Graph theory analyses on connectivity maps obtained by partial directed coherence using EEG data of dyslexic and healthy children


Tezin Türü: Yüksek Lisans

Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Mühendislik Fakültesi, Elektrik ve Elektronik Mühendisliği Bölümü, Türkiye

Tezin Onay Tarihi: 2017

Öğrenci: EKİN CAN ERKUŞ

Danışman: İLKAY ULUSOY

Özet:

Dyslexia is a common brain disorder which is defined as reading and sometimes learning disability. In this thesis study, EEG data which were collected from 31 dyslexics and 27 non-dyslexic children during reading task were used. First, multivariate autoregressive modelling was made. Then using MVAR models, brain connectivity networks were obtained with partial directed coherence (PDC) algorithms. Using brain connectivity networks, graph theory properties such as “characteristic path length”, “clustering coefficient”, “global efficiency” and “small-world measure” were calculated. Finally, group analyses were done based on the graph theory properties using statistical analyses. Between groups, for after stimulus condition, there was a significant difference in terms of “small-world measure”. Between before stimulus and after stimulus conditions, “global efficiency” was found to have significant difference in control group. Similarly, “characteristic path length” and “clustering coefficient” properties were found to have significant difference in dyslexic group. Also, main hub nodes were discovered for each subject using connectivity maps. Hub nodes distributions had differences between groups in right frontal and right occipito-parietal regions of brains. All the results were compared with literature and discussed.