| Datensatz noch nicht abgeschlossen | |
| Zugriffsnummer | 56036 |
| Dokumenttyp | Konferenzartikel |
| Sprache | Englisch |
| Titel | Zero-shot self-supervised Greedy learning for magnitude-phase reconstruction in MR elastography |
| Autor(in); Institution |
Martin, Stefan; 8.1, Biomedizinische Magnetresonanz, PTB-Berlin
Guastini, Mara; 8.1, Biomedizinische Magnetresonanz, PTB-Berlin
Schattenfroh, Jakob; Charité – Universitätsmedizin Berlin, GERMANY
Sack, Ingolf; Charité – Universitätsmedizin Berlin, GERMANY
|
| Quelle/Jahr | Proceedings of the International Society for Magnetic Resonance in Medicine:(2026) |
| Availability | [online only] |
| ISSN | 1065-9889 (online) |
| URL | |
| Verlag | ISMRM |
| Konferenzangaben | Annual Meeting of the ISMRM, Capetown, 9-14, Mai, 2026, South Africa |
| Freie Schlagworte | Physics-Informed Deep Learning ; Accelerated MRE Reconstruction ; Zero-Shot |
| Zusammenfassung | Soft-tissue stiffness provides important diagnostic information, as many diseases alter mechanical properties. MRE quantifies these by encodingoscillatory tissue motion as phase variations in the MR signal, yet accurate reconstruction under strong undersampling remains challenging. Here, we proposean iterative approach in which we decouple the regularization of magnitude and phase images using two distinct learned neural-network priors. |
