Multi atlas based segmentations: a comparison between binary and probabilistic methods
Tijl A. van der Velden · 2012
Manually segmenting medical images is a time consuming process. Automating this process is therefore desired. This can be done by using the information from previously segmented images, atlases. Multiple methods have been proposed to perform this task. This paper reviews different approaches and compares them with an emphasis on the relation between probabilistic and binary methods. In this paper, the articles listed in table 1 are described and analyzed. Section 2 gives a short introduction to atlas based segmentation. In section 3 and 4 the different methods are described. In section 3 the imaged based methods are described with a subdivision between the binary and probabilistic methods. The same subdivision is made in section 4, that describes the segmentation based methods. The different methods are compared in section 5 and a conclusion can be found in section 6.