Multiphase level set algorithm for coupled segmentation of multiple regions. Application to MRI segmentation

Abstract

Classic geometric active contour algorithms have the limitation of segmenting the image into only two regions: background and object of interest. A new multiphase level set algorithm for the segmentation of two or more regions of interest is proposed. This algorithm avoids by construction the presence of overlapped and void regions and no additional coupling terms are required. In addition, the number of iterations needed to reach convergence is small. The algorithm has been tested against a state-of-the-art multiphase method on both simulated and real Magnetic Resonance Imaging (MRI) volumes with favorable results.

Intelligent Segmentation Algorithm for Diagnosis of Meniere’s Disease in the Inner Auditory Canal Using MRI Images with Three-Dimensional Level Set

“…When X and Y were not independent of each other, the joint entropy was less than the sum of their respective entropy. e above was expressed in (5). e gradient descent method was selected as the optimization algorithm of the registration process, and the objective function was optimized by an iterative method to achieve the best conditions.…”

“…is method can effectively prevent missed detection of image edges and can process images of various types of lesions. However, the accuracy and operability of this method need to be improved [5]. Statistical shape models are the models that employ the high-resolution images as training samples and are compared with the images to be tested, so as to obtain accurate solutions through appropriate registration methods [6].…”

Computational anatomy for multi-organ analysis in medical imaging: A review

“…Alternatively, Merino-Caviedes et al (Merino-Caviedes et al, 2010) and Ma et al (Ma et al, 2013) used the region competition speed term proposed by Brox and Weickert (Brox and Weickert, 2004), to segment the brain and the female pelvic region, respectively. The use of competing forces was also used by Yan et al (Yan et al, 2009), as the strategy to deal with the gaps between subcortical structures.…”

“…GlM► Akhoundi- Asl and Soltanian-Zadeh, 2007;Bossa and Olmos, 2006;Bossa and Olmos, 2007;Bossa et al, 2011;Cootes et al, 1994;Duta and Sonka, 1998;Gorczowski et al, 2010;Poupon et al, 1998;Styner et al, 2006;InM► Akhoundi-Asl and Soltanian-Zadeh, 2007;Asl and Soltanian-Zadeh, 2008;CoDeM► Bazin and Pham, 2006;Fan et al, 2008a;Fan et al, 2008b;Gao et al, 2011;Gao et al, 2017;Han and Prince, 2003;Han et al, 2002;Ho and Shi, 2004;Holtzman-Gazit et al, 2003;Jeon et al, 2005;Kim et al, 2014;Litvin and Karl, 2005;MacDonald et al, 1994;Mangin et al, 1995;Merino-Caviedes et al, 2010;Pohl et al, 2007;Samson et al, 2000;Tsai et al, 2001;Uzunbas et al, 2010;Vese and Chan, 2002;Yan et al, 2009;Yang et al, 2004;Zeng et al, 1999;MuLeM► Cerrolaza et al, 2011;Cerrolaza et al, 2012;Cerrolaza et al, 2014;Cerrolaza et al, 2015;Cerrolaza et al, 2016;Joohwi Leea, Sun Hyung Kimb, 2016;Shen et al, 2001;…”

“…A similar trend is observed in the context of multi-organ analysis, with the brain being the subject of study in 33% of the works reviewed here. The study of the human brain as a modular structure ranged from the simpler gray/white matter division of the brain tissue (Merino-Caviedes et al, 2010;Song et al, 2006;Zeng et al, 1999) to the more detailed partition into dozens of structures (Batrancourt et al, 2006;Heckemann et al, 2006;Lancaster et al, 1997;Milletari et al, 2017). The increasing resolution of brain imaging technology offering detailed and accurate representation of this organ also encouraged the development of new computational approaches to characterize multiple neuroanatomical structures.…”