Statistical analyses of motion-corrupted MRI relaxometry data computed from multiple scans.

BACKGROUND: Consistent noise variance across data points (i.e. homoscedasticity) is required to ensure the validity of statistical analyses of MRI data conducted using linear regression methods. However, head motion leads to degradation of image quality, introducing noise heteroscedasticity into ordinary-least square analyses. NEW METHOD: The recently introduced QUIQI method restores noise homoscedasticity by means of weighted least square analyses in which the weights, specific for each dataset of an analysis, are computed from an index of motion-induced image quality degradation. QUIQI was first demonstrated in the context of brain maps of the MRI parameter R2 * , which were computed from a single set of images with variable echo time. Here, we extend this framework to quantitative maps of the MRI parameters R1, R2 * , and MTsat, computed from multiple sets of images. RESULTS: QUIQI restores homoscedasticity in analyses of quantitative MRI data computed from multiple scans. QUIQI allows for optimization of the noise model by using metrics quantifying heteroscedasticity and free energy. COMPARISON WITH EXISTING METHODS: QUIQI restores homoscedasticity more effectively than insertion of an image quality index in the analysis design and yields higher sensitivity than simply removing the datasets most corrupted by head motion from the analysis. CONCLUSION: QUIQI provides an optimal approach to group-wise analyses of a range of quantitative MRI parameter maps that is robust to inherent homoscedasticity.

In vivo Estimation of Axonal Morphology From Magnetic Resonance Imaging and Electroencephalography Data.

Purpose: We present a novel approach that allows the estimation of morphological features of axonal fibers from data acquired in vivo in humans. This approach allows the assessment of white matter microscopic properties non-invasively with improved specificity. Theory: The proposed approach is based on a biophysical model of Magnetic Resonance Imaging (MRI) data and of axonal conduction velocity estimates obtained with Electroencephalography (EEG). In a white matter tract of interest, these data depend on (1) the distribution of axonal radius [P(r)] and (2) the g-ratio of the individual axons that compose this tract [g(r)]. P(r) is assumed to follow a Gamma distribution with mode and scale parameters, M and θ, and g(r) is described by a power law with parameters α and ß. Methods: MRI and EEG data were recorded from 14 healthy volunteers. MRI data were collected with a 3T scanner. MRI-measured g-ratio maps were computed and sampled along the visual transcallosal tract. EEG data were recorded using a 128-lead system with a visual Poffenberg paradigm. The interhemispheric transfer time and axonal conduction velocity were computed from the EEG current density at the group level. Using the MRI and EEG measures and the proposed model, we estimated morphological properties of axons in the visual transcallosal tract. Results: The estimated interhemispheric transfer time was 11.72 ± 2.87 ms, leading to an average conduction velocity across subjects of 13.22 ± 1.18 m/s. Out of the 4 free parameters of the proposed model, we estimated θ - the width of the right tail of the axonal radius distribution - and ß - the scaling factor of the axonal g-ratio, a measure of fiber myelination. Across subjects, the parameter θ was 0.40 ± 0.07 µm and the parameter ß was 0.67 ± 0.02 µm-α. Conclusion: The estimates of axonal radius and myelination are consistent with histological findings, illustrating the feasibility of this approach. The proposed method allows the measurement of the distribution of axonal radius and myelination within a white matter tract, opening new avenues for the combined study of brain structure and function, and for in vivo histological studies of the human brain.

Restoring statistical validity in group analyses of motion-corrupted MRI data.

Motion during the acquisition of magnetic resonance imaging (MRI) data degrades image quality, hindering our capacity to characterise disease in patient populations. Quality control procedures allow the exclusion of the most affected images from analysis. However, the criterion for exclusion is difficult to determine objectively and exclusion can lead to a suboptimal compromise between image quality and sample size. We provide an alternative, data-driven solution that assigns weights to each image, computed from an index of image quality using restricted maximum likelihood. We illustrate this method through the analysis of quantitative MRI data. The proposed method restores the validity of statistical tests, and performs near optimally in all brain regions, despite local effects of head motion. This method is amenable to the analysis of a broad type of MRI data and can accommodate any measure of image quality.

