Dataset Information

Dimensionality reduction of diffusion MRI measures for improved tractometry of the human brain.

ABSTRACT: Various diffusion MRI (dMRI) measures have been proposed for characterising tissue microstructure over the last 15 years. Despite the growing number of experiments using different dMRI measures in assessments of white matter, there has been limited work on: 1) examining their covariance along specific pathways; and on 2) combining these different measures to study tissue microstructure. Indeed, it quickly becomes intractable for existing analysis pipelines to process multiple measurements at each voxel and at each vertex forming a streamline, highlighting the need for new ways to visualise or analyse such high-dimensional data. In a sample of 36 typically developing children aged 8-18 years, we profiled various commonly used dMRI measures across 22 brain pathways. Using a data-reduction approach, we identified two biologically-interpretable components that capture 80% of the variance in these dMRI measures. The first derived component captures properties related to hindrance and restriction in tissue microstructure, while the second component reflects characteristics related to tissue complexity and orientational dispersion. We then demonstrate that the components generated by this approach preserve the biological relevance of the original measurements by showing age-related effects across developmentally sensitive pathways. In summary, our findings demonstrate that dMRI analyses can benefit from dimensionality reduction techniques, to help disentangling the neurobiological underpinnings of white matter organisation.

SUBMITTER: Chamberland M

PROVIDER: S-EPMC6711466 | biostudies-literature | 2019 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Dimensionality reduction of diffusion MRI measures for improved tractometry of the human brain.

Chamberland Maxime M Raven Erika P EP Genc Sila S Duffy Kate K Descoteaux Maxime M Parker Greg D GD Tax Chantal M W CMW Jones Derek K DK

NeuroImage 20190620

Various diffusion MRI (dMRI) measures have been proposed for characterising tissue microstructure over the last 15 years. Despite the growing number of experiments using different