Učitavanje...
Multimodal MR Synthesis via Modality-Invariant Latent Representation
We propose a multi-input multi-output fully convolutional neural network model for MRI synthesis. The model is robust to missing data, as it benefits from, but does not require, additional input modalities. The model is trained end-to-end, and learns to embed all input modalities into a shared modal...
Spremljeno u:
| Izdano u: | IEEE Trans Med Imaging |
|---|---|
| Glavni autori: | , , , |
| Format: | Artigo |
| Jezik: | Inglês |
| Izdano: |
2017
|
| Teme: | |
| Online pristup: | https://ncbi.nlm.nih.gov/pmc/articles/PMC5904017/ https://ncbi.nlm.nih.gov/pubmed/29053447 https://ncbi.nlm.nih.govhttp://dx.doi.org/10.1109/TMI.2017.2764326 |
| Oznake: |
Dodaj oznaku
Bez oznaka, Budi prvi tko označuje ovaj zapis!
|