Prostate tumor segmentation for gamma image using region growing approaches

Prostate is a gland of male reproductive system to store semen. The prostate cancer is prevalent among the male which may cause mortality. It is usually unpredictable in the clinical course as the prostate cancer mostly slows grow and do not manifest in the early stage. Recent imaging technique is usually focused on the local or regional imaging so that the tumor can be more precisely identified. The measurement of tumor size can be used to inspect the progress of the severity. Gamma imaging that employs the radiotracer is widely used in the imaging of prostate cancer. However, the imaging technique is still unable to show clearly the edge of the tumor where it may cause wrong diagnosis and wrong measurement of the tumor size. Therefore, in order to increase the image quality, Gabor filter is used to reduce the noise of the image and to smooth the image. Segmentation with region growing method will be implemented to subdivide the image into the region of interest (tumor) to facilitate the radiologist in identifying and measuring the tumor size to make a more precise decision in provision of appropriate therapy. This technique is verified by five other images with prostate cancer from different modalities in radiology. The results show that the tumor can be accurately partitioned alone from the surrounding normal tissues by varying the intensity range. However, there are some cases cannot really isolate the tumor alone but it still can show clearly the tumor shape and edge. Hence, it can be concluded that this technique is valid to be applied in the clinical field to assist in the interpretation process. Key-Words: Prostate cancer, gamma imaging, Gabor filter, region growing segmentation