MinesParisTech, PSL Research University, Center for Mathematical Morphology and CBIO-Centre for Computational Biology, Fontainebleau, FranceAuthors:
Peter Naylor, Vaia Machairas, Etienne Decenciere, and Thomas WalterAbstract:
In this paper, we present our method to address the two tasks of the challenge CAMELYON16: tumor detection and slide classification. Unless specified, all steps are performed at resolution 2 of slides images. Our method is based on supervised pixel classification, where features are calculated on superpixel supports of different scales. For classification, we tested Random Forests and Support Vector Machines.Results:
The following figure shows the receiver operating characteristic (ROC) curve of the method.
The following figure shows the free-response receiver operating characteristic (FROC) curve of the method.
The table below presents the average sensitivity of the developed system at 6 predefined false positive rates: 1/4, 1/2, 1, 2, 4, and 8 FPs per whole slide image.