Random forest dissimilarity based multi-view learning for Radiomics application. (April 2019)

• Record Type: Journal Article
• Authors: Cao, Hongliu; Bernard, Simon; Sabourin, Robert; Heutte, Laurent
• Is Part Of: Pattern recognition. Volume 88(2019:Apr.)
• Page Start: 185
• Page End: 197
• Publication Date: 2019-04
• Subjects: Radiomics -- Dissimilarity space -- Random forest -- Machine learning -- Feature selection -- Multi-view learning -- High dimension -- Low sample size; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2018.11.011
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: Propose a Random forest dissimilarity based method for multi-view learning. Study the effect of hyperparameters on the quality of random forest dissimilarity. Compare the proposed method to the state of art Radiomics solutions. Compare the proposed method to multi-view learning approaches Show that the proposed approach outperforms the state-of-the-art methods. Abstract: Radiomics is a medical imaging technique that aims at extracting a large amount of features from one or several modalities of medical images, in order to help diagnose and treat diseases like cancers. Many recent studies have shown that Radiomics features can offer a lot of useful information that physicians cannot extract from these images, and can be efficiently associated with other information like gene or protein data. However, most of the classification studies in Radiomics report the use of feature selection methods without identifying the underlying machine learning challenges. In this paper, we first show that the Radiomics classification problem should be viewed as a high dimensional, low sample size, multi-view learning problem. Then, we propose a dissimilarity-based method for merging the information from the different views, based on Random Forest classifiers. The proposed approach is compared to different state-of-the-art Radiomics and multi-view solutions, on different public multi-view datasets as well as on Radiomics datasets. In particular, our experiments show that the proposed approach works better than the state-of-the-art methods from the Radiomics, as well as from the multi-view learning literature.