The Cancer Imaging Archive (TCIA) has a publically available FDA database consisting of just over thousand CT scans intended for facilitating the assessment of lung nodule size estimation methodologies, the development of image analysis software, as well as a wide range of different analyses. The use of these scans would be greatly facilitated by the availability of phantom nodule location data that could be input into methods that require them. This paper outlines a new image processing method to locate the phantom nodules in these CT scans in order to supplant manual location prior to their use. We present a method for extracting the phantom nodules, which involves phantom lung wall removal and separation of the phantom nodules from surrounding phantom blood vessels. Nodule locations are described by rectangular boxes bounding their positions in the scans and volume estimations.
Proceedings Title: Proceedings of the 2012 International Conference on Image Processing, Computer Vision, & Pattern Recognition, IPCV 2012
Conference Dates: July 16-19, 2012
Conference Location: Las Vegas, NV
Pub Type: Conferences
anthropomorphic phantom, image processing, Lung nodules, thoracic CT