کنفرانس مهندسی برق ایران

صفحه اصلی / سی و دومین کنفرانس بین المللی مهندسی برق

Effect of structural connectivity weightings in graph-based analysis in Schizophrenia

نویسندگان :

Sara Khamseh¹ Farzaneh Keyvanfard²

1- Amirkabir University of Technology (Tehran Polytechnic) 2- Institute for Research in Fundamental Sciences (IPM),

کلمات کلیدی :

structural connectivity،schizophrenia،brain network،thresholding method،graph theory

چکیده :

Graph theory models the brain as a complex network to sets of vertices linked by connections. The connections between vertices can be determined by the structural connectivity (SC) of the brain. SC, based on tractography, is represented through different measurements including global fractional anisotropy and fiber density. To apply graph theory, matrix thresholding is required to delete the weak connections in the data and then the graph parameters such as nodal strength are computed. In this pipeline, different metrics of representing structural connectivity and thresholding methods can be selected. Here we aim to investigate whether this selection affects the discrimination results between healthy control and schizophrenia patients. Our findings indicate that applying density-based thresholding to the global fractional anisotropy metric can identify the alternation more thoroughly, while the significant difference in graph measurements may not be detected using other approached.