Showing papers by "Tulay Adali published in 2004"
TL;DR: This method is applied to data from experiments designed to stimulate visual cortex, motor cortex or both visual and motor cortices, and several intergroup and intragroup metrics are proposed for assessing the utility of the components for comparisons of group ICA data.
170 citations
TL;DR: In this article, the authors present an approach for constrained or semi-blind independent component analysis (ICA) analysis of fMRI data of an auditory oddball paradigm using prior information with weak constraints.
135 citations
17 May 2004
TL;DR: A number of complex nonlinear functions are proposed for the independent component analysis of complex-valued data and their efficiency in generating the higher order statistics needed for ICA is shown.
Abstract: A number of complex nonlinear functions are proposed for the independent component analysis (ICA) of complex-valued data. We discuss the properties of these nonlinearities and show their efficiency in generating the higher order statistics needed for ICA.
61 citations
17 May 2004
TL;DR: Simulation results with an accurate receiver model and all-order PMD show the success of the MAP equalizer in reducing the ISI due to PMD, and that the analytical conditional PDF the authors derive provides a good match to the actual distribution.
Abstract: A maximum a posteriori equalizer is presented for optical communication systems Assuming that the span of the intersymbol interference (ISI) does not extend beyond the neighboring bits - as is typically the case for the distortion introduced by polarization mode dispersion (PMD) - we derive the conditional probability distribution function in the electrical domain in the presence of PMD and amplifier spontaneous emission dominated noise Simulation results with an accurate receiver model and all-order PMD show the success of the MAP equalizer in reducing the ISI due to PMD, and that the analytical conditional PDF we derive provides a good match to the actual distribution
9 citations
15 Apr 2004
TL;DR: It is shown that the complex infomax using these efficient nonlinearities demonstrates superior performance compared to analysis of the magnitude data with either ICA or linear regression, and provides a potentially powerful method for the analysis of fMRI data.
Abstract: Independent component analysis (ICA) for separating complex-valued sources is needed for convolutive source-separation in the frequency domain, or for performing source separation on complex-valued data. Functional magnetic resonance imaging (fMRI) is a technique that produces complex-valued data; however the vast majority of fMRI analyses utilize only magnitude images due in large part to the difficulty of developing a temporal phase model. We have successfully applied ICA to complex fMRI data but there is a need to further optimize the complex ICA. We recently proposed a number of complex nonlinear functions for ICA of complex valued data. We apply two of these functions to fMRI data and examine the properties of these nonlinearities and their efficiency in generating the higher order statistics needed for ICA. We show that the complex infomax using these efficient nonlinearities demonstrates superior performance compared to analysis of the magnitude data with either ICA or linear regression. Complex ICA thus provides a potentially powerful method for the analysis of fMRI data.
7 citations
29 Sep 2004
TL;DR: It is shown that ICA with a smooth filtering scheme can improve the estimation of the smooth image sources from a mixture of images, as well asThe estimation of a smooth visual activation map in a hybrid functional magnetic resonance imaging (fMRI) data set.
Abstract: In this contribution, we propose a feature selective filtering scheme for independent component analysis (ICA) to improve the estimation of the sources of interest (SOI), i.e., sources that have certain desired features in their sample space. As an example, we show that ICA with a smooth filtering scheme can improve the estimation of the smooth image sources from a mixture of images, as well as the estimation of a smooth visual activation map in a hybrid functional magnetic resonance imaging (fMRI) data set. Hence, the technique can potentially be used in the analysis of fMRI data to improve the ICA estimation of functional activation regions that are expected to be smooth
1 citations
01 Jun 2004
TL;DR: This paper poses the TEQ design problem completely in the frequency domain by defining a frequency-domain weighted least-squares cost function and presents corresponding algorithms for these two objectives and simulation results, which show the shortening (channel or channel and echo) as well as noise suppression objectives effectively.