Jiří Jan

Exploring task-related variability in fMRI data using fluctuations in power spectrum of simultaneously acquired EEG

BACKGROUND The paper deals with joint analysis of fMRI and scalp EEG data, simultaneously acquired during event-related oddball experiment. The analysis is based on deriving temporal sequences of EEG powers in individual frequency bands for the selected EEG electrodes and using them as regressors in the general linear model (GLM). NEW METHOD Given the infrequent use of EEG spectral changes to explore task-related variability, we focused on the aspects of parameter setting during EEG regressor calculation and searched for such parameters that can detect task-related variability in EEG-fMRI data. We proposed a novel method that uses relative EEG power in GLM. RESULTS Parameter, the type of power value, has a direct impact as to whether task-related variability is detected or not. For relative power, the final results are sensitive to the choice of frequency band of interest. The electrode selection also has certain impact; however, the impact is not crucial. It is insensitive to the choice of EEG power series temporal weighting step. Relative EEG power characterizes the experimental task activity better than the absolute power. Absolute EEG power contains broad spectrum component. Task-related relative power spectral formulas were derived. COMPARISON WITH EXISTING METHODS For particular set of parameters, our results are consistent with previously published papers. Our work expands current knowledge by new findings in spectral patterns of different brain processes related to the experimental task. CONCLUSIONS To make analysis to be sensitive to task-related variability, the parameters type of power value and frequency band should be set properly.

Optimization methods for registration of multimodal images of retina

Registration of multimodal images of retina is essential for correct diagnosis of the optic nerve head and retina. For reliable vessel segmentation, it is also important to use information from both, colour photographs and Heidelberg Retina Tomograph (HRT) scans. Mutual information was tested as a coincidence measure and has proven functional and reliable. This paper compares several methods for finding the correct transformation parameters.

Extended time-frequency granger causality for evaluation of functional network connectivity in event-related FMRI data

In this article, we show that adaptive multivariate autoregressive (AMVAR) modeling accompanied by proper estimation of the delay and the width of hemodynamic response function is an effective technique for evaluation of spectral Granger causality among different functional brain networks identified by independent component analysis from event-related fMRI data. The entire concept is demonstrated on 28 subjects auditory oddball fMRI data.

3D simulation of diffraction in ultrasonic computed tomography

The contribution presents further results in developing the exact means for simulating the realistic situation in the USCT (ultrasonic computed tomography) imaging system, aiming both at evaluating the approximations used in the existing USCT image reconstruction methods as to their precision and also (in a longer perspective) at iterative improvement of the obtained images via continuum mechanics based feedback. The mathematical models, generalised in comparison with [1], emerging from the transparent physical background, are presented for inhomogeneous media incorporating both the object tissue and the surrounding fluid. The equations are already general enough to employ complex nonlinear phenomena in three-dimensional space; and linearised 3D simulations (giving rise to wave equation, WE) have been performed enabling first conclusions on the feasibility of this approach with respect to the available computing resources. Some of the results of the numerical solution of the WE in 3D by means of the finite-element method show in local detail the diffraction phenomena on acoustic-impedance inhomogeneities. The spatial extent of the simulations is basically limited only by the available computing resources. The hardware requirements and related practical limitations are mentioned together with a few examples of presently available results. Together with conclusions, further perspectives of this branch of the USCT research are suggested.

Stable Scalp EEG Spatiospectral Patterns Across Paradigms Estimated by Group ICA

Electroencephalography (EEG) oscillations reflect the superposition of different cortical sources with potentially different frequencies. Various blind source separation (BSS) approaches have been developed and implemented in order to decompose these oscillations, and a subset of approaches have been developed for decomposition of multi-subject data. Group independent component analysis (Group ICA) is one such approach, revealing spatiospectral maps at the group level with distinct frequency and spatial characteristics. The reproducibility of these distinct maps across subjects and paradigms is relatively unexplored domain, and the topic of the present study. To address this, we conducted separate group ICA decompositions of EEG spatiospectral patterns on data collected during three different paradigms or tasks (resting-state, semantic decision task and visual oddball task). K-means clustering analysis of back-reconstructed individual subject maps demonstrates that fourteen different independent spatiospectral maps are present across the different paradigms/tasks, i.e. they are generally stable.

Evaluation of functional network connectivity in event-related FMRI data based on ICA and time-frequency granger causality

In this article we show that Adaptive Multivariate Autoregressive (AMVAR) modeling accompanied by proper preprocessing is an effective technique for evaluation of spectral Granger causality among functional brain networks identified by independent component analysis from eventrelated fMRI data.

Tissue models and speckle reduction in medical ultrasound images

This paper presents a new method for speckle noise reduction in medical ultrasound images. It is based on the statistical description of the envelope of ultrasound signal by the virtue of the Nakagami-m distribution. Parameter of this distribution is used to adjust an adaptive filter.

Stable EEG Spatiospectral Sources Using Relative Power as Group-ICA Input

Within the last decade, various blind source separation algorithms (BSS) isolating distinct EEG oscillations were derived and implemented. Group Independent Component Analysis (group-ICA) is a promising tool for decomposing spatiospectral EEG maps across multiple subjects. However, researchers are faced with many preprocessing options prior to performing group-ICA, which potentially influences the results. To examine the influence of preprocessing steps, within this article we compare results derived from group-ICA using the absolute power of spatiospectral maps and the relative power of spatiospectral maps. Within a previous study, we used K-means clustering to demonstrate group-ICA of absolute power spatiospectral maps generates sources which are stable across different paradigms (i.e. resting-state, semantic decision, visual oddball) Within the current study, we compare these maps with those obtained using relative power of spatiospectral maps as input to group-ICA. We find that relative EEG power contains 10 stable spatiospectral patterns which were similar to those observed using absolute power as inputs. Interestingly, relative power revealed two γ-band (20–40 Hz) patterns which were present across 3 paradigms, but not present using absolute power. This finding suggests that relative power potentially emphasizes low energy signals which are obscured by the high energy low frequency which dominates absolute power measures.

Spatial-temporal-spectral EEG patterns of BOLD functional network connectivity dynamics

OBJECTIVE Growing interest in the examination of large-scale brain network functional connectivity dynamics is accompanied by an effort to find the electrophysiological correlates. The commonly used constraints applied to spatial and spectral domains during electroencephalogram (EEG) data analysis may leave part of the neural activity unrecognized. We propose an approach that blindly reveals multimodal EEG spectral patterns that are related to the dynamics of the BOLD functional network connectivity. APPROACH The blind decomposition of EEG spectrogram by parallel factor analysis has been shown to be a useful technique for uncovering patterns of neural activity. The simultaneously acquired BOLD fMRI data were decomposed by independent component analysis. Dynamic functional connectivity was computed on the components time series using a sliding window correlation, and between-network connectivity states were then defined based on the values of the correlation coefficients. ANOVA tests were performed to assess the relationships between the dynamics of between-network connectivity states and the fluctuations of EEG spectral patterns. MAIN RESULTS We found three patterns related to the dynamics of between-network connectivity states. The first pattern has dominant peaks in the alpha, beta, and gamma bands and is related to the dynamics between the auditory, sensorimotor, and attentional networks. The second pattern, with dominant peaks in the theta and low alpha bands, is related to the visual and default mode network. The third pattern, also with peaks in the theta and low alpha bands, is related to the auditory and frontal network. SIGNIFICANCE Our previous findings revealed a relationship between EEG spectral pattern fluctuations and the hemodynamics of large-scale brain networks. In this study, we suggest that the relationship also exists at the level of functional connectivity dynamics among large-scale brain networks when no standard spatial and spectral constraints are applied on the EEG data.

46. Could it be possible to distinguish bending and crossing fibers in diffusion MRI data

After diffusion tensor imaging (DTI) model (Basser et al., 1994), several approaches which are able to detect two or more crossing fibers in diffusion MRI (dMRI) data have been invented (e.g. Q-ball imaging, ball and stick model) (Tuch, 2004; Behrens et al., 2003). After that, some fiber bundles which had not been seen with DTI model were suddenly observed (e.g. in corpus callosum). Although it brought an improvement it seems that about 50% of detected fiber bundles are false positive results after tractography (Ciccarelli et al., 2008). One crucial problem is that tractography cannot decide if the bundles are crossing or bending because models are not estimating bending-tensor. For 2 crossing fibers, the tractography algorithm can trace from one point to three different places. For 2 bending fibers, there is only one possible way. We would like to introduce how the difference between dMRI data coming from crossing or bending fibers could be detected. Imagine a population of water molecules in the centre of crossing or bending and some applied gradient of diffusion measurement. For crossing fibers, the population of molecules can diffuse in all directions of fiber spreading, thus the phase of molecules can be affected by the whole gradient range. Contrary for one bending fiber, the population can diffuse only in directions of the fiber, thus the phase can be affected only by the narrower gradient range. It applies similarly for second bended fiber. From this point of view, phase distributions should differ for crossing and bending fibers respectively also resulting dMRI data should differ. For this statement testing, the dMRI data simulator which generates dMRI data based on Brownian motion of water molecules inside and outside axons per one voxel volume was created. Although there is several technical problems and aspects (e.g. periodic character of gradient space phase distribution) we are looking for sequence settings of dMRI measurement where the dMRI data would be statistically significantly different for crossing and bending fiber geometries. Acknowledgement Computational resources were provided by the MetaCentrum under the program LM2010005 and the CERIT-SC under the program Centre CERIT Scientific Cloud, part of the Operational Program Research and Development for Innovations, Reg. No. CZ.1.05/3.2.00/08.0144.