Dual Meta-Learning with Longitudinally Generalized Regularization for One-Shot Brain Tissue Segmentation Across the Human Lifespan

TL;DR

This paper introduces a dual meta-learning approach for brain tissue segmentation that maintains longitudinal consistency across the human lifespan, improving segmentation accuracy on longitudinal datasets.

Contribution

It proposes a novel dual meta-learning framework with regularizations to enhance longitudinal generalization in brain tissue segmentation tasks.

Findings

Effective on iSeg2019 and ADNI datasets

Outperforms existing methods in longitudinal segmentation accuracy

Maintains consistency across different age groups

Abstract

Brain tissue segmentation is essential for neuroscience and clinical studies. However, segmentation on longitudinal data is challenging due to dynamic brain changes across the lifespan. Previous researches mainly focus on self-supervision with regularizations and will lose longitudinal generalization when fine-tuning on a specific age group. In this paper, we propose a dual meta-learning paradigm to learn longitudinally consistent representations and persist when fine-tuning. Specifically, we learn a plug-and-play feature extractor to extract longitudinal-consistent anatomical representations by meta-feature learning and a well-initialized task head for fine-tuning by meta-initialization learning. Besides, two class-aware regularizations are proposed to encourage longitudinal consistency. Experimental results on the iSeg2019 and ADNI datasets demonstrate the effectiveness of our method. Our code is available at https://github.com/ladderlab-xjtu/DuMeta.