Bridging 3D Deep Learning and Curation for Analysis and High-Quality Segmentation in Practice

TL;DR

VessQC is an open-source tool that uses uncertainty maps to guide manual curation of 3D microscopy segmentations, significantly improving error detection while maintaining efficiency in real-world bioimage analysis.

Contribution

The paper introduces VessQC, a novel tool that integrates uncertainty-guided visualization for efficient human-in-the-loop correction of 3D microscopy segmentations.

Findings

Uncertainty-guided correction increased error detection recall from 67% to 94%.

VessQC enables efficient refinement of volumetric segmentations.

The tool is freely available for the bioimaging community.

Abstract

Accurate 3D microscopy image segmentation is critical for quantitative bioimage analysis but even state-of-the-art foundation models yield error-prone results. Therefore, manual curation is still widely used for either preparing high-quality training data or fixing errors before analysis. We present VessQC, an open-source tool for uncertainty-guided curation of large 3D microscopy segmentations. By integrating uncertainty maps, VessQC directs user attention to regions most likely containing biologically meaningful errors. In a preliminary user study uncertainty-guided correction significantly improved error detection recall from 67% to 94.0% (p=0.007) without a significant increase in total curation time. VessQC thus enables efficient, human-in-the-loop refinement of volumetric segmentations and bridges a key gap in real-world applications between uncertainty estimation and practical human-computer interaction. The software is freely available at github.com/MMV-Lab/VessQC.