Segmentation of Liver using Abdominal CT Scan to Detection Liver Desease Area
Faizatul Himmah, Riyanto Sigit, Tri Harsono · 2018
Liver disease is a disease that can reduce liver function and affect the production process of protein, hormone and nutrients in the human body. One way of knowing liver disease by doing abdominal CT scan. CT Scan (Computerized Axial Tomography scan) can produce images of organs that cannot be seen by the standard x-ray photo equipment and the resulting image has good and accurate resolution. However, the problems found in Abdominal CT Scan are not able to ensure the image of the heart properly. This is because Abdominal CT Scan has a weakness that there are images that should not participate recorded, so the results cannot be used as a reference. From the problem, an idea came up to detect liver disease using Abdominal CT Scan automatically. Watershed transform algorithm used in the segmentation process to produces liver locations that can distinguish objects by background. Then the image will be segmented again using binary threshold method to separate the liver image as the observed object. The final step is done by doing a calculation to determine the affected area percentage. The output of this paper is a wide percentage of the area of liver disease that can be useful as a radiology doctor's analysis. From the result, the wide segmentation of the liver has an average accuracy of 81.15% and segmentation of disease has an average accuracy of 98.28%. So it can be concluded that the watershed method can be used for segmentation process on CT scan abdominal image.