Browsing by Subject "Connectivity"

Now showing 1 - 5 of 5

Open Access
A plug-in graph neural network to boost temporal sensitivity in fMRI analysis
(IEEE, 2024-09) Şıvgın, Irmak; Bedel, Hasan Atakan; Ozturk, Saban; Çukur, Tolga
Learning-based methods offer performance leaps over traditional methods in classification analysis of high-dimensional functional MRI (fMRI) data. In this domain, deep-learning models that analyze functional connectivity (FC) features among brain regions have been particularly promising. However, many existing models receive as input temporally static FC features that summarize inter-regional interactions across an entire scan, reducing the temporal sensitivity of classifiers by limiting their ability to leverage information on dynamic FC features of brain activity. To improve the performance of baseline classification models without compromising efficiency, here we propose a novel plug-in based on a graph neural network, GraphCorr, to provide enhanced input features to baseline models. The proposed plug-in computes a set of latent FC features with enhanced temporal information while maintaining comparable dimensionality to static features. Taking brain regions as nodes and blood-oxygen-level-dependent (BOLD) signals as node inputs, GraphCorr leverages a node embedder module based on a transformer encoder to capture dynamic latent representations of BOLD signals. GraphCorr also leverages a lag filter module to account for delayed interactions across nodes by learning correlational features of windowed BOLD signals across time delays. These two feature groups are then fused via a message passing algorithm executed on the formulated graph. Comprehensive demonstrations on three public datasets indicate improved classification performance for several state-of-the-art graph and convolutional baseline models when they are augmented with GraphCorr.
Open Access
BolT: Fused window transformers for fMRI time series analysis
(Elsevier B.V., 2023-05-18) Bedel, Hasan Atakan; Şıvgın, Irmak; Dalmaz, Onat; Ul Hassan Dar, Salman ; Çukur, Tolga
Deep-learning models have enabled performance leaps in analysis of high-dimensional functional MRI (fMRI) data. Yet, many previous methods are suboptimally sensitive for contextual representations across diverse time scales. Here, we present BolT, a blood-oxygen-level-dependent transformer model, for analyzing multi-variate fMRI time series. BolT leverages a cascade of transformer encoders equipped with a novel fused window attention mechanism. Encoding is performed on temporally-overlapped windows within the time series to capture local representations. To integrate information temporally, cross-window attention is computed between base tokens in each window and fringe tokens from neighboring windows. To gradually transition from local to global representations, the extent of window overlap and thereby number of fringe tokens are progressively increased across the cascade. Finally, a novel cross-window regularization is employed to align high-level classification features across the time series. Comprehensive experiments on large-scale public datasets demonstrate the superior performance of BolT against state-of-the-art methods. Furthermore, explanatory analyses to identify landmark time points and regions that contribute most significantly to model decisions corroborate prominent neuroscientific findings in the literature.
Open Access
Connectivity analysis of an AUV network with OFDM based communications
(IEEE, 2017) Bereketli, A.; Tümçakır, M.; Yazgı, İ.; Yeni, B.; Köseoğlu, M.; Duman, Tolga M.
Autonomous underwater vehicle (AUV) networks play a crucial role in tactical, commercial, and scientific applications, where reliable and robust communication protocols are needed due to the challenging characteristics of the channel. With this motivation, connectivity of AUV networks in different regions with varying transducer characteristics are analyzed through simulations based on real-life orthogonal frequency division multiplexing (OFDM) based communication experiments over noisy and Doppler-distorted channels. Doppler compensation is performed according to the autocorrelation using the cyclic prefix. Using binary and quadrature phase shift keying (BPSK and QPSK) modulation schemes in conjunction with low density parity check (LDPC) coding, error rate levels are investigated through shallow water pond and at-sea experiments. It is shown that, the utilized transmission scheme is capable of correcting all bit errors among nearly one million bits transmitted up to a distance of 1 km, yielding a payload rate of 15.6 kbps with 4096 subcarriers and QPSK modulation. The simulations provide key parameters that must be taken into account in the design of scalable and connected AUV networks.
Open Access
Genetic influences on cortical structure and function
(2018-07) Demirayak, Pınar
Structure and function of the human central nervous system is determined by both genetic and environmental in uences. One of the fundamental quests in neuroscience studies is to determine to what degree each of these two factors in uence the development and function of the nervous system and, ultimately, human behavior. However, this is an inherently di cult problem to tackle as it is nearly impossible to tease apart the individual contributions of genes and environment, since they interact heavily throughout an organism's life. Recently, our understanding about the role that speci c genes play in the development of brain structure and function, has been greatly advanced by studies that combine genetic and neuroimaging methods to investigate congenital neurodevelopmental disorders. These studies of patients, that are homozygous for a speci c mutation, allow to single out contributions of individual genes in the neurodevelopmental process and have the potential to reveal gene-based alterations in cortical structure and function that can not be compensated by mechanisms of cortical plasticity or mitigating environmental e ects. In this thesis I pursue this promising approach further and investigate the e ects of three di erent single gene mutations on brain structure and function - namely RAD51, LAMC3 and HTRA2. Each of these genes is highly expressed during neurodevelopment, and each in uences cortical structure and function di erently. Overall, I nd that these genes are all highly associated with abnormal structural and functional connectivity patterns, however, and surprisingly, highly abnormal structure does not necessarily predict highly abnormal behavior.
Open Access
Resting-state network dysconnectivity in ADHD: a system-neuroscience-based meta-analysis
(Taylor and Francis, 2020) Sütçübaşı, B.; Metin, B.; Kurban, Mustafa Kerem; Metin, Z. E.; Beşer, B.; Sonuga-Barke, E.
Objectives: Neuroimaging studies report altered resting-state functional connectivity in attention deficit/hyperactivity disorder (ADHD) across multiple brain systems. However, there is inconsistency among individual studies. Methods: We meta-analyzed seed-based resting state studies of ADHD connectivity within and between four established resting state brain networks (default mode, cognitive control, salience, affective/motivational) using Multilevel Kernel Density Analysis method. Results: Twenty studies with 944 ADHD patients and 1121 controls were included in the analysis. Compared to controls, ADHD was associated with disrupted within-default mode network (DMN) connectivity – reduced in the core (i.e. posterior cingulate cortex seed) but elevated in the dorsal medial prefrontal cortex sub-system (i.e. temporal pole-inferior frontal gyrus). Connectivity was elevated between nodes in the cognitive control system. When the analysis was restricted to children and adolescents, additional reduced connectivity was detected between DMN and cognitive control and affective/motivational and salience networks. Conclusions: Our data are consistent with the hypothesis that paediatric ADHD is a DMN-dysconnectivity disorder with reduced connectivity both within the core DMN sub-system and between that system and a broad set of nodes in systems involved in cognition and motivation.