SYSTEM AND METHOD FOR CLASSIFYING AND SEGMENTING M

专利名称：SYSTEM AND METHOD FOR CLASSIFYING

AND SEGMENTING MICROSCOPY IMAGESWITH DEEP MULTIPLE INSTANCE LEARNING

发明人：Oren KRAUS,Jimmy BA,Brendan FREY申请号：US15352821申请日：20161116

公开号：US20180137338A1公开日：20180517

专利附图：

摘要：Systems and methods that receive as input microscopy images, extract features,and apply layers of processing units to compute one or more set of cellular phenotype

features, corresponding to cellular densities and/or fluorescence measured underdifferent conditions. The system is a neural network architecture having a convolutionalneural network followed by a multiple instance learning (MIL) pooling layer. The systemdoes not necessarily require any segmentation steps or per cell labels as theconvolutional neural network can be trained and tested directly on raw microscopyimages in real-time. The system computes class specific feature maps for every

phenotype variable using a fully convolutional neural network and uses multiple instancelearning to aggregate across these class specific feature maps. The system producespredictions for one or more reference cellular phenotype variables based on microscopyimages with populations of cells.

申请人：THE GOVERNING COUNCIL OF THE UNIVERSITY OF TORONTO

地址：Toronto CA

国籍：CA

更多信息请下载全文后查看