduced bias, particularly in the abdomen region.
Radiotherapy requires precise segmentation of organs at risk (OARs) and of the Clinical Target Volume (CTV) to maximize treatment efficacy and minimize toxicity. While deep learning (DL) has significantly advanced automatic contouring, complex targets like CTVs remain challenging. This study explores the use of simpler, well-segmented structures (e.g., OARs) as Anatomical Prior (AP) information to improve CTV segmentation. We investigate gender bias in segmentation models and the mitigation effect of the prior information. Findings indicate that incorporating prior knowledge with the discussed strategies enhances segmentation quality in female patients and reduces gender bias, particularly in the abdomen region. This research provides a comparative analysis of new encoding strategies and highlights the potential of using AP to achieve fairer segmentation outcomes.