https://scholars.lib.ntu.edu.tw/handle/123456789/413030
標題: | Joint sequence learning and cross-modality convolution for 3D biomedical segmentation | 作者: | Tseng K.-L. Lin Y.-L. Hsu W. WINSTON HSU CHUNG-YANG HUANG |
公開日期: | 2017 | 卷: | 2017-January | 起(迄)頁: | 3739-3746 | 來源出版物: | 30th IEEE Conference on Computer Vision and Pattern Recognition | 摘要: | Deep learning models such as convolutional neural network have been widely used in 3D biomedical segmentation and achieve state-of-the-art performance. However, most of them often adapt a single modality or stack multiple modalities as different input channels. To better leverage the multimodalities, we propose a deep encoder-decoder structure with cross-modality convolution layers to incorporate different modalities of MRI data. In addition, we exploit convolutional LSTM to model a sequence of 2D slices, and jointly learn the multi-modalities and convolutional LSTM in an end-to-end manner. To avoid converging to the certain labels, we adopt a re-weighting scheme and two-phase training to handle the label imbalance. Experimental results on BRATS-2015 [13] show that our method outperforms state-of-the-art biomedical segmentation approaches. ? 2017 IEEE. |
URI: | https://scholars.lib.ntu.edu.tw/handle/123456789/413030 | ISBN: | 9781538604571 | DOI: | 10.1109/CVPR.2017.398 |
顯示於: | 資訊工程學系 |
檔案 | 描述 | 大小 | 格式 | |
---|---|---|---|---|
1704.07754.pdf | 1.12 MB | Adobe PDF | 檢視/開啟 |
在 IR 系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。