Detection of Diaphragmatic Rupture using Computed Tomography Images

• Main Authors: Yi-shan Tu, 杜怡珊
• Other Authors: Ruey-feng Chang
• Format: Others
• Language: en_US
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/60630666095198382830

碩士 === 國立中正大學 === 資訊工程所 === 96 === There is a high possibility that injuries to the diaphragm occur after penetrating or blunt trauma. With the passing of the time, there will be a progressive damage. Complex diaphragmatic rupture will lead to higher morbidity and mortality, thus early diagnosis and treatment is extremely important. In most cases, diaphragmatic injuries are not seen in isolation but with some other concurrent injuries. Thus doctors often pay more attention on coronal view of the computed tomography images and miss the examination of the diaphragm. In this paper, a detection method is developed to examine the diaphragmatic rupture using the computed tomography images. In this diaphragm detection system, the image resolution is reduced to decrease the computation time of detections. Then, two kinds of detections are used to detect different regions in computed tomography images. One is lung air boundary detection; the feature that the intensities of liquid and organs are different to the intensities of air is utilized to detect the border between air and liquid. The principle that the boundary should be continued is used to improve the correctness of the detected result. The other is the diaphragm detection; the diaphragm dome has different intensity with other tissues on coronal view of the computed tomography images, and the Sobel operator is used to enhance the diaphragm dome. The characteristic of diaphragm shape is also used as a guided rule for the diaphragm detection. Finally, the results of lung air boundary detection and diaphragm detection are compared to find out whether the diaphragm is injured or not. In our experiments, there are total 18 cases examined, the detection accuracy rate is up to 95%. The results show that the proposed system almost could detect the diaphragmatic rupture out in the images with a good accuracy.