Review of brain MRI image segmentation methods

Abstract

Brain image segmentation is one of the most important parts of clinical diagnostic tools. Brain images mostly contain noise, inhomogeneity and sometimes deviation. Therefore, accurate segmentation of brain images is a very difficult task. However, the process of accurate segmentation of these images is very important and crucial for a correct diagnosis by clinical tools. We presented a review of the methods used in brain segmentation. The review covers imaging modalities, magnetic resonance imaging and methods for noise reduction, inhomogeneity correction and segmentation. We conclude with a discussion on the trend of future research in brain segmentation.

Skull Segmentation in MRI by a Support Vector Machine Combining Local and Global Features

“…In contrast, great effort has been devoted to skull stripping, the process of segmenting brain from nonbrain tissue without further considering the composition of the non-brain tissue [3], as well as the subsequent segmentation of the brain into different types of soft tissue [4]. What sets skull segmentation apart from soft tissue segmentation and skull stripping is the need to distinguish between bone and air, neither of which provides a signal in the MR sequences used routinely.…”

Automatic brain segmentation using fractional signal modeling of a multiple flip angle, spoiled gradient-recalled echo acquisition

“…For example, tissue separation may be critical for complete assessment of the function in a given structure [1]. Separation of different compartments can also be utilized in the actual calculation of parametric maps [2], while, in other applications, it may be important to know the volume of the structures of interest [3].…”

“…These methods generally employ multiple mathematically advanced algorithms, which make them difficult to implement. The wide variety of existing segmentation methods has been thoroughly reviewed in the literature [1,6].…”

Brain MR Segmentation through Fuzzy Expectation Maximization and Histogram Based K-Means

“…An accurate and robust tissue classification is the basis for many applications such as: quantitative measurement of tissue volume in normal and diseased brain, morphological analysis, or visualization. Manual or even semi-automatic, Classification performed by a trained expert is labor-intensive and hence impractical for processing large amounts of data, highly subjective, and non-reproducible [1]. The main difficulties found in the automatic segmentation of MR brain images are due to the fact that image intensities are not necessarily constant for each tissue class.…”