Organ Segmentation Of Pediatric Computed Tomography (CT) With Generative Adversarial Networks
Chi Nok Enoch Kan
Accurately segmenting organs in abdominal computed tomography (CT) is crucial for many clinical applications such as organ-specific dose estimation. With the recent emergence of deep learning techniques for computer vision, many powerful frameworks are proposed for organ segmentation in abdominal CT images. A major problem with these state-of-the-art methods is that they depend on large amounts of training data to achieve high segmentation accuracy. Pediatric abdominal CTs are particularly hard to obtain since these children are much more sensitive to ionizing radiation than adults. It is extremely challenging to train automatic segmentation algorithms on pediatric CT volumes. To address these issues, we propose 2 new GAN architectures for abdominal CT synthesis and a combined segmentation-synthesis network with a built-in auxiliary classifier generative adversarial network (ACGAN) that conditionally generates additional features during training. All 3 frameworks are tested on a pediatric abdominal CT dataset collected by the Medical College of Wisconsin. Both of our proposed GAN architectures can generate quantitatively and qualitatively realistic abdominal CT images and patches. 2.5D segmentation experiments with 4-fold cross validation confirms our proposed segmentation framework, CFG-SegNet, is indeed high- performing and able precisely segment reproductive organs in abdominal CTs across multiple patient ages.
Read Full Article Here: https://epublications.marquette.edu/theses_open/631/