Structure Prediction for Gland Segmentation With Hand-Crafted and Deep Convolutional Features
Structure Prediction for Gland Segmentation With Hand-Crafted and Deep Convolutional Features
复制标题
DOI：
10.1109/tmi.2017.2750210
复制
发表时间：
2018-01-01
影响因子：
10.6
通讯作者：
McKenna, Stephen J.
中科院分区：
文献类型：
--
作者：
Manivannan, Siyamalan;Li, Wenqi;McKenna, Stephen J.
我们提出了一种新的方法来分割结肠组织病理学图像的腺体结构的实例。我们使用一种结构学习方法，它代表了类标签的局部空间配置，捕获通常被滑动窗口方法忽略的结构信息。这允许我们揭示像素标签的不同空间结构（例如，相邻腺体之间的位置或远离腺体的位置），并将相邻的腺体结构正确地识别为单独的实例。通过聚类得到标签结构的示例，并用于训练支持向量机分类器。然后将预测的标签结构组合并进行后处理以获得分割图。我们将联合收割机手工制作的多尺度图像特征与经过训练的深度卷积网络计算的特征相结合，以将图像映射到分割图。我们评估所提出的方法在公共领域的GlaS数据集，它允许广泛的比较，最近的替代方法。使用GlaS竞赛协议，我们的方法实现了整体最佳性能。
We present a novel method to segment instances of glandular structures from colon histopathology images. We use a structure learning approach which represents local spatial configurations of class labels, capturing structural information normally ignored by sliding-window methods. This allows us to reveal different spatial structures of pixel labels (e.g., locations between adjacent glands, or far from glands), and to identify correctly neighboring glandular structures as separate instances. Exemplars of label structures are obtained via clustering and used to train support vector machine classifiers. The label structures predicted are then combined and post-processed to obtain segmentation maps. We combine handcrafted, multi-scale image features with features computed by a deep convolutional network trained to map images to segmentation maps. We evaluate the proposed method on the public domain GlaS data set, which allows extensive comparisons with recent, alternative methods. Using the GlaS contest protocol, our method achieves the overall best performance.