Abstract | ||
---|---|---|
•It is an end to end network for osteosarcoma CT image segmentation.•The feature extraction part of the network was deep, and rich hierarchical features could be learned directly from the images by the network.•Multiple supervised side output modules were added to the network for guiding the learning of shape features and semantic features.•Our method achieved higher F1-measure compared with the FCN and U-Net method. |
Year | DOI | Venue |
---|---|---|
2018 | 10.1016/j.compmedimag.2018.01.006 | Computerized Medical Imaging and Graphics |
Keywords | Field | DocType |
Osteosarcoma segmentation,Deep residual network,Multiple supervised networks | Residual,Computer vision,Gradient descent,Segmentation,Image segmentation,Ground truth,Artificial intelligence,Dice,Medicine,Feature learning | Journal |
Volume | ISSN | Citations |
63 | 0895-6111 | 4 |
PageRank | References | Authors |
0.38 | 16 | 6 |