Name
Abstract | ||
|---|---|---|
In radiotherapy planning, multi-organ contouring is the most complex task especially in the part of abdomen, and now this work is primarily conducted manually. Since there exist many aliasing artifacts in abdominal Magnetic Resonance (MR) images, manual contouring is often subject to operator's experience. Besides, manual contouring is sometimes labor-intensive and time-consuming. Thus, it is significant to develop an intelligent auto-contouring method. In our study, auto-contouring is divided into two steps. One is to effectively segment image voxels into several groups, and the other automatically contours target organs according to the group labels. In clinical trials, cases in which a few accurately labeled data together with numerous unlabeled data from target MR images are often met, which limits the applicability of classic supervised algorithms. Laplacian Support Vector Machines (LapSVM), a well-established semi-supervised classification algorithm, is adopted in our study to reliably segment tissue/organ groups and thereby to assist the routine MR guided adaptive radiation therapy (MR-ART). The experimental results verify that our LapSVM-based auto-contouring method achieves satisfied contouring accuracies upon abdominal MR images. |
Year | DOI | Venue |
|---|---|---|
2019 | 10.1166/jmihi.2019.2737 | JOURNAL OF MEDICAL IMAGING AND HEALTH INFORMATICS |
Keywords | DocType | Volume |
Auto-Contouring,Semi-Supervised Learning,Laplacian Support Vector Machines | Journal | 9 |
Issue | ISSN | Citations |
7 | 2156-7018 | 0 |
PageRank | References | Authors |
0.34 | 0 | 7 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Leyuan Zhou | 1 | 7 | 3.13 |
| Kaifa Zhao | 2 | 19 | 3.65 |
| Yang Ding | 3 | 5 | 1.40 |
| Jiamin Zheng | 4 | 2 | 2.05 |
| Yangyang Chen | 5 | 0 | 0.68 |
| Yizhang Jiang | 6 | 2 | 3.73 |
| Pengjiang Qian | 7 | 133 | 11.25 |