Zheng Qing, Milthorpe Bruce K, Jones Allan S
Graduate School of Biomedical Engineering, University of New South Wales, Sydney, Australia.
Cytometry A. 2004 Jan;57(1):1-9. doi: 10.1002/cyto.a.10106.
Automated cell recognition from histologic images is a very complex task. Traditionally, the image is segmented by some methods chosen to suit the image type, the objects are measured, and then a classifier is used to determine cell type from the object's measurements. Different classifiers have been used with reasonable success, including neural networks working with data from morphometric analysis.
Image data of cells were input directly into neural networks to determine the feasibility of direct classification by using pixel intensity information. Several types of neural network and their ability to work with cells in a complex patterned background were assessed for a variety of images and cell types and for the accuracy of classification.
Inflammatory cells from animal biomaterial implants in rabbit paravertebral muscle were imaged in histologic sections. Simple, three-layer, fully connected, back-propagation neural networks and four-layer networks with two layers of a shared-weights neural network were most successful at classifying the cells from the images, with 97% and 98% correct recognition rates, respectively.
The high accuracy recognition rate shows the potential for direct classification of visual image pixel data by neural networks.
从组织学图像中自动识别细胞是一项非常复杂的任务。传统上，通过选择适合图像类型的一些方法对图像进行分割，测量对象，然后使用分类器根据对象的测量结果确定细胞类型。已经使用了不同的分类器并取得了一定的成功，包括使用来自形态计量分析数据的神经网络。
将细胞的图像数据直接输入神经网络，以利用像素强度信息确定直接分类的可行性。针对各种图像和细胞类型以及分类的准确性，评估了几种类型的神经网络及其在复杂图案背景下处理细胞的能力。
对兔椎旁肌中动物生物材料植入物的炎性细胞进行了组织学切片成像。简单的三层全连接反向传播神经网络和具有两层共享权重神经网络的四层网络在对图像中的细胞进行分类方面最为成功，正确识别率分别为97%和98%。
高准确率表明神经网络对视觉图像像素数据进行直接分类具有潜力。
