Fig. 1
From: Non-small-cell lung cancer classification via RNA-Seq and histology imaging probability fusion
Pipeline for sample prediction. (i) Both the WSI and the RNA-Seq data for that case ID are obtained. (ii) Non-overlapping \(512\times 512\) tiles are extracted for the WSI, filtering the background. For RNA-Seq, we took the set of DEGs selected by the mRMR ranking. (iii) For the WSI, the probabilities are obtained by averaging the number of tiles predicted per class and the total number of them. For RNA-Seq data, the probabilities are returned by the SVM. (iv) We fuse the probabilities by averaging the ones obtained by each classifier per class, and the final prediction is the class with the higher probability