Calculation of phantom volume for computed tomography (CT) scan images

Abstract

CT scan is a radiological examination that uses X-ray to produce cross-sectional images of an object. Its aim is to determine abnormalities such as cancers in human organs using ionizing radiation X-ray. Cancer is a term that refers to abnormal cell growth. Currently there are more than 100 types of cancer. Cancer cells can form any body tissue and continue to grow uncontrolled. Cancer cells are derived from normal body cells' damaged DNA (deoxyribonucleic acid), a cellular material that controls the characteristics and growth of cells. Approximately 14.1 million new cancer patients are diagnosed every year and about 8.2 million of them die it. The top five cancers in men are lung, prostate, colon, stomach and liver cancers, whereas in women, there are breast, colon, cervix, lung and uterus cancers. On the other hand, available cancer treatment modalities include surgery, chemotherapy, radiation therapy, hormonal therapy, immune therapy and stem cell transplant. Cancer treatment options depend on the type, stage of cancer, patient's physical condition and preferences. In general, different treatment options come with different results and side effects. Recently, there has been rapid development in cancer therapy modalities and a decrease in cancer related mortality. Calculation of volume changes in cancer treatment processes is very important to know the success level of the therapy. Therefore, cancer volume calculations prior and after treatment are important. This research developed a method of image processing to calculate phantom volume. The phantom material which is used is polymethyl methacrylate (PMMA). This phantom is assumed as cancer cell in patient's body that the volume will be calculated. Image processing and area calculation were conducted on each phantom image slice by thresholding and trapezoidal integration method. Then phantom volume was calculated by integrating all areas with slice thickness. These calculation results were then compared with those from manual calculation. This yielded an error value of 3.63%.