Enforcing connectivity of 3D linear structures using their 2D projections

TL;DR

This paper introduces a method to enhance the connectivity of 3D curvilinear structures in medical imaging by applying topology-aware loss functions on their 2D projections, improving accuracy and reducing annotation effort.

Contribution

It proposes a novel approach that enforces 3D structure connectivity through 2D projection-based topology-aware losses, addressing limitations of voxel-wise training methods.

Findings

Improved 3D connectivity in neural network outputs.

Reduced annotation effort for training data.

Enhanced accuracy of 3D structure delineation.

Abstract

Many biological and medical tasks require the delineation of 3D curvilinear structures such as blood vessels and neurites from image volumes. This is typically done using neural networks trained by minimizing voxel-wise loss functions that do not capture the topological properties of these structures. As a result, the connectivity of the recovered structures is often wrong, which lessens their usefulness. In this paper, we propose to improve the 3D connectivity of our results by minimizing a sum of topology-aware losses on their 2D projections. This suffices to increase the accuracy and to reduce the annotation effort required to provide the required annotated training data. Code is available at https://github.com/doruk-oner/ConnectivityOnProjections.