Deep Learning Based Instance Segmentation in 3D Biomedical Images Using Weak Annotation

TL;DR

This paper introduces a weak annotation method for 3D biomedical image instance segmentation that reduces annotation effort while maintaining high performance, using only bounding boxes and limited voxel masks.

Contribution

A novel two-stage 3D segmentation model that leverages minimal annotation data, significantly reducing labeling costs compared to full voxel annotation methods.

Findings

Achieves comparable performance to fully annotated models with less annotation effort.

Outperforms fully annotated methods in similar annotation time.

Effective on multiple biomedical datasets.

Abstract

Instance segmentation in 3D images is a fundamental task in biomedical image analysis. While deep learning models often work well for 2D instance segmentation, 3D instance segmentation still faces critical challenges, such as insufficient training data due to various annotation difficulties in 3D biomedical images. Common 3D annotation methods (e.g., full voxel annotation) incur high workloads and costs for labeling enough instances for training deep learning 3D instance segmentation models. In this paper, we propose a new weak annotation approach for training a fast deep learning 3D instance segmentation model without using full voxel mask annotation. Our approach needs only 3D bounding boxes for all instances and full voxel annotation for a small fraction of the instances, and uses a novel two-stage 3D instance segmentation model utilizing these two kinds of annotation, respectively. We evaluate our approach on several biomedical image datasets, and the experimental results show that (1) with full annotated boxes and a small amount of masks, our approach can achieve similar performance as the best known methods using full annotation, and (2) with similar annotation time, our approach outperforms the best known methods that use full annotation.