EEG TABANLI FONKSİYONEL BEYİN AĞLARININ ARİTMETİK BAŞARI VE CİNSİYET AÇISINDAN İNCELENMESİ
Yükleniyor...
Dosyalar
Tarih
2021-06-18
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Erişim Hakkı
info:eu-repo/semantics/openAccess
Özet
İnsan beyni, yapısal ve fonksiyonel olarak karmaşık bir ağdan oluşmaktadır. Beynin öğrenme, hafıza ve odaklanma gibi mekanizmalarındaki düzensizlikler üzerinden çeşitli beyin hastalıklarının açığa çıkarılması, güncel araştırma alanları içerisindedir. Bu araştırmaların önemli bir bölümü, graf teorisi ve kompleks ağ bilimi temelli bir yaklaşım olan fonksiyonel beyin ağları ile yapılmaktadır. İleri görüntüleme ve veri işleme teknikleri yardımı ile elde edilebilen ve “connectome” adı verilen beynin tam çözünürlüklü topolojisinin aksine fonksiyonel beyin ağları, beynin belirli bölgelerindeki sinirsel aktivitelerin istatistiksel uyumu üzerinden, daha düşük çözünürlüklü fakat erişimi daha kolay olan ağ gösterimleri oluşturmakta kullanılmaktadır. Bu çalışmaların veri altyapılarını, EEG, fMRI ve MEG gibi nörolojik görüntüleme yöntemleri oluşturmaktadır. Bu çalışmada, milisaniye düzeyinde sunduğu yüksek zamansal çözünürlük nedeniyle fonksiyonel beyin ağı çalışmalarında tercih edilen EEG sinyalleri kullanılmıştır. Çalışma içerisinde gönüllü deneklerden dinlenme ve bilişsel görev performansları olmak üzere iki farklı durum altında alınan EEG verileri kullanılmıştır. Deneklere bilişsel görev olarak beyin aktivitesinin yoğun ortaya çıktığı matematiksel çıkarma işlemi serisi verilmiştir. Veri seti cinsiyet, bilişsel aktivitenin varlığı ve bu aktivitedeki başarı durum bilgilerini içermektedir. EEG’nin her bandı sinirbilim araştırmaları için farklı bilgiler sağladığından, analizler teta (4–8 Hz), alfa (8–12 Hz), beta (12–30 Hz) ve gama (30–100 Hz) alt frekans bantlarına ayrılarak yapılmıştır. Deneklere ait EEG verileri üzerinden, PLV ve COH yöntemleri uygulanarak beynin fonksiyonel ağ gösterimleri elde edilmiştir. Bu gösterimler cinsiyet, bilişsel aktivitenin varlığı (veya dinlenme) ve aritmetik başarı (veya başarısızlık) durumları için ayrı ayrı sunulmuştur. Sinyal aktivitelerinin korelasyonunun göstergesi olan bağlantı ağırlıklarına 0-1 aralığında çeşitli eşikler uygulanarak ağlar gürültü içeren bağlantılardan temizlenmiş, bu versiyonlar üzerinde de çeşitli ağ parametrelerinin, uygulanan eşiğe göre değişimi incelenmiştir. Çalışma sonucunda özellikle aritmetik işlemlerde başarılı bireylerin beyinlerinin dinlenme süresince de bağlılık düzeyinin daha iyi olduğu gözlenmiştir. Dinlenme anında bağlılığı düşük seviyede olan erkek bireylere ait beyinlerin, bilişsel süreç başladığında bu bağlılık düzeyini önemli oranda artırdığı gözlenmiştir. Kadın bireylerin beyinleri ise dinlenme anında daha bağlantılı durumdadır. İşlevsel beyin ağlarının bağlılık düzeylerinin detaylı olarak incelenip sunulduğu çalışmamızda ayrıca lokal ölçekte hangi bağlantı örüntülerinin bilişsel süreçlerle ilişkili olduğu da irdelenmiştir.
