Square-Cut: A Segmentation Algorithm on the Basis of a Rectangle Shape

TL;DR

This paper introduces a rectangle-based graph cut segmentation algorithm that improves object shape preference, demonstrated on MRI vertebrae images with high accuracy, addressing limitations of uniform node distribution.

Contribution

The novel sampling strategy based on rectangle shape enhances segmentation accuracy in challenging MRI images compared to traditional uniform sampling methods.

Findings

Achieved an average Dice Similarity Coefficient of 90.97%.

Effectively segmented vertebrae in MRI images with indistinguishable regions.

Outperformed traditional graph cut methods in shape preference.

Abstract

We present a rectangle-based segmentation algorithm that sets up a graph and performs a graph cut to separate an object from the background. However, graph-based algorithms distribute the graph's nodes uniformly and equidistantly on the image. Then, a smoothness term is added to force the cut to prefer a particular shape. This strategy does not allow the cut to prefer a certain structure, especially when areas of the object are indistinguishable from the background. We solve this problem by referring to a rectangle shape of the object when sampling the graph nodes, i.e., the nodes are distributed nonuniformly and non-equidistantly on the image. This strategy can be useful, when areas of the object are indistinguishable from the background. For evaluation, we focus on vertebrae images from Magnetic Resonance Imaging (MRI) datasets to support the time consuming manual slice-by-slice segmentation performed by physicians. The ground truth of the vertebrae boundaries were manually extracted by two clinical experts (neurological surgeons) with several years of experience in spine surgery and afterwards compared with the automatic segmentation results of the proposed scheme yielding an average Dice Similarity Coefficient (DSC) of 90.97\pm62.2%.