Detection of Hepatolithiasis in CT Images for Image-Guided Surgery

摘要

Hepatolithiasis is endemic in the Asia-Pacific region, where its prevalence can be as high as 30-50%. Complete cure of hepatolithiasis requires both extraction of the stones and elimination of bile stasis through surgery, and therefore it is essential to detect hepatolithiasis at an earlier stage. The easiest way to diagnose hepatolithiasis is by Computed Tomography (CT) images. The detection/segmentation of hepatolithiasis from CT images is an important prerequisite for computer-aided diagnosis, treatment planning, and monitoring of hepatolithiasis. In this paper, a novel and efficient method is proposed for the easy detection of this deadly disease in CT volumes based on the well-known segmentation methods and morphological operations. The detection process involves the pre-processing of input images, image segmentation and morphological operations which helps to identify the presence of hepatolithiasis which in this case is the presence of small stones. The proposed method is validated on sixteen liver CT images provided by the First Affiliated Hospital of Guangzhou Medical University. A sensitivity of 100% and an accuracy of 95% were obtained.