Towards Robust Partially Supervised Multi-Structure Medical Image Segmentation on Small-Scale Data

TL;DR

This paper introduces VLUU, a framework that enhances partially supervised medical image segmentation on small datasets by generating vicinal labels, effectively transforming partial labels into a fully supervised setting.

Contribution

VLUU leverages human structure similarity to improve partially supervised segmentation, addressing data scarcity and dataset shift in medical imaging.

Findings

VLUU outperforms previous models on small-scale datasets.

It effectively handles dataset shift and class imbalance.

The approach advances label-efficient deep learning in medical imaging.

Abstract

The data-driven nature of deep learning (DL) models for semantic segmentation requires a large number of pixel-level annotations. However, large-scale and fully labeled medical datasets are often unavailable for practical tasks. Recently, partially supervised methods have been proposed to utilize images with incomplete labels in the medical domain. To bridge the methodological gaps in partially supervised learning (PSL) under data scarcity, we propose Vicinal Labels Under Uncertainty (VLUU), a simple yet efficient framework utilizing the human structure similarity for partially supervised medical image segmentation. Motivated by multi-task learning and vicinal risk minimization, VLUU transforms the partially supervised problem into a fully supervised problem by generating vicinal labels. We systematically evaluate VLUU under the challenges of small-scale data, dataset shift, and class imbalance on two commonly used segmentation datasets for the tasks of chest organ segmentation and optic disc-and-cup segmentation. The experimental results show that VLUU can consistently outperform previous partially supervised models in these settings. Our research suggests a new research direction in label-efficient deep learning with partial supervision.