Abstract—The purpose of this research work is to segment Human Lung CT Scan images for early detection of cancer. As a part of this work combination of ‘Region growing’ and ‘Watershed Technique’ are implemented as the ‘Segmentation’ method. These methods are based on the filters available in the ‘Insight Segmentation and Registration Toolkit’ (ITK). Image segmentation plays a vital role in several medical imaging programs by automating or assisting the delineation of physiological structures along with other parts of interest. NIH/NCI Lung Image Database Consortium (LIDC) served as the repository of lung images in the DICOM format. In this short article, different methods that can be used for efficient visualization as well as automatically extracting the organ regions from abdominal CT (Computerized tomography) data especially from lung that can be further used in various medical diagnosis applications like CBMIR (Content-based medical image retrieval) have been suggested. The achieved results support the use of the proposed algorithm for extraction of lung nodules from the image.

Index Terms—CAD, DICOM, CBMIR, ITK, MATITK,computerized tomography, LIDC

The authors is in CEC Landran, Punjab (e- mail: shubh_86@ymail.com)

Cite: Shubhpreet Kaur and Gagandeep Jindal, "Watershed Segmentation of Lung CT Scan Images for Early Diagnosis of Cancer," International Journal of Computer and Electrical Engineering vol. 3, no. 6, pp. 850-852, 2011.