Abstract
Recently, optical device such as CT, MRI, CR, and some other devices are developed in medical field. These devices provide us advantages that filming time decrease and image resolution is higher than past system. Simultaneously, exposure time is shorter than before. But slices which medical doctor should diagnosis are increase. By this reason, their burden is huge and oversight of abnormal area may be caused. On the other hand, segmentation for region of interest is important task to extract abnormal area correctly. In this paper, we propose a method for segmentation and extracting abnormal area on obtained lung region employing thorax Multi-detector raw CT images. Furthermore, extracted abnormal areas are marked and display on PC. We applied our technique onto 12 real MDCT images and satisfactory results are obtained. To analyze the experimental results, we evaluate our systems based on ROC analysis.