Does whole-body Patlak 18F-FDG PET imaging improve lesion detectability in clinical oncology?

PURPOSE: Single-pass whole-body (WB) 18F-FDG PET/CT imaging is routinely employed for the clinical assessment of malignant, infectious, and inflammatory diseases. Our aim in this study is the systematic clinical assessment of lesion detectability in multi-pass WB parametric imaging enabling direct imaging of the highly quantitative 18F-FDG influx rate constant Ki, as a complement to standard-of-care standardized uptake value (SUV) imaging for a range of oncologic studies. METHODS: We compared SUV and Ki images of 18 clinical studies of different oncologic indications (lesion characterization and staging) including standard-of-care SUV and dynamic WB PET protocols in a single session. The comparison involved both the visual assessment and the quantitative evaluation of SUVmean, SUVmax, Kimean, Kimax, tumor-to-background ratio (TBRSUV, TBRKi), and contrast-to-noise ratio (CNRSUV, CNRKi) quality metrics. RESULTS: Overall, both methods provided good-quality images suitable for visual interpretation. A total of 118 lesions were detected, including 40 malignant (proven) and 78 malignant (unproven) lesions. Of those, 111 were detected on SUV and 108 on Ki images. One proven malignant lesion was detected only on Ki images whereas none of the proven malignant lesions was visible only on SUV images. The proven malignant lesions had overall higher Ki TBR and CNR scores. One unproven lesion, which was later confirmed as benign, was detected only on the SUV images (false-positive). Overall, our results from 40 proven malignant lesions suggested improved sensitivity (from 92.5 to 95%) and accuracy (from 90.24 to 95.12%) and potentially enhanced specificity with Ki over SUV imaging. CONCLUSION: Oncologic WB Patlak Ki imaging may achieve equivalent or superior lesion detectability with reduced false-positive rates when complementing standard-of-care SUV imaging. KEY POINTS: • The whole-body spatio-temporal distribution of 18 F-FDG uptake may reveal clinically useful information on oncologic diseases to complement the standard-of-care SUV metric. • Parametric imaging resulted in less false-positive indications of non-specific 18 F-FDG uptake relative to SUV. • Parametric imaging may achieve equivalent or superior 18 F-FDG lesion detectability than standard-of-care SUV imaging in oncology.

Preparation of translationally cold neutral molecules.

Efforts at EPFL to obtain translationally cold neutral molecules are described. Active deceleration of polar molecules is performed by confining the molecules in moving three-dimensional electrostatic traps, and by appropriately choosing the velocity of those traps. Alternatively, cold molecules can be obtained by velocity filtering. Here, the velocity of the molecules is not changed, but instead the cold molecules are extracted from a thermal sample by using the competition between the electrostatic force and the centrifugal force inside a bent electrostatic guide for polar molecules.

Enhanced magnetic field sensitivity of spin-dependent transport in cluster-assembled metallic nanostructures.

The emerging field of spintronics explores the many possibilities offered by the prospect of using the spin of the electrons for fast, nanosized electronic devices. The effect of magnetization acting on a current is the essence of giant or tunnel magnetoresistance. Although such spintronics effects already find technological applications, much of the underlying physics remains to be explored. The aim of this article is to demonstrate the importance of spin mixing in metallic nanostructures. Here we show that magnetic clusters embedded in a metallic matrix exhibit a giant magnetic response of more than 500% at low temperature, using a recently developed thermoelectric measurement. This method eliminates the dominating resistivity component of the magnetic response and thus reveals an intrinsic spin-dependent process: the conduction-electron spin precession about the exchange field as the electron crosses the clusters, giving rise to a spin-mixing mechanism with strong field dependence. This effect appears sensibly only in the smallest clusters, that is, at the level of less than 100 atoms per cluster.