Segmentation And Classification Of Tumor Cell From Fused Images

Abstract

Multi-modality imaging technologies have been routinely used in the clinical practice nowadays. Information fusion of multi-modality medical images can reduce randomness and redundancy, and has been proved to be useful for medical diagnosis, analysis, treatment and outcome assessment. A restoration process is further integrated into the co-segmentation process to handle the uncertainty introduced by the blurred tumor edges in the MRI image. The new information fusion strategy can automatically decide which modality should be more trustful forlocalizing the tumor boundary, in accord to the medical knowledge the images conveyed. In this
 proposed system, two input images are given namely CT images and MRI images. The QWT is one of the effective multi scale image fusion method. Active Contour segmentation is designed in the proposed area. Here the threshold required for segmenting adjusts itself according to the segmented area and position. The trained data are then used to reconstruct the fused image to reduce the noise. The deep neural networks are used to train the input medical images for detecting the tumor whether it is benign or malignant.