Brain Tumor Segmentation Based on Minimum Spanning Tree

Abstract

In this paper, we propose a minimum spanning tree-based method for segmenting brain tumors. The proposed method performs interactive segmentation based on the minimum spanning tree without tuning parameters. The steps involve preprocessing, making a graph, constructing a minimum spanning tree, and a newly implemented way of interactively segmenting the region of interest. In the preprocessing step, a Gaussian filter is applied to 2D images to remove the noise. Then, the pixel neighbor graph is weighted by intensity differences and the corresponding minimum spanning tree is constructed. The image is loaded in an interactive window for segmenting the tumor. The region of interest and the background are selected by clicking to split the minimum spanning tree into two trees. One of these trees represents the region of interest and the other represents the background. Finally, the segmentation given by the two trees is visualized. The proposed method was tested by segmenting two different 2D brain T1-weighted magnetic resonance image data sets. The comparison between our results and the gold standard segmentation confirmed the validity of the minimum spanning tree approach. The proposed method is simple to implement and the results indicate that it is accurate and efficient.

Brain Tumor Segmentation Based on α‐Expansion Graph Cut

“…The specificity in our work compared with the work stated in [21] is nearly the same. We believe that due to the usage of the same dataset used by the authors, our work is optimal and the comparison is more logical, in addition, in some cases the Jaccard indices are low, but still more than half.…”

“…In recent studies, Mayala et al [21] proposed an interactive segmentation technique based on the minimum spanning tree (MST) algorithm. The method involves a pre-processing step followed by the construction of the MST and an interactive process for determining the background and region of interest (ROI) to complete the segmentation.…”

GUBS: Graph-Based Unsupervised Brain Segmentation in MRI Images

“…The adjacency graph size is defined by the number of voxels in the MRI image, corresponding to vertices in the adjacency graph. The experimental timing for the adjacency graph and MST construction is efficient and presented in [ 36 ]. For the adjacency graph constructed from MRI image 134 × 134 × 134 (with 2,406,104 nodes in the adjacency graph), we update the MST twice to disconnect a path connecting the brain and the non-brain tissues as well as disconnect the path connecting non-brain tissues and the background.…”

“…We extended the segmentation criteria used in the paper [ 36 ] by collapsing subgraphs of the voxels adjacency graph before constructing the MST and presented the GUBS method for segmenting the brain from MRI images. GUBS works by representing MRI volume into an adjacency graph, followed by collapsing representative nodes sampled from the brain, no-brain, and background regions, respectively.…”