Segmentation of Tumors from Ultrasound Images with PAORGB
Ms. Maya Eapena, Miss. S. Joy Angel Ancelita, G. Geetha · Procedia Computer Science · 2015
Segmentation is an unavoidable process for computer-aided diagnosis in medical image processing. Due to complex artifacts accurate extraction of tumors in medical images remains a challenge. The existing system deals with the manual cut of the tumor centered image along with robust graph based segmentation for accurate extraction of tumor. The input image can be obtained from any type of modality. There are different ways to segment breast tumor images. This paper focuses on a novel technique called parameter automatically optimized robust graph based (PAORGB) segmentation. Particle swam optimization is incorporated with robust graph based segmentation in order to obtain a global solution. The new method may be a effective way to segment breast tumors than the robust graph based method and to also reduce the computational time.