Tumor demarcation by VQ based clustering and augmentation with KMCG and KFCG codebook generation algorithms

Ultrasound (US) imaging is important modality to examine the clinical problems and also used as complimentary to the mammogram images to understand nature and shape of the breast tumor. Accurate and efficient segmentation method helps radiologists to understand and observe the volume of a tumor (growth or shrinkage). Inherent artifact present in US images, such as speckle, attenuation and shadows are major hurdles in achieving proper segmentation. Along with the accuracy, computational efficiency is also major concern in the segmentation process. Here, in this paper, VQ based clustering technique is proposed for US image segmentation with KMCG and KFCG as codebook generation algorithms. A novel technique of sequential cluster clubbing is used on clusters obtained from codebook generation algorithms and appropriate cluster has been selected as segmentation result. Besides original KMCG and KFCG, augmented KMCG and KFCG are also proposed for clustering with different block sizes. The results of all proposed methods are compared with each other and best result is selected based on two criteria's, one is computational efficiency and other is accuracy. Finally, best results amongst our methods are compared with results of original watershed and improved watershed transforms.