Segmentation of low-grade gliomas based on the growing region and level sets techniques
Zaouche Rahima, Ahror Belaid, Basel Solaiman, D. Ben Salem, Souhil Tliba · 2018
In this paper, we propose a novel semi-automatic segmentation method based on the local image properties. Its originality is twofold, the first stands on the intensity invariant of phase-local information for the purpose of low-grade gliomas segmentation in MR images. In a second time, a level set method driven is combined to growing region so as to improve tumor detection. Experiments were conducted on a set of medical images. A comparison between the obtained results and the manual segmentation collected from experts is performed. The preliminary results are interesting and encouraging.