Design of Computer-Aided Intelligent Diagnosis System for Urological Tumors: from MRI Image to PET-CET Image Segmentation Algorithm

In this paper, a new MR urinary system image segmentation based on Kernel Fuzzy c-means clustering algorithm (KFCM) is proposed. The influence of the selection of the membership degree m parameter and the number of clusters J} in the KavCM algorithm on the image segmentation is discussed. Simulation experiments show that the membership function value for MRI brain images is
$2\leqslant \mathrm{m}\leqslant \text{ll}$
integers. Aiming at the characteristics of low resolution and blurred boundary contour of PET images, combined with the needs of clinical diagnosis, this paper uses a bilateral cross filter to construct an edge filter of CT images to filter PET images. The proliferation term in the model is improved to improve tumor prediction. The accuracy reaches 7.6%