Keypoint Transfer Segmentation
We present an image segmentation method that transfers label maps of entire organs from the training images to the novel image to be segmented. The transfer is based on sparse correspondences between keypoints that represent automatically identified distinctive image locations. Our segmentation algo...
| Main Authors: | , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | en_US |
| Published: |
2017
|
| Online Access: | http://hdl.handle.net/1721.1/110912 https://orcid.org/0000-0002-3652-1874 https://orcid.org/0000-0003-2516-731X |
| _version_ | 1826192577712357376 |
|---|---|
| author | Toews, M. Wachinger, Christian Langs, Georg Wells, William M Golland, Polina |
| author2 | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory |
| author_facet | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Toews, M. Wachinger, Christian Langs, Georg Wells, William M Golland, Polina |
| author_sort | Toews, M. |
| collection | MIT |
| description | We present an image segmentation method that transfers label maps of entire organs from the training images to the novel image to be segmented. The transfer is based on sparse correspondences between keypoints that represent automatically identified distinctive image locations. Our segmentation algorithm consists of three steps: (i) keypoint matching, (ii) voting-based keypoint labeling, and (iii) keypoint-based probabilistic transfer of organ label maps. We introduce generative models for the inference of keypoint labels and for image segmentation, where keypoint matches are treated as a latent random variable and are marginalized out as part of the algorithm. We report segmentation results for abdominal organs in whole-body CT and in contrast-enhanced CT images. The accuracy of our method compares favorably to common multi-atlas segmentation while offering a speed-up of about three orders of magnitude. Furthermore, keypoint transfer requires no training phase or registration to an atlas. The algorithm’s robustness enables the segmentation of scans with highly variable field-of-view. |
| first_indexed | 2024-09-23T09:21:09Z |
| format | Article |
| id | mit-1721.1/110912 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T09:21:09Z |
| publishDate | 2017 |
| record_format | dspace |
| spelling | mit-1721.1/1109122022-09-26T11:19:10Z Keypoint Transfer Segmentation Toews, M. Wachinger, Christian Langs, Georg Wells, William M Golland, Polina Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Wachinger, Christian Langs, Georg Wells, William M Golland, Polina We present an image segmentation method that transfers label maps of entire organs from the training images to the novel image to be segmented. The transfer is based on sparse correspondences between keypoints that represent automatically identified distinctive image locations. Our segmentation algorithm consists of three steps: (i) keypoint matching, (ii) voting-based keypoint labeling, and (iii) keypoint-based probabilistic transfer of organ label maps. We introduce generative models for the inference of keypoint labels and for image segmentation, where keypoint matches are treated as a latent random variable and are marginalized out as part of the algorithm. We report segmentation results for abdominal organs in whole-body CT and in contrast-enhanced CT images. The accuracy of our method compares favorably to common multi-atlas segmentation while offering a speed-up of about three orders of magnitude. Furthermore, keypoint transfer requires no training phase or registration to an atlas. The algorithm’s robustness enables the segmentation of scans with highly variable field-of-view. National Alliance for Medical Image Computing (U.S.) (U54-EB005149) National Center for Image Guided Therapy (P41-EB015898) 2017-08-02T16:05:49Z 2017-08-02T16:05:49Z 2015-06 2017-08-02 2015-06 Article http://purl.org/eprint/type/ConferencePaper 978-3-319-19991-7 978-3-319-19992-4 0302-9743 1611-3349 http://hdl.handle.net/1721.1/110912 Wachinger, C.; Toews, M.; Langs, G.; Wells, W. et al. “Keypoint Transfer Segmentation.” Information Processing in Medical Imaging (2015): 233–245 © Springer International Publishing Switzerland 2015 https://orcid.org/0000-0002-3652-1874 https://orcid.org/0000-0003-2516-731X en_US http://dx.doi.org/10.1007/978-3-319-19992-4_18 Information Processing in Medical Imaging Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf PMC |
| spellingShingle | Toews, M. Wachinger, Christian Langs, Georg Wells, William M Golland, Polina Keypoint Transfer Segmentation |
| title | Keypoint Transfer Segmentation |
| title_full | Keypoint Transfer Segmentation |
| title_fullStr | Keypoint Transfer Segmentation |
| title_full_unstemmed | Keypoint Transfer Segmentation |
| title_short | Keypoint Transfer Segmentation |
| title_sort | keypoint transfer segmentation |
| url | http://hdl.handle.net/1721.1/110912 https://orcid.org/0000-0002-3652-1874 https://orcid.org/0000-0003-2516-731X |
| work_keys_str_mv | AT toewsm keypointtransfersegmentation AT wachingerchristian keypointtransfersegmentation AT langsgeorg keypointtransfersegmentation AT wellswilliamm keypointtransfersegmentation AT gollandpolina keypointtransfersegmentation |