Sungheon Kim

Correlation between BOLD-fMRI and EEG signal changes in response to visual stimulus frequency in humans

The correlation between signals acquired using electroencephalography (EEG) and fMRI was investigated in humans during visual stimulation. Evoked potential EEG and BOLD fMRI data were acquired independently under similar conditions from eight subjects during stimulation by a checkerboard flashed at frequencies ranging from 2–12 Hz. The results indicate highly correlated changes in the strength of the EEG signal averaged over two occipital electrodes and the BOLD signal within the occipital lobe as a function of flash frequency for 7/8 subjects (average linear correlation coefficient of 0.76). Both signals peaked at approximately 8 Hz. For one subject the correlation coefficient was 0.20; the EEG signal peaked at 6 Hz and the BOLD signal peaked at 10 Hz. Overall, the EEG and BOLD signals, each averaged over 40‐sec stimulation periods, appear to be coupled linearly during visual stimulation by a flashing checkerboard. Magn Reson Med 49:108–114, 2003.

Influence of conductivity tensors on the scalp electrical potential: study with 2-D finite element models

The influence of conductivity tensor on the forward solution of electroencephalography was assessed in 2-D head models of a human subject. The conductivity tensors of different regions of the head were estimated from magnetic resonance-diffusion tensor images by linearly mapping the mean trace values to the published conductivity values. The anisotropic conductivity model was compared with the isotropic conductivity model in terms of the difference between the scalp potentials. The differences were measured by the cross correlation (CC) and the relative error (RE) between two models. We have also proposed a new measure, scaling-removed RE (SRRE) as a more effective indicator of the difference. The results with 354 individual dipole sources show that there are remarkable differences between the anisotropic conductivity tensor and the isotropic model (CC=0.96, RE=30.73% and SRRE=19.34%). Although the CC is high, the large RE and SRRE indicate that this difference may also affect the accuracy of inverse solutions in localizing the current dipole sources.

EEG distributed source imaging with a realistic finite-element head model

We have investigated electroencephalography (EEG) distributed source imaging with a realistic finite-element (FE) head model. The performance of different FE imaging methods was evaluated and compared in two- (2-D) and three-dimensional (3-D) simulation studies. The results demonstrate the feasibility of EEG distributed source imaging with FE head models using FE inverse methods. We also show that incorporating prior knowledge of sources significantly improves the inverse solutions. As an application of the EEG imaging methods, the sources of voluntary movement-related evoked potentials of the human brain were imaged. The reconstructed EEG sources show good spatial correlation with the functional magnetic imaging resonance (fMRI)-mapped motor activity in the brain, thus validating the FE inverse methods and adding temporal information to the fMRI-mapped regions of the brain.

Influence of conductivity tensors in the finite element model of the head on the forward solution of EEG

The influence of conductivity tensor on the forward solution of electroencephalography (EEG) was assessed in the 2D head models of a human subject. The conductivity tensors of the different regions of the head were estimated from the diffusion tensor images by scaling the mean trace values to the published conductivity values. The anisotropic conductivity model was compared with the isotropic conductivity model in terms of the difference between the scalp potentials. The differences were measured by the cross correlation (CC) and the relative error (RE) between two models. The results with 252 individual dipole source show the CC is 0.96 and the RE is 31.24%. Although the CC is high, the large RE indicates that there is a significant amount of asymmetrical magnitude change on the scalp potential by the conductivity tensor. It indicates that the influence will also affect the inverse solutions of localizing the current dipole source.

Application of independent component analysis with mixture density model to localize brain alpha activity in fMRI and EEG

Independent component analysis (ICA) is an approach to solve the blind source separation problem. In the original and extended versions of ICA, nonlinearity functions are fixed to have specific density forms such as super‐Gaussian or sub‐Gaussian, thereby limiting their performance when sources with different classes of densities are mixed in multichannel data. In this article, we have incorporated a mixture density model such that no assumption about source density would be required. We show that this leads to better source separation due to increased flexibility in handling source‐ densities with flexible parametric nonlinearity. The algorithm was validated through simulation studies and its performance was compared to other versions of ICA. The modified mixture density ICA was then applied to functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) data to localize independent sources of alpha activity in the human brain. A good spatial correlation was found in the spatial distribution of alpha sources derived independently from fMRI and EEG, suggesting that spontaneous alpha rhythm can be imaged by fMRI using ICA without concurrent acquisition of EEG.

DT-MRI regularization using 3D nonlinear gradient vector flow anisotropic diffusion

In DT-MRI, diffusion-weighted multislice echoplanar images (EPIs) are processed to represent water diffusion characteristics as a diffusion tensor, reflecting the amount of diffusion in 3D. However imaging quality is generally compromised by several factors including the number of imaging slices, averages, diffusion sensitization steps (b-values), voxel size, and gradient directions, resulting in suboptimal SNR. In this study, we focus on improving imaging quality and SNR by denoising and reducing systematic and random errors through nonlinear anisotropic regularization. The raw EPIs are directly regularized through a newly proposed nonlinear anisotropic diffusion regularization method in 3D utilizing the gradient vector flow fields and its performance is compared to conventional 2D and vector-valued 2D anisotropic regularization methods. The effects of these variants of anisotropic regularization are examined through the maps of color-coded fractional anisotropy and tracked neural fibers. The results show that DT-MRI regularization using the proposed 3D anisotropic diffusion significantly improves the quality of fiber tracking and diffusion indices such as fractional anisotropy.

Multispectral excitation of time-resolved fluorescence of biological compounds: variation of fluorescence lifetime with excitation and emission wavelengths

We investigated the time-resolved fluorescence spectra of tissue structural proteins elastin and collagen I for different excitation wavelengths to assess the usefulness of multispectral excitation for characterization of fluorescent structural proteins in tissue. Laser excitation pulses (4 ns FWHM) at wavelengths between 337 nm and 422 nm were directed to the samples with a multifiber fluorescence probe. The fluorescence emission was measured with a photomultiplier and deconvoluted from the emission pulse with an algorithm based on the Laguerre expansion of kernels technique. A multiexponential approximation of the intrinsic fluorescence decay was computed using a global approach in which the decay constants were held fixed across wavelengths of emission.

Estimation of multiple fiber orientations from diffusion tensor MRI

A novel method has been investigated to identify the orientations of multiple fiber tracts in a voxel of diffusion tensor MRI (DT-MRI). Conventionally, white matter fiber tracking has been performed based on the assumption that the major eigenvector represents the orientation of the fiber tract. However, it has been acknowledged widely that the assumption is valid only if there is a unidirectional single fiber tract per voxel. Recent attempts to resolve the problem of multiple fibers often require long acquisition times and strong gradients with greater than 50 gradient-directions. Our approach is to keep the same imaging sequence as conventional DT-MRI with a relatively small (six) number of diffusion gradient directions, and to apply independent component analysis technique to separate the raw data into multiple data sets corresponding to multiple fiber tracts respectively. A simulation study was carried out to test the method, followed by application to human data.

Spatial correspondence of brain alpha activity component in fMRI and EEG

This paper presents a new approach to investigate the spatial correlation of brain alpha activity in functional magnetic resonance imaging (fMRI) and electroencephalography (EEG). To avoid potential problems of simultaneous fMRI and EEG acquisitions in imaging brain alpha activity, data from each modality were acquired separately under a “three conditions” setup where one of the conditions involved closing eyes and relaxing, thus making it conducive to generation of alpha activity. The other two conditions -- eyes open in a lighted room or engaged in a mental arithmetic task, were designed to attenuate alpha activity. Using the Mixture Density Independent Component Analysis (MD-ICA) that incorporates flexible non-linearity functions into the conventional ICA framework, we could identify the spatiotemporal components of fMRI activations and EEG activities associated with the alpha rhythm. The sources of the individual EEG alpha activity component were localized by a Maximum Entropy (ME) method that solves an inverse problem in the framework of a classical four-sphere head model. The resulting dipole sources of EEG alpha activity were spatially transformed to 3D MRIs of the subject and compared to fMRI ICA-determined alpha activity maps.