Synthetic CT images for semi-sequential detection and segmentation of lung nodules
- PDF / 7,428,600 Bytes
- 13 Pages / 595.224 x 790.955 pts Page_size
- 61 Downloads / 170 Views
Synthetic CT images for semi-sequential detection and segmentation of lung nodules Mohammad Hesam Hesamian1
· Wenjing Jia1 · Xiangjian He1 · Qingqing Wang1 · Paul J. Kennedy2
© Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Accurately detecting and segmenting lung nodules from CT images play a critical role in the earlier diagnosis of lung cancer and thus have attracted much interest from the research community. However, due to the irregular shapes of nodules, and the low-intensity contrast between the nodules and other lung areas, precisely segmenting nodules from lung CT images is a very challenging task. In this paper, we propose a highly effective and robust solution to this problem by innovatively utilizing the changes of nodule shapes over continuous slices (inter-slice changes) and develop a deep learning based endto-end system. Different from the existing 2.5D or 3D methods that attempt to explore the inter-slice features, we propose to create a novel synthetic image to depict the unique changing pattern of nodules between slices in distinctive colour patterns. Based on the new synthetic images, we then adopt the deep learning based image segmentation techniques and develop a modified U-Net architecture to learn the unique color patterns formed by nodules. With our proposed approach, the detection and segmentation of nodules can be achieved simultaneously with an accuracy significantly higher than the state of the arts by 10% without introducing high computation cost. By taking advantage of inter-slice information and form the proposed synthetic image, the task of lung nodule segmentation is done more accurately and effectively. Keywords Nodule segmentation · Deep learning · Synthetic CT images · Semi-sequential segmentation
1 Introduction Lung cancer is recognized as one of the most common types of cancer, yet by far the deadliest [29]. Early detection of malignant nodules can improve the effectiveness of treatment and help to increase the survival rate of patients by 50% to 65% [16]. The boundaries of lung nodules are considered as the vital criteria for cancer diagnosis and analysis. Therefore, accurate segmentation of lung nodules from CT images provides critical information about the location, size and shape of the nodules. This information is critical for early diagnosis, therapy selection and screening of the patient response to therapy. Traditionally, lung and lung nodules are delineated by radiologists on a slice-by-slice basis, which is time Mohammad Hesam Hesamian
[email protected] 1
School of Electrical and Data Engineering, University of Technology Sydney, Sydney, Australia
2
School of Computer Science, Australian Artificial Intelligence Institute, University of Technology Sydney, Sydney, Australia
consuming and prone to inter- and intra-rater variations. The considerable number of radiologic errors that occur, highlight the high demand in the clinical practice of developing a computer-aided lung nodule segmentation technique. This could help the radio
Data Loading...
