Building medical image classifiers with very limited data using segmentation networks.
Medical Image Analysis(2018)
摘要
•Building medical image classifiers using features from segmentation networks.•Important for 3D image analysis as ImageNet pre-trained CNNs are unavailable in 3D.•Tested on 3D brain tumor type and 2D cardiac semantic level classifications.•Compared to classifiers trained from scratch and with ImageNet pre-trained CNNs.•82% and 86% accuracies on brain tumor and cardiac semantic level classifications.
更多查看译文
关键词
Image segmentation,Image classification,Deep learning,Transfer learning,Convolutional neural network,Fully convolutional network
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络