Removal of Unwanted Object in 3D CT Kidney Stone Images and 3D Visualization

摘要

For the study of anatomical structure and image processing of CT images, noise removal techniques have become a vital role in a medical imaging application. In this paper, an automatic unwanted object removing for kidney stone detection is proposed together with 3D visualization. For the removal of surrounding unwanted objects, there are three steps in this proposed scheme. The first step is hypodense and isodense region removing using intensity-based thresholding. In the second step, size-based thresholding is used to remove the bones of the abdomen. In the third step, geometric feature-based thresholding is developed for false-positive reducing. The proposed scheme can effectively provide the structural information of the kidney stone CT image when removing surrounding unwanted objects. It gives the performance with 95.7% in sensitivity. To represent better visual results, simulation experiment results are shown using 3D visualization.