Human brain structurally and functionally consists of a complex network. Detecting brain disorders with the aid of anomalies on the mechanisms of a brain such as learning, memory and focus are among the current research areas. A substantial part of these research areas is carried out with graph theory and functional brain networks, which is a complex network science-based approach. Unlike the full resolution topology of the brain called ""connectome"", which can be obtained with the help of advanced imaging and data processing techniques, functional brain networks are used to create network representations with lower resolution but easier to access through the statistical coherence of neural activity in certain regions of the brain. The data infrastructures of these studies are composed of neurological imaging methods such as EEG, fMRI and MEG. In this study, EEG signals, which are preferred in functional brain network studies, were used due to their high temporal resolution at the millisecond level. Within the scope of the study, EEG data obtained from volunteer subjects under two different conditions, namely resting and cognitive task performances, were used. As a cognitive task, a series of mathematical subtraction processes where brain activity occurs intensely was given to the subjects. The dataset includes information on gender, presence of cognitive activity, and success in this activity. Since each EEG frequency band provides different information for neuroscience research, the analysis was done by dividing it into sub-frequency bands as theta (4–8 Hz), alpha (8–12 Hz), beta (12–30 Hz), and gamma (30–100 Hz). By applying PLV and COH methods over the EEG data of the subjects, functional network representations of the brain were obtained. These representations are given separately for gender, presence of cognitive activity or resting state, and arithmetic success or failure. By applying various thresholds in the range of 0-1 to the link weights, which are indicative of the correlation of signal activities, the networks were cleared of links containing noise. The variation of various network parameters was investigated according to the threshold applied on these cleared versions. As a result of the study, it was observed that individuals who were successful in arithmetic operations had a better connectivity level in the resting state of their brains. It has been also observed that the brains of male individuals with low level of connectivity at rest significantly increase this level of connectivity when the cognitive process begins. The brains of female individuals are more connected at rest. This study examines the connectivity levels of functional brain networks and present them in detail as well as which connection patterns are related to cognitive processes at the local scale."
Human brain structurally and functionally consists of a complex network. Detecting brain disorders with the aid of anomalies on the mechanisms of a brain such as learning, memory and focus are among the current research areas. A substantial part of these research areas is carried out with graph theory and functional brain networks, which is a complex network science-based approach. Unlike the full resolution topology of the brain called ""connectome"", which can be obtained with the help of advanced imaging and data processing techniques, functional brain networks are used to create network representations with lower resolution but easier to access through the statistical coherence of neural activity in certain regions of the brain. The data infrastructures of these studies are composed of neurological imaging methods such as EEG, fMRI and MEG. In this study, EEG signals, which are preferred in functional brain network studies, were used due to their high temporal resolution at the millisecond level. Within the scope of the study, EEG data obtained from volunteer subjects under two different conditions, namely resting and cognitive task performances, were used. As a cognitive task, a series of mathematical subtraction processes where brain activity occurs intensely was given to the subjects. The dataset includes information on gender, presence of cognitive activity, and success in this activity. Since each EEG frequency band provides different information for neuroscience research, the analysis was done by dividing it into sub-frequency bands as theta (4–8 Hz), alpha (8–12 Hz), beta (12–30 Hz), and gamma (30–100 Hz). By applying PLV and COH methods over the EEG data of the subjects, functional network representations of the brain were obtained. These representations are given separately for gender, presence of cognitive activity or resting state, and arithmetic success or failure. By applying various thresholds in the range of 0-1 to the link weights, which are indicative of the correlation of signal activities, the networks were cleared of links containing noise. The variation of various network parameters was investigated according to the threshold applied on these cleared versions. As a result of the study, it was observed that individuals who were successful in arithmetic operations had a better connectivity level in the resting state of their brains. It has been also observed that the brains of male individuals with low level of connectivity at rest significantly increase this level of connectivity when the cognitive process begins. The brains of female individuals are more connected at rest. This study examines the connectivity levels of functional brain networks and present them in detail as well as which connection patterns are related to cognitive processes at the local scale."
Açıklama
Anahtar Kelimeler
Functional brain networks, EEG, Graph theory, Complex networks., Fonksiyonel beyin ağları, EEG, Graf teorisi, Kompleks ağlar
