We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

physics.med-ph

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Physics > Medical Physics

Title: Model-based Learning for Quantitative Susceptibility Mapping

Abstract: Quantitative susceptibility mapping (QSM) is a magnetic resonance imaging (MRI) technique that estimates magnetic susceptibility of tissue from Larmor frequency offset measurements. The generation of QSM requires solving a challenging ill-posed field-to-source inversion problem. Inaccurate field-to-source inversion often causes large susceptibility estimation errors that appear as streaking artifacts in the QSM, especially in massive hemorrhagic regions. Recently, several deep learning (DL) QSM techniques have been proposed and demonstrated impressive performance. Due to the inherent non-existent ground-truth QSM references, these DL techniques used either calculation of susceptibility through multiple orientation sampling (COSMOS) maps or synthetic data for network training. Therefore, they were constrained by the availability and accuracy of COSMOS maps, or suffered from performance drop when the training and testing domains were different. To address these limitations, we present a model-based DL method, denoted as uQSM. Without accessing to QSM labels, uQSM is trained using the well-established physical model. When evaluating on multi-orientation QSM datasets, uQSM achieves higher levels of quantitative accuracy compared to TKD, TV-FANSI, MEDI, and DIP approaches. When qualitatively evaluated on single-orientation datasets, uQSM outperforms other methods and reconstructed high quality QSM.
Comments: Accepted by MICCAI WORKSHOP MLMIR 2020
Subjects: Medical Physics (physics.med-ph); Image and Video Processing (eess.IV)
Cite as: arXiv:2004.06259 [physics.med-ph]
  (or arXiv:2004.06259v3 [physics.med-ph] for this version)

Submission history

From: Juan Liu [view email]
[v1] Tue, 14 Apr 2020 01:30:31 GMT (8192kb,D)
[v2] Fri, 14 Aug 2020 03:59:15 GMT (14314kb,D)
[v3] Mon, 17 Aug 2020 04:03:59 GMT (14497kb,D)

Link back to: arXiv, form interface, contact.