Abstract
| Original language | English |
|---|---|
| Title of host publication | Brainlesion |
| Subtitle of host publication | Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - 3rd International Workshop, BrainLes 2017, Held in Conjunction with MICCAI 2017, Revised Selected Papers |
| Place of Publication | Switzerland |
| Publisher | Springer |
| Pages | 263-273 |
| Number of pages | 11 |
| ISBN (Electronic) | 9783319752389 |
| ISBN (Print) | 9783319752372 |
| DOIs | |
| Publication status | Published - 1 Jan 2018 |
| Event | Medical Image Computing and Computer-Assisted Intervention 2017 - Quebec City Convention Centre, Quebec, Canada Duration: 10 Sep 2017 → 14 Sep 2017 Conference number: 20th http://www.miccai2017.org/ |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 10670 LNCS |
| ISSN (Print) | 0302-9743 |
| ISSN (Electronic) | 1611-3349 |
Conference
| Conference | Medical Image Computing and Computer-Assisted Intervention 2017 |
|---|---|
| Abbreviated title | MICCAI 2017 |
| Country | Canada |
| City | Quebec |
| Period | 10/09/17 → 14/09/17 |
| Internet address |
Keywords
- CNN
- Computer vision
- Deep learning
- Image segmentation
Cite this
}
Residual encoder and convolutional decoder neural network for glioma segmentation. / Pawar, Kamlesh; Chen, Zhaolin; Shah, N. Jon; Egan, Gary.
Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - 3rd International Workshop, BrainLes 2017, Held in Conjunction with MICCAI 2017, Revised Selected Papers. Switzerland : Springer, 2018. p. 263-273 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10670 LNCS).Research output: Chapter in Book/Report/Conference proceeding › Conference Paper › Research › peer-review
TY - GEN
T1 - Residual encoder and convolutional decoder neural network for glioma segmentation
AU - Pawar, Kamlesh
AU - Chen, Zhaolin
AU - Shah, N. Jon
AU - Egan, Gary
PY - 2018/1/1
Y1 - 2018/1/1
N2 - A deep learning approach to glioma segmentation is presented. An encoder and decoder pair deep learning network is designed which takes T1, T2, T1-CE (contrast enhanced) and T2-Flair (fluid attenuation inversion recovery) images as input and outputs the segmented labels. The encoder is a 49 layer deep residual learning architecture that encodes the 240 × 240 × 4 input images into 8 × 8 × 2048 feature maps. The decoder network takes these feature maps and extract the segmented labels. The decoder network is fully convolutional network consisting of convolutional and upsampling layers. Additionally, the input images are downsampled using bilinear interpolation and are inserted into the decoder network through concatenation. This concatenation step provides spatial information of the tumor to the decoder, which was lost due to pooling/downlsampling during encoding. The network is trained on the BRATS-17 training dataset and validated on the validation dataset. The dice score, sensitivity and specificity of the segmented whole tumor, core tumor and enhancing tumor is computed on validation dataset. The mean dice score for whole tumor, core tumor and enhancing tumor for validation dataset were 0.824, 0.627 and 0.575, respectively.
AB - A deep learning approach to glioma segmentation is presented. An encoder and decoder pair deep learning network is designed which takes T1, T2, T1-CE (contrast enhanced) and T2-Flair (fluid attenuation inversion recovery) images as input and outputs the segmented labels. The encoder is a 49 layer deep residual learning architecture that encodes the 240 × 240 × 4 input images into 8 × 8 × 2048 feature maps. The decoder network takes these feature maps and extract the segmented labels. The decoder network is fully convolutional network consisting of convolutional and upsampling layers. Additionally, the input images are downsampled using bilinear interpolation and are inserted into the decoder network through concatenation. This concatenation step provides spatial information of the tumor to the decoder, which was lost due to pooling/downlsampling during encoding. The network is trained on the BRATS-17 training dataset and validated on the validation dataset. The dice score, sensitivity and specificity of the segmented whole tumor, core tumor and enhancing tumor is computed on validation dataset. The mean dice score for whole tumor, core tumor and enhancing tumor for validation dataset were 0.824, 0.627 and 0.575, respectively.
KW - CNN
KW - Computer vision
KW - Deep learning
KW - Image segmentation
UR - http://www.scopus.com/inward/record.url?scp=85042544528&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-75238-9_23
DO - 10.1007/978-3-319-75238-9_23
M3 - Conference Paper
AN - SCOPUS:85042544528
SN - 9783319752372
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 263
EP - 273
BT - Brainlesion
PB - Springer
CY - Switzerland
ER -