ICA-based denoising for ASL perfusion imaging.

Arterial Spin Labelling (ASL) imaging derives a perfusion image by tracing the accumulation of magnetically labeled blood water in the brain. As the image generated has an intrinsically low signal to noise ratio (SNR), multiple measurements are routinely acquired and averaged, at a penalty of increased scan duration and opportunity for motion artefact. However, this strategy alone might be ineffective in clinical settings where the time available for acquisition is limited and patient motion are increased. This study investigates the use of an Independent Component Analysis (ICA) approach for denoising ASL data, and its potential for automation. 72 ASL datasets (pseudo-continuous ASL; 5 different post-labeling delays: 400, 800, 1200, 1600, 2000 m s; total volumes = 60) were collected from thirty consecutive acute stroke patients. The effects of ICA-based denoising (manual and automated) where compared to two different denoising approaches, aCompCor, a Principal Component-based method, and Enhancement of Automated Blood Flow Estimates (ENABLE), an algorithm based on the removal of corrupted volumes. Multiple metrics were used to assess the changes in the quality of the data following denoising, including changes in cerebral blood flow (CBF) and arterial transit time (ATT), SNR, and repeatability. Additionally, the relationship between SNR and number of repetitions acquired was estimated before and after denoising the data. The use of an ICA-based denoising approach resulted in significantly higher mean CBF and ATT values (p < 0.001), lower CBF and ATT variance (p < 0.001), increased SNR (p < 0.001), and improved repeatability (p < 0.05) when compared to the raw data. The performance of manual and automated ICA-based denoising was comparable. These results went beyond the effects of aCompCor or ENABLE. Following ICA-based denoising, the SNR was higher using only 50% of the ASL-dataset collected than when using the whole raw data. The results show that ICA can be used to separate signal from noise in ASL data, improving the quality of the data collected. In fact, this study suggests that the acquisition time could be reduced by 50% without penalty to data quality, something that merits further study. Independent component classification and regression can be carried out either manually, following simple criteria, or automatically.

Cortical structural involvement and cognitive dysfunction in early Parkinson's disease.

Magnetic resonance imaging (MRI) studies in early Parkinson's disease (PD) have shown promise in the detection of disease-related brain changes in the white and deep grey matter. We set out to establish whether intrinsic cortical involvement in early PD can be detected with quantitative MRI. We collected a rich, multi-modal dataset, including diffusion MRI, T1 relaxometry and cortical morphometry, in 20 patients with early PD (disease duration, 1.9 ± 0.97 years, Hoehn & Yahr 1-2) and in 19 matched controls. The cortex was reconstructed using FreeSurfer. Data analysis employed linked independent component analysis (ICA), a novel data-driven technique that allows for data fusion and extraction of multi-modal components before further analysis. For comparison, we performed standard uni-modal analysis with a general linear model (GLM). Linked ICA detected multi-modal cortical changes in early PD (p = 0.015). These comprised fractional anisotropy reduction in dorsolateral prefrontal, cingulate and premotor cortex and the superior parietal lobule, mean diffusivity increase in the mesolimbic, somatosensory and superior parietal cortex, sparse diffusivity decrease in lateral parietal and right prefrontal cortex, and sparse changes to the cortex area. In PD, the amount of cortical dysintegrity correlated with diminished cognitive performance. Importantly, uni-modal analysis detected no significant group difference on any imaging modality. We detected microstructural cortical pathology in early PD using a data-driven, multi-modal approach. This pathology is correlated with diminished cognitive performance. Our results indicate that early degenerative processes leave an MRI signature in the cortex of patients with early PD. The cortical imaging findings are behaviourally meaningful and provide a link between cognitive status and microstructural cortical pathology in patients with early PD.

Altered and asymmetric default mode network activity in a "hypnotic virtuoso": an fMRI and EEG study.

Very highly hypnotizable subjects are rare, easily induced, and able to manifest the whole spectrum of hypnotic phenomena, including post-hypnotic amnesia. The aim of this study was to detect and localize by means of quantitative functional MRI and EEG changes in cortical activity during hypnosis induction and deep "pure hypnosis" in a hypnotic "virtuoso" subject. We focused on areas forming the default mode network (DMN), since previous studies found that very highly suggestible subjects in hypnosis showed decreased activity in anterior DMN. During undisturbed hypnosis, our "virtuoso" subject showed not only detectable changes in DMN, but also peculiar activations of non-DMN areas and hemispheric asymmetries of frontal lobe connectivity. Our findings confirm that hypnosis is associated with significant modulation of connectivity and activity which involve the DMN but are not limited to it, depending on the depth of the hypnotic state, the type of mental content and emotional involvement.

Impact of automated ICA-based denoising of fMRI data in acute stroke patients.

Different strategies have been developed using Independent Component Analysis (ICA) to automatically de-noise fMRI data, either focusing on removing only certain components (e.g. motion-ICA-AROMA, Pruim et al., 2015a) or using more complex classifiers to remove multiple types of noise components (e.g. FIX, Salimi-Khorshidi et al., 2014 Griffanti et al., 2014). However, denoising data obtained in an acute setting might prove challenging: the presence of multiple noise sources may not allow focused strategies to clean the data enough and the heterogeneity in the data may be so great to critically undermine complex approaches. The purpose of this study was to explore what automated ICA based approach would better cope with these limitations when cleaning fMRI data obtained from acute stroke patients. The performance of a focused classifier (ICA-AROMA) and a complex classifier (FIX) approaches were compared using data obtained from twenty consecutive acute lacunar stroke patients using metrics determining RSN identification, RSN reproducibility, changes in the BOLD variance, differences in the estimation of functional connectivity and loss of temporal degrees of freedom. The use of generic-trained FIX resulted in misclassification of components and significant loss of signal (< 80%), and was not explored further. Both ICA-AROMA and patient-trained FIX based denoising approaches resulted in significantly improved RSN reproducibility (p < 0.001), localized reduction in BOLD variance consistent with noise removal, and significant changes in functional connectivity (p < 0.001). Patient-trained FIX resulted in higher RSN identifiability (p < 0.001) and wider changes both in the BOLD variance and in functional connectivity compared to ICA-AROMA. The success of ICA-AROMA suggests that by focusing on selected components the full automation can deliver meaningful data for analysis even in population with multiple sources of noise. However, the time invested to train FIX with appropriate patient data proved valuable, particularly in improving the signal-to-noise ratio.

Individual Thresholding of Voxel-based Functional Connectivity Maps. Estimation of Random Errors by Means of Surrogate Time Series.

INTRODUCTION: This article is part of the Focus Theme of Methods of Information in Medicine on "Biosignal Interpretation: Advanced Methods for Neural Signals and Images". BACKGROUND: Voxel-based functional connectivity analysis is a common method for resting state fMRI data. However, correlations between the seed and other brain voxels are corrupted by random estimate errors yielding false connections within the functional connectivity map (FCmap). These errors must be taken into account for a correct interpretation of single-subject results. OBJECTIVES: We estimated the statistical range of random errors and propose two methods for an individual setting of correlation threshold for FCmaps. METHODS: We assessed the amount of random errors by means of surrogate time series and described its distribution within the brain. On the basis of these results, the FCmaps of the posterior cingulate cortex (PCC) from 15 healthy subjects were thresholded with two innovative methods: the first one consisted in the computation of a unique (global) threshold value to be applied to all brain voxels, while the second method is to set a different (local) threshold of each voxel of the FCmap. RESULTS: The distribution of random errors within the brain was observed to be homogeneous and, after thresholding with both methods, the default mode network areas were well identifiable. The two methods yielded similar results, however the application of a global threshold to all brain voxels requires a reduced computational load. The inter-subject variability of the global threshold was observed to be very low and not correlated with age. Global threshold values are also almost independent from the number of surrogates used for their computation, so the analyses can be optimized using a reduced number of surrogate time series. CONCLUSIONS: We demonstrated the efficacy of FCmaps thresholding based on random error estimation. This method can be used for a reliable single-subject analysis and could also be applied in clinical setting, to compute individual measures of disease progression or quantitative response to pharmacological or rehabilitation treatments.