Lesion Demarcation of CT-Scan Images using Image Processing Technique

The Computed Tomography (CT) scan image is the most reliable method and mostly used in radiotherapy as it is of low cost yet efficient compared to the Magnetic Resonance Imaging (MRI) scan. The characteristics of brain tumor cases in the CT scan image are explored to obtain useful information according to their grey level intensity and Hounsfield Unit (HU). Therefore, the rapid development of the image processing technique has been widely proposed to explore more to further reveal the information of the CT scan grey level intensity. This study proposes a computed technique in processing the CT scan images including the segmentation, enhancement and filtering processes. The clustering algorithm of Fuzzy C-Mean (FCM) is used to cluster the CT scan images into 3 regions. The FCM is used to separate the CT scan images into three parts namely the bone, brain and tumor regions. Then, the intensity of the regions obtained is measured and compared with HU. The result shows that there is a strong correlation between the grey level intensity value and HU. This computer-aided technique hopefully could help the radiologists to perform a better diagnose on the brain tumor images.