Single-subject Multi-contrast MRI Super-resolution via Implicit Neural Representations
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2023
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866913186210381824 |
|---|---|
| author | McGinnis, Julian Shit, Suprosanna Li, Hongwei Bran Sideri-Lampretsa, Vasiliki Graf, Robert Dannecker, Maik Pan, Jiazhen Ansó, Nil Stolt Mühlau, Mark Kirschke, Jan S. Rueckert, Daniel Wiestler, Benedikt |
| author_facet | McGinnis, Julian Shit, Suprosanna Li, Hongwei Bran Sideri-Lampretsa, Vasiliki Graf, Robert Dannecker, Maik Pan, Jiazhen Ansó, Nil Stolt Mühlau, Mark Kirschke, Jan S. Rueckert, Daniel Wiestler, Benedikt |
| contents | Clinical routine and retrospective cohorts commonly include multi-parametric Magnetic Resonance Imaging; however, they are mostly acquired in different anisotropic 2D views due to signal-to-noise-ratio and scan-time constraints. Thus acquired views suffer from poor out-of-plane resolution and affect downstream volumetric image analysis that typically requires isotropic 3D scans. Combining different views of multi-contrast scans into high-resolution isotropic 3D scans is challenging due to the lack of a large training cohort, which calls for a subject-specific framework. This work proposes a novel solution to this problem leveraging Implicit Neural Representations (INR). Our proposed INR jointly learns two different contrasts of complementary views in a continuous spatial function and benefits from exchanging anatomical information between them. Trained within minutes on a single commodity GPU, our model provides realistic super-resolution across different pairs of contrasts in our experiments with three datasets. Using Mutual Information (MI) as a metric, we find that our model converges to an optimum MI amongst sequences, achieving anatomically faithful reconstruction. Code is available at: https://github.com/jqmcginnis/multi_contrast_inr/ |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2303_15065 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Single-subject Multi-contrast MRI Super-resolution via Implicit Neural Representations McGinnis, Julian Shit, Suprosanna Li, Hongwei Bran Sideri-Lampretsa, Vasiliki Graf, Robert Dannecker, Maik Pan, Jiazhen Ansó, Nil Stolt Mühlau, Mark Kirschke, Jan S. Rueckert, Daniel Wiestler, Benedikt Image and Video Processing Computer Vision and Pattern Recognition Clinical routine and retrospective cohorts commonly include multi-parametric Magnetic Resonance Imaging; however, they are mostly acquired in different anisotropic 2D views due to signal-to-noise-ratio and scan-time constraints. Thus acquired views suffer from poor out-of-plane resolution and affect downstream volumetric image analysis that typically requires isotropic 3D scans. Combining different views of multi-contrast scans into high-resolution isotropic 3D scans is challenging due to the lack of a large training cohort, which calls for a subject-specific framework. This work proposes a novel solution to this problem leveraging Implicit Neural Representations (INR). Our proposed INR jointly learns two different contrasts of complementary views in a continuous spatial function and benefits from exchanging anatomical information between them. Trained within minutes on a single commodity GPU, our model provides realistic super-resolution across different pairs of contrasts in our experiments with three datasets. Using Mutual Information (MI) as a metric, we find that our model converges to an optimum MI amongst sequences, achieving anatomically faithful reconstruction. Code is available at: https://github.com/jqmcginnis/multi_contrast_inr/ |
| title | Single-subject Multi-contrast MRI Super-resolution via Implicit Neural Representations |
| topic | Image and Video Processing Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2303.15065 |