Title | ||
---|---|---|
MIDeepSeg: Minimally interactive segmentation of unseen objects from medical images using deep learning |
Abstract | ||
---|---|---|
•A novel deep learning-based interactive framework for medical image segmentation, with high accuracy, efficiency and good generalizability to unseen objects.•A context-aware and parameter-free method to encode user interactions for convolutional neural networks.•A novel information fusion method to efficiently refine segmentation obtained by CNNs.•Extensive experiments show our interactive method is ready-to-use for various previously unseen 2D and 3D images of different modalities. |
Year | DOI | Venue |
---|---|---|
2021 | 10.1016/j.media.2021.102102 | Medical Image Analysis |
Keywords | DocType | Volume |
Interactive image segmentation,Convolutional neural network,Geodesic distance,Generalization | Journal | 72 |
ISSN | Citations | PageRank |
1361-8415 | 2 | 0.64 |
References | Authors | |
0 | 9 |
Name | Order | Citations | PageRank |
---|---|---|---|
Xiangde Luo | 1 | 7 | 3.13 |
Guotai Wang | 2 | 119 | 10.33 |
Tao Song | 3 | 17 | 4.34 |
Jingyang Zhang | 4 | 4 | 3.71 |
Michael Aertsen | 5 | 81 | 6.21 |
Jan Deprest | 6 | 123 | 20.45 |
Sébastien Ourselin | 7 | 576 | 57.16 |
Tom Vercauteren | 8 | 115 | 13.99 |
Shaoting Zhang | 9 | 2 | 0.98 |