Watershed-Based Segmentation of the Midsagittal Section of the Corpus Callosum in Diffusion MRI

Abstract

Fig. 1. The main steps of the corpus callosum segmentation: (a) the diffusion tensor image, (b) the fractional anisotropy map, (c) the watershed transform and (d) the final segmentation resultAbstract-The corpus callosum (CC) is one of the most important white matter structures of the brain, interconnecting the two cerebral hemispheres. The corpus callosum is related to several neurodegenerative diseases and, as segmentation is usually the first step for studies in this structure, it is important to have a robust method for CC segmentation. We propose here a new approach for fully automatic segmentation of the CC in the magnetic resonance diffusion tensor images. The method uses the watershed transform and is performed on the fractional anisotropy (FA) map weighted by the projection of the principal eigenvector in the left-right direction. It first computes the section of the CC in the midsagittal slice and uses it as a seed for the 3D volume segmentation. Experiments with real diffusion MRI data showed that the proposed method is able to quickly segment the CC without any user intervention, with great results when compared to manual segmentation. Since it is simple, fast and does not require parameter settings, the proposed method is well suited for clinical applications.

Automatic DTI-based parcellation of the corpus callosum through the watershed transform

“…The automatic segmentation of the CC in the midsagittal slice, presented in previous work (Freitas et al, 2011) includes the automatic determination of the midsagittal slice of the brain, the weighting of the FA map, the computation of the morphological gradient and the watershed transform segmentation.…”

“…A threshold T = 0.2 of the weighted FA average computed for each region is used to classify the regions that form the CC. The sensibility of the proposed method to the variation of both parameters (n and T) were accessed and discussed in previous work (Freitas et al, 2011).…”

Can incomplete silver standard labels improve performance of DTI-based volumetric segmentation of the corpus callosum?

“…Among the existing semi-automated and automated segmentation methods working directly on DTI, 15 most of them segment the CC on the midsagittal slice such as the ones proposed by Freitas et al, 21 Niogi et al 22 Nazem-Zadeh et al, 23 Luis-Garcıa et al, 24 Cover et al 25 and Herrera et al 26 For volumetric CC segmentation on DTI, there are only five automated methods: Rittner et al 27 extended Freitas et al' 28 approach and achieved a hierarchical segmentation by applying the 3D watershed using automatically selected markers; Kong et al 29 proposed a method to automatically learn an adaptive distance metric, through a semi-supervised learning model based on graphs; Wang et al 30 introduced the gray information of multiple atlases and the prior information of target shapes into the Active Shape Model (ASM) and proposed the Multi-Atlas Active Shape Model (MA-ASM) approach for segmentation of 7 Region of Interest (ROI), including the CC; Rodrigues et al 1 developed a volumetric segmentation method for the CC using Convolutional Neural Networks (CNN) using as input diffusion maps, such as fractional anisotropy (FA), mean diffusivity (MD) and mode of anisotropy (MO), separately and combined; Rodrigues et al 31 developed a subsequent method based on 3D convolutions, presenting better performance in comparison to the 2D model.…”

Universal and stable medical image generation for tissue segmentation \newline(The unistable method)

“…A bias field correction algorithm was used as a presegmentation step for MRI data to avoid wrong initial parameter estimation through intensity normalization. Freitas et al [24] proposed an approach for segmentation of the midsagittal section of the corpus callosum using the watershed transform. The results were compared with manual segmentation.…”

“…Region-based segmentation rely are used on clustering methods such as Otsu [20], K-means [21], expectation maximization (EM) [22], C-means and fuzzy C-means (FCM) [23], Markov random field, and iterated conditional modes (ICM) [4,5,18,19]. It should be taken into consideration that there is no unique solution for the segmentation problem where different results are produced by changing the clustering method and/or the selected numbers of clusters [2,4,5,13,14,16,19,[23][24][25][26][27]. The human brain is taken as a case study for this research.…”