Semi-supervised Learning for Fetal Brain MRI Quality Assessment with ROI Consistency
© 2020, Springer Nature Switzerland AG. Fetal brain MRI is useful for diagnosing brain abnormalities but is challenged by fetal motion. The current protocol for T2-weighted fetal brain MRI is not robust to motion so image volumes are degraded by inter- and intra- slice motion artifacts. Besides, man...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer International Publishing
2021
|
| Online Access: | https://hdl.handle.net/1721.1/137502 |
| _version_ | 1811087718252281856 |
|---|---|
| author | Xu, J Lala, S Gagoski, B Abaci Turk, E Grant, PE Golland, P Adalsteinsson, E |
| author_facet | Xu, J Lala, S Gagoski, B Abaci Turk, E Grant, PE Golland, P Adalsteinsson, E |
| author_sort | Xu, J |
| collection | MIT |
| description | © 2020, Springer Nature Switzerland AG. Fetal brain MRI is useful for diagnosing brain abnormalities but is challenged by fetal motion. The current protocol for T2-weighted fetal brain MRI is not robust to motion so image volumes are degraded by inter- and intra- slice motion artifacts. Besides, manual annotation for fetal MR image quality assessment are usually time-consuming. Therefore, in this work, a semi-supervised deep learning method that detects slices with artifacts during the brain volume scan is proposed. Our method is based on the mean teacher model, where we not only enforce consistency between student and teacher models on the whole image, but also adopt an ROI consistency loss to guide the network to focus on the brain region. The proposed method is evaluated on a fetal brain MR dataset with 11,223 labeled images and more than 200,000 unlabeled images. Results show that compared with supervised learning, the proposed method can improve model accuracy by about 6% and outperform other state-of-the-art semi-supervised learning methods. The proposed method is also implemented and evaluated on an MR scanner, which demonstrates the feasibility of online image quality assessment and image reacquisition during fetal MR scans. |
| first_indexed | 2024-09-23T13:50:54Z |
| format | Article |
| id | mit-1721.1/137502 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T13:50:54Z |
| publishDate | 2021 |
| publisher | Springer International Publishing |
| record_format | dspace |
| spelling | mit-1721.1/1375022021-11-06T03:18:40Z Semi-supervised Learning for Fetal Brain MRI Quality Assessment with ROI Consistency Xu, J Lala, S Gagoski, B Abaci Turk, E Grant, PE Golland, P Adalsteinsson, E © 2020, Springer Nature Switzerland AG. Fetal brain MRI is useful for diagnosing brain abnormalities but is challenged by fetal motion. The current protocol for T2-weighted fetal brain MRI is not robust to motion so image volumes are degraded by inter- and intra- slice motion artifacts. Besides, manual annotation for fetal MR image quality assessment are usually time-consuming. Therefore, in this work, a semi-supervised deep learning method that detects slices with artifacts during the brain volume scan is proposed. Our method is based on the mean teacher model, where we not only enforce consistency between student and teacher models on the whole image, but also adopt an ROI consistency loss to guide the network to focus on the brain region. The proposed method is evaluated on a fetal brain MR dataset with 11,223 labeled images and more than 200,000 unlabeled images. Results show that compared with supervised learning, the proposed method can improve model accuracy by about 6% and outperform other state-of-the-art semi-supervised learning methods. The proposed method is also implemented and evaluated on an MR scanner, which demonstrates the feasibility of online image quality assessment and image reacquisition during fetal MR scans. 2021-11-05T15:02:50Z 2021-11-05T15:02:50Z 2020 2020-11-20T18:08:41Z Article http://purl.org/eprint/type/ConferencePaper https://hdl.handle.net/1721.1/137502 Xu, J, Lala, S, Gagoski, B, Abaci Turk, E, Grant, PE et al. 2020. "Semi-supervised Learning for Fetal Brain MRI Quality Assessment with ROI Consistency." Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 12266 LNCS. en 10.1007/978-3-030-59725-2_37 Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf Springer International Publishing arXiv |
| spellingShingle | Xu, J Lala, S Gagoski, B Abaci Turk, E Grant, PE Golland, P Adalsteinsson, E Semi-supervised Learning for Fetal Brain MRI Quality Assessment with ROI Consistency |
| title | Semi-supervised Learning for Fetal Brain MRI Quality Assessment with ROI Consistency |
| title_full | Semi-supervised Learning for Fetal Brain MRI Quality Assessment with ROI Consistency |
| title_fullStr | Semi-supervised Learning for Fetal Brain MRI Quality Assessment with ROI Consistency |
| title_full_unstemmed | Semi-supervised Learning for Fetal Brain MRI Quality Assessment with ROI Consistency |
| title_short | Semi-supervised Learning for Fetal Brain MRI Quality Assessment with ROI Consistency |
| title_sort | semi supervised learning for fetal brain mri quality assessment with roi consistency |
| url | https://hdl.handle.net/1721.1/137502 |
| work_keys_str_mv | AT xuj semisupervisedlearningforfetalbrainmriqualityassessmentwithroiconsistency AT lalas semisupervisedlearningforfetalbrainmriqualityassessmentwithroiconsistency AT gagoskib semisupervisedlearningforfetalbrainmriqualityassessmentwithroiconsistency AT abaciturke semisupervisedlearningforfetalbrainmriqualityassessmentwithroiconsistency AT grantpe semisupervisedlearningforfetalbrainmriqualityassessmentwithroiconsistency AT gollandp semisupervisedlearningforfetalbrainmriqualityassessmentwithroiconsistency AT adalsteinssone semisupervisedlearningforfetalbrainmriqualityassessmentwithroiconsistency |