Estimating Dynamic Connectivity States in fMRI Using Regime-Switching Factor Models
KAUST DepartmentComputer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division
Permanent link to this recordhttp://hdl.handle.net/10754/626407
MetadataShow full item record
AbstractWe consider the challenges in estimating state-related changes in brain connectivity networks with a large number of nodes. Existing studies use sliding-window analysis or time-varying coefficient models which are unable to capture both smooth and abrupt changes simultaneously, and rely on ad-hoc approaches to the high-dimensional estimation. To overcome these limitations, we propose a Markov-switching dynamic factor model which allows the dynamic connectivity states in functional magnetic resonance imaging (fMRI) data to be driven by lower-dimensional latent factors. We specify a regime-switching vector autoregressive (SVAR) factor process to quantity the time-varying directed connectivity. The model enables a reliable, data-adaptive estimation of change-points of connectivity regimes and the massive dependencies associated with each regime. We develop a three-step estimation procedure: 1) extracting the factors using principal component analysis, 2) identifying connectivity regimes in a low-dimensional subspace based on the factor-based SVAR model, 3) constructing high-dimensional state connectivity metrics based on the subspace estimates. Simulation results show that our estimator outperforms K-means clustering of time-windowed coefficients, providing more accurate estimate of time-evolving connectivity. It achieves percentage of reduction in mean squared error by 60% when the network dimension is comparable to the sample size. When applied to resting-state fMRI data, our method successfully identifies modular organization in resting-state networks in consistency with other studies. It further reveals changes in brain states with variations across subjects and distinct large-scale directed connectivity patterns across states.
CitationTing C-M, Ombao H, Samdin SB, Salleh S-H (2017) Estimating Dynamic Connectivity States in fMRI Using Regime-Switching Factor Models. IEEE Transactions on Medical Imaging: 1–1. Available: http://dx.doi.org/10.1109/TMI.2017.2780185.
Showing items related by title, author, creator and subject.
MobiSeq: De novo SNP discovery in model and non-model species through sequencing the flanking region of transposable elementsRey-Iglesia, Alba; Gopalakrishan, Shyam; Carøe, Christian; Alquezar-Planas, David E.; Ahlmann Nielsen, Anne; Röder, Timo; Bruhn Pedersen, Lene; Næsborg-Nielsen, Christina; Sinding, Mikkel-Holger S.; Fredensborg Rath, Martin; Li, Zhipeng; Petersen, Bent; Gilbert, M. Thomas P.; Bunce, Michael; Mourier, Tobias; Hansen, Anders Johannes (Molecular Ecology Resources, Wiley, 2018-12-21) [Article]In recent years, the availability of reduced representation library (RRL) methods has catalysed an expansion of genome-scale studies to characterize both model and non-model organisms. Most of these methods rely on the use of restriction enzymes to obtain DNA sequences at a genome-wide level. These approaches have been widely used to sequence thousands of markers across individuals for many organisms at a reasonable cost, revolutionizing the field of population genomics. However, there are still some limitations associated with these methods, in particular the high molecular weight DNA required as starting material, the reduced number of common loci among investigated samples, and the short length of the sequenced site-associated DNA. Here, we present MobiSeq, a RRL protocol exploiting simple laboratory techniques, that generates genomic data based on PCR targeted enrichment of transposable elements and the sequencing of the associated flanking region. We validate its performance across 103 DNA extracts derived from three mammalian species: grey wolf (Canis lupus), red deer complex (Cervus sp.) and brown rat (Rattus norvegicus). MobiSeq enables the sequencing of hundreds of thousands loci across the genome and performs SNP discovery with relatively low rates of clonality. Given the ease and flexibility of MobiSeq protocol, the method has the potential to be implemented for marker discovery and population genomics across a wide range of organisms—enabling the exploration of diverse evolutionary and conservation questions.
Towards macroscopic modeling of electro-thermo-mechanical couplings in PEDOT/PSS: Modeling of moisture absorption kineticsZhanshayeva, Lyazzat (2018-07) [Thesis]
Advisor: Lubineau, Gilles
Committee members: Farooq, Aamir; Baran, DeryaOrganic conducting polymer, poly(3,4-ethylene dioxythiophene)-poly(styrene sulfonate) (PEDOT:PSS), is widely recognized for its electro-actuation mechanism and is used in flexible electronics. Its high potential as actuator is based on a strong coupling between chemical, mechanical and electrical properties which directly depends on external stimuli. There is no model today to describe the interplay between moisture absorption, mechanical expansion and electrical stimulus. Elucidating the role of each component in the effective actuation properties is needed to further optimize and tailor such materials. The objective of this thesis is to develop a macroscopic model to describe water sorption kinetics of the PEDOT:PSS film. We used gravimetric analysis of pure PEDOT:PSS film of three different thicknesses to investigate absorption kinetics over a broad range of temperatures and relative humidity. Our results revealed that the moisture uptake of PEDOT:PSS film does not follow Fickian diffusion law due to the retained amount of water after desorption process. We used an existing diffusionreaction model to describe this behavior, and COMSOL Multiphysics and MATLAB software programs to implement it. We observed that the generic model we used in our work could predict polymer behavior with 95% accuracy. However, our model was not able to properly represent the data at very high relative humidity at low temperature, which was attributed to the excessive swelling of the film. Also, we examined a relation between the moisture content of PEDOT:PSS and its mechanical strain and electrical conductivity. The results presented here are the first step towards a general multiphysics electro-thermo-mechanical description of PEDOT:PSS based actuators.
Control of modeling error in calibration and validation processes for predictive stochastic modelsOden, J. Tinsley; Prudhomme, Serge (International Journal for Numerical Methods in Engineering, Wiley, 2010-09-28) [Article]The idea of adaptive control of modeling error is expanded to include ideas of statistical calibration, validation, and uncertainty quantification. © 2010 John Wiley & Sons, Ltd.