An Efficient Hybrid Wavelet-ICA algorithm for Analyzing Simulated fMRI Data in Noisy Environment

Abstract

The performance ofICA algorithms in correct separation sequences such as system noise, patient motion, physiological ofindependent sources can be highly affected by existence ofnoises noise (breathing and heart beat), ghost artifact in fMRI in the observation data. In this paper a hybrid Wavelet-ICA method images, long-term instability of the scanner baseline, local for improving the functionality of noise free ICA algorithms in changes in magnetic field due to short term scanner noisy environment is proposed. At first the robustness of two most instabilities, and slow phase variation in MR images has made frequent ICA algorithms, named Fast ICA and Information . ' maximization ICA, for extracting true activated spatial and the detectlon of bram acivations a challengig task by temporal sources offMRI signals in the presence of different noise different algorithms. Many different methods have been levels are evaluated. These algorithms are applied on simulated proposed for analyzing fMRI signals which can be categorized fMRI datasets consisting of different activated sources with various into two approaches: model based analysis such as General temporal patterns, different levels of activation, trend and noise. linear model, GLM and data driven analysis such as principal Then, a hybrid wavelet-Fast ICA model to transform the signals component analysis, PCA, Factor analysis, FA and into a domain, allowing for simultaneous un-mixing and wavelet Independent component analysis (ICA) [1 ]. Among data based de-noising is proposed. As the results show this combination * * t has significantly improved the sensitivity of extracted sources in drivesndmetds, the mowrf estCa method whIc different SNR levels, in particular in low SNR s. To measure the assumes idep enddcy of the sources that constitute feitRI accuracy of source separation, the correlation coefficients between signal. ICA has been used for fMRI analysis to separate either extracted activation signals and simulated temporal patterns are spatially independent brain maps or their activation time also measured. As the results suggest the proposed hybrid method is courses (Spatial ICA) and temporally independent time more robust in comparison with noise free ICA for noisy courses and the corresponding spatial maps (Temporal ICA) observation in extracting more accurate independent sources. from the fMRI signal. There have been many algorithms proposed to implement the ICA method but the most Index Term: Functional magnetic resonance imaging, frequently used for analyzing of them are Fast ICA and Independent Component Analysis, Infomax, Wavelet, Infomax [1,2]. One of the main factors which influence the Wiener Filter Fast ICA, PICA, SNR performances of different ICA algorithms is the amount of noise in the observation data, here fMRI datasets, which I. INTRODUCTION makes the estimation of un-mixing matrix more complicated. s a noninvasive technique to measure brain activity, Furthermore, the estimation of true dimension of independent fMRI based on Blood Oxygenation Le...

Spatial and Transform Domain Filtering Method for Image De-noising: A Review

“…These algorith ms are applied on simu lated FM RI dataset consisting of different activated sources with various temporal pattern, different level of activation, trend and noise. Then a hybrid wavelet-Fast ICA model to transform the signals into a do main, allo wing for simu ltaneous un-mixing and wavelet basedde-noising is proposed [47].In Th is paper authors presented a novel adaptive method of image denoising based on the dual-tree co mplex wavelet transform (DTCWT) and independent component analysis (ICA). This method extracted the highfrequency component of the image with the DTCWT, then combining with the princip le of ICA v irtual observed noise channel de-noise.…”

Wavelet-based fMRI analysis: 3-D denoising, signal separation, and validation metrics

“…Previously, there have been a few attempts to combine the power of wavelet processing (in order to decorrelate data), with a widely applied fMRI analysis technique such as ICA (Azzerboni et al, 2004; Azzerboni et al, 2005; Boroomand et al, 2007; Johnson et al, 2007). A more recent technique that utilizes a combination of wavelet-domain ICA with Wiener filtering of wavelet coefficients is introduced in (Boroomand et al, 2007). They utilize a hybrid approach in order to remove effects of noise on accurate separation of data using ICA.…”

“…It may be intuitive to say that the application of Wiener filter in wavelet domain improves the performance of time-domain Wiener filtering, introducing least amount of temporal smoothing, eventually resulting in a good estimate of noise in a least squares sense. (Boroomand et al, 2007; Xu et al, 1994) did not report the false discovery rate (FDR) that is expected to be high due to the aforementioned low-smoothing characteristics of proposed wavelet-wiener filter, and neither compared their results for varying noise frames against other denoising methods such as Wiener filtering on observation data in the spatial domain.…”