Rim Lesion Segmentation on 1MM QSM Positive Source : a Comparison between Deep Learning and Conventional Methods.
Motivation: To automate rim lesion segmentation in multiple sclerosisGoal(s): To compare deep learning and conventional methods for rim lesion segmentation in multiple sclerosisApproach: We compare Unet with chan-vese and Grabcut segmention of MR rim positive lesions. Results: Deep learning achieve the highest Dice score among the compared methods. Impact: Automate rim lesion segmentation in Multiple Sclerosis may allow determine those patient with persistent inflammation.
Authors
- İlhami Kovanlıkaya (ORCID: https://orcid.org/0000-0001-6339-9862)
- Mert Şişman (ORCID: https://orcid.org/0000-0003-0862-4478)
- Susan A. Gauthier (ORCID: https://orcid.org/0000-0002-4015-1781)
- Pascal Spincemaille (ORCID: https://orcid.org/0000-0002-8821-1341)
- Ha Manh Luu (ORCID: https://orcid.org/0000-0002-5678-6212)
- Yi Wang (ORCID: https://orcid.org/0000-0002-2031-087X)
- Thanh Nguyen (ORCID: https://orcid.org/0000-0002-6799-5132)
Publication Details
- Journal
- Proceedings on CD-ROM - International Society for Magnetic Resonance in Medicine. Scientific Meeting and Exhibition/Proceedings of the International Society for Magnetic Resonance in Medicine, Scientific Meeting and Exhibition
- Published
- 2024-11-26
- DOI
- https://doi.org/10.58530/2024/2506
- Primary Topic
- Oral and Maxillofacial Pathology
- Type
- article
- Field-Weighted Citation Impact
- 0.00