Fast and Automatic Segmentation of Anatomical Chest Structures from CT Images Based on Image-Guided Lung Biopsy Surgery

摘要

The segmentation of chest organs from CT images has a great significance for finding the optimal surgery path in the image-guided lung biopsy. In this paper, we proposed a fast and fully automated multi-organ segmentation method. It is used to segment and recognize the different chest internal organ and tissue regions for the lung biopsy base on the CT images, including skin, chest, lung, trachea, bronchus, blood vessels and bones. 35-datasets of patients' CT images had been used for-evaluation, and 32 datasets-were viable for this method The average processing time was 150 seconds in one dataset and the average coincidence degree for all organs is 92.20%. Experimental results show that our segmentation method is fast and accurate enough to meet the requirement of the image-guided lung biopsy surgery.