Cell Segmentation in 3D Confocal Images using Supervoxel Merge-Forests with CNN-based Hypothesis Selection

TL;DR

This paper introduces a novel 3D cell segmentation method using supervoxels and CNN-based hypothesis selection, significantly improving accuracy and reducing manual correction in large-scale confocal microscopy images.

Contribution

The proposed approach combines supervoxel merging with CNN-based hypothesis selection, outperforming existing methods in 3D cell segmentation accuracy.

Findings

Outperforms current segmentation methods in accuracy

Reduces manual correction effort

Validated on Arabidopsis confocal images

Abstract

Automated segmentation approaches are crucial to quantitatively analyze large-scale 3D microscopy images. Particularly in deep tissue regions, automatic methods still fail to provide error-free segmentations. To improve the segmentation quality throughout imaged samples, we present a new supervoxel-based 3D segmentation approach that outperforms current methods and reduces the manual correction effort. The algorithm consists of gentle preprocessing and a conservative super-voxel generation method followed by supervoxel agglomeration based on local signal properties and a postprocessing step to fix under-segmentation errors using a Convolutional Neural Network. We validate the functionality of the algorithm on manually labeled 3D confocal images of the plant Arabidopis thaliana and compare the results to a state-of-the-art meristem segmentation algorithm.