In our previous study, fat tissues on abdominal CT tended to be concentrated around colon cancer sites, and the intra-abdominal fat area in patients with colon cancer presented a small secondary peak in the pelvis, which we called the sub-peak. We assessed the amount of the sub-peak to develop a novel method for early detection of colon cancer. The profile of intra-abdominal fat data was determined first by segmentation of fat components in CT images, and was approximated by removing the sub-peak part manually by use of six-degree polynominal approximation. The sub-peak value was obtained by subtracting the approximate data from the original data. The amount of the sub-peak was quantified by the sub-peak ratio which is defined by the ratio of the maximum value to the sub-peak value in the profile of intra-abdominal fat data. The subjects consisted of patients with rectal cancer (60 cases) , sigmoid colon cancer (72 cases) , and non-tumors in the pelvic cavity (59 cases) . Our result indicated that the accuracy in detecting colon cancer was 76.1% when the sub-peak ratio was equal to or greater than 10%. This amount of the sub-peak ratio may be useful in the development of a computer-aided diagnosis (CAD) scheme for early detection of colon cancer.
