Regional radiomics similarity networks (R2SNs) in the human brain: reproducibility, small-world properties and a biological basis

Background: A structural covariance network (SCN) has been used successfully to structural magnetic resonance imaging (MRI) studies. However, most SCNs were constructed by a unitary marker that was insensitive for discriminating different disease phases. The aim of this study was to devise a novel regional radiomics similarity network (R2SN) that could provide more comprehensive information in morphological network analysis. Methods: R2SNs were constructed by computing the Pearson correlations between the radiomics features extracted from any pair of regions for each subject. We further assessed the small-world property of R2SNs using the graph theory method, and we evaluated the reproducibility in different datasets and the reliability of the R2SNs through test-retest analysis. The relationship between the R2SNs and interregional coexpression of enriched genes was also explored, as well as the relationship with general intelligence. Results: R2SNs could be replicated in different datasets, regardless of the use of different feature subsets. R2SNs showed high reliability in the test-retest analysis (intraclass correlation coefficient (ICC)>0.7). In addition, the small-word property (σ>2) and the high correlation between gene expression (R=0.24, P