OC10.01: Fetal MRI‐based network medicine: deciphering the phenome of callosal agenesis

Callosal agenesis (CCA) is one of the most common brain malformations detected prenatally. Rather than a single entity, CCA is an umbrella term defined by anatomy, independent of etiology or outcome. There is a substantial phenotypic diversity in cases of callosal agenesis – even if presenting as “isolated” finding therefore the complex imaging associations of callosal agenesis are difficult to understand. Network analysis is a novel approach to address such complexity and is increasingly applied in the novel field of “network medicine”. Single-centre study including 100 fetuses with various types of corpus callosum agenesis (CCA). (e.g. complete, partial, hypoplastic with intrahemaspheric cysts or lipomas) between 18 and 29 GW (55 female and 45 male fetuses). The observed phenotypes included: Colpocephaly, hippocampal verticalization, loss of temporal lobe asymmetry, lamination, opercularisation, cortical malformation, hypertelorism and others. Graph theory was used to examine brain-related co-occurring phenotypes of CCA in fetal MRI. We used Gephi a software for exploratory data and visual analytics with complementary statistics. Graph theory-based network analysis displays the highly interconnected nature of co-occurring phenotypes as clusters in 53 complete CCA and 31 in partial CCA, (7 intrahemaspheric cysts, 5 short thin callosal dysgenesis and 4 with lipomas) implying the existence of topological modules, representing local regions in the network as well as a high betweenness centrality (a measure of shortest paths that go through each node). Network-based approaches to human disease are important since images are complex data that can be analysed in to a network and provide visualization of these complex disease phenotypes, such a corpus callosum agenesis (CCA). Furthermore, it allows us to of identify common co- occurrence patterns within a disease and to link these findings to certain outcome groups which may serve as biomarkers for better or unfavourable motor or cognitive outcomes.