Segmenting the ischemic penumbra: a spatial Random Forest approach with automatic threshold finding

We propose a fully automatic method for segmenting the ischemic penumbra, using image texture and spatial features and a modified Random Forest algorithm, which we call Segmentation Forests, which has been designed to adapt the original Random Forests algorithm of Breiman to the segmentation of medical images. The method is fast, taking approximately six minutes to segment a new case, and has yields convincing results (An out-of-sample average Dice coefficient of 0.85, with a standard deviation of 0.06).