Computer Aided Diagnosis And Detection Technology Research In Liver Cell Carcinoma Operationg

In the whole world,an average of about 7 million people die each year from cancer,such as hepatocellular carcinoma(HCC).Liver cancer is a very dangerous type of cancer.Multi-temporal CT maps can provide clinicians with some important information about liver cancer.With contrast enhancement,hepatocellular carcinoma is mainly shown as a bright area on the arterial phase,and in the portal phase diagram,the main manifestations are dark areas.In this paper,an automatic diagnosis of hepatocellular carcinoma based on edge detection and differential operation is proposed.According to the scheme,the black area is selected by subtracting the original phase diagram from the corresponding registered phase diagram.Through these black regions,the Sobel and LOG edge detection filters can be used to extract the region without edge into a suspected region of hepatocellular carcinoma.False positive regions can be eliminated by six characteristic values extracted from the cancerous and liver regions.Other false positive can be further eliminated by using the opening operation.Finally,the three-dimensional shape of hepatocellular carcinoma can be obtained by expansion treatment.In our study,CT images of 249 patients with suspected hepatocellular carcinoma were taken.In order to overcome the deviations of the tumor locations in different phases due to different imaging times,we used the image registration method to correct,so that the accuracy of diagnosis reached 73.1%.The experimental results show that our proposed method of computerized detection using the information obtained from different time phase maps to locate cancerous regions is effective.