The Cholangiocarcinoma (CCA) is the primary liver cancer that met worldwide. It cans surveillance by the Periductal fibrosis (PDF) ultrasound images which applied to the CCA surveillance system. The performance of the CCA surveillance system is the PDF regions detection. Several techniques of the image processing methods can be applied in this task. This paper presented the new method for the Computer Aided Detection system for automatic PDF detection (CADPDF) which combined the preprocessing technique of the PDF ultrasound images, edge detection technique of the PDF ultrasound images, and identify the PDF region. Adopting the PDF Ultrasound images database for experimental consist of 30 PDF ultrasound images. Moreover, the performance evaluation of the CADPDF technique used the Sensitivity parameter. The CADPDF technique performance evaluation sensitivity parameter is 82.37%. The CADPDF system has the potential to increase the performance for detection the PDF in ultrasound images for surveillance Cholangiocarcinoma.