Data Augmentation Using Learned Transformations for One-Shot Medical Image Segmentation
Image segmentation is an important task in many medical applications. Methods based on convolutional neural networks attain state-of-the-art accuracy; however, they typically rely on supervised training with large labeled datasets. Labeling medical images requires significant expertise and time, and...
| Main Authors: | , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Institute of Electrical and Electronics Engineers (IEEE)
2021
|
| Online Access: | https://hdl.handle.net/1721.1/129978 |
| _version_ | 1826202227898843136 |
|---|---|
| author | Zhao, Amy (Xiaoyu Amy) Balakrishnan, Guha Durand, Fredo Guttag, John V Dalca, Adrian Vasile |
| author2 | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory |
| author_facet | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Zhao, Amy (Xiaoyu Amy) Balakrishnan, Guha Durand, Fredo Guttag, John V Dalca, Adrian Vasile |
| author_sort | Zhao, Amy (Xiaoyu Amy) |
| collection | MIT |
| description | Image segmentation is an important task in many medical applications. Methods based on convolutional neural networks attain state-of-the-art accuracy; however, they typically rely on supervised training with large labeled datasets. Labeling medical images requires significant expertise and time, and typical hand-tuned approaches for data augmentation fail to capture the complex variations in such images. We present an automated data augmentation method for synthesizing labeled medical images. We demonstrate our method on the task of segmenting magnetic resonance imaging (MRI) brain scans. Our method requires only a single segmented scan, and leverages other unlabeled scans in a semi-supervised approach. We learn a model of transformations from the images, and use the model along with the labeled example to synthesize additional labeled examples. Each transformation is comprised of a spatial deformation field and an intensity change, enabling the synthesis of complex effects such as variations in anatomy and image acquisition procedures. We show that training a supervised segmenter with these new examples provides significant improvements over state-of-the-art methods for one-shot biomedical image segmentation. |
| first_indexed | 2024-09-23T12:04:12Z |
| format | Article |
| id | mit-1721.1/129978 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T12:04:12Z |
| publishDate | 2021 |
| publisher | Institute of Electrical and Electronics Engineers (IEEE) |
| record_format | dspace |
| spelling | mit-1721.1/1299782022-09-27T23:53:49Z Data Augmentation Using Learned Transformations for One-Shot Medical Image Segmentation Zhao, Amy (Xiaoyu Amy) Balakrishnan, Guha Durand, Fredo Guttag, John V Dalca, Adrian Vasile Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Image segmentation is an important task in many medical applications. Methods based on convolutional neural networks attain state-of-the-art accuracy; however, they typically rely on supervised training with large labeled datasets. Labeling medical images requires significant expertise and time, and typical hand-tuned approaches for data augmentation fail to capture the complex variations in such images. We present an automated data augmentation method for synthesizing labeled medical images. We demonstrate our method on the task of segmenting magnetic resonance imaging (MRI) brain scans. Our method requires only a single segmented scan, and leverages other unlabeled scans in a semi-supervised approach. We learn a model of transformations from the images, and use the model along with the labeled example to synthesize additional labeled examples. Each transformation is comprised of a spatial deformation field and an intensity change, enabling the synthesis of complex effects such as variations in anatomy and image acquisition procedures. We show that training a supervised segmenter with these new examples provides significant improvements over state-of-the-art methods for one-shot biomedical image segmentation. 2021-02-23T20:37:04Z 2021-02-23T20:37:04Z 2019-06 2019-04 2020-12-11T17:15:44Z Article http://purl.org/eprint/type/ConferencePaper 9781728132938 1063-6919 https://hdl.handle.net/1721.1/129978 Zhao, Amy et al. “Data Augmentation Using Learned Transformations for One-Shot Medical Image Segmentation.” 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2019, Long Beach, California, Institute of Electrical and Electronics Engineers (IEEE), June 2019. © 2019 The Author(s) en 10.1109/CVPR.2019.00874 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf Institute of Electrical and Electronics Engineers (IEEE) arXiv |
| spellingShingle | Zhao, Amy (Xiaoyu Amy) Balakrishnan, Guha Durand, Fredo Guttag, John V Dalca, Adrian Vasile Data Augmentation Using Learned Transformations for One-Shot Medical Image Segmentation |
| title | Data Augmentation Using Learned Transformations for One-Shot Medical Image Segmentation |
| title_full | Data Augmentation Using Learned Transformations for One-Shot Medical Image Segmentation |
| title_fullStr | Data Augmentation Using Learned Transformations for One-Shot Medical Image Segmentation |
| title_full_unstemmed | Data Augmentation Using Learned Transformations for One-Shot Medical Image Segmentation |
| title_short | Data Augmentation Using Learned Transformations for One-Shot Medical Image Segmentation |
| title_sort | data augmentation using learned transformations for one shot medical image segmentation |
| url | https://hdl.handle.net/1721.1/129978 |
| work_keys_str_mv | AT zhaoamyxiaoyuamy dataaugmentationusinglearnedtransformationsforoneshotmedicalimagesegmentation AT balakrishnanguha dataaugmentationusinglearnedtransformationsforoneshotmedicalimagesegmentation AT durandfredo dataaugmentationusinglearnedtransformationsforoneshotmedicalimagesegmentation AT guttagjohnv dataaugmentationusinglearnedtransformationsforoneshotmedicalimagesegmentation AT dalcaadrianvasile dataaugmentationusinglearnedtransformationsforoneshotmedicalimagesegmentation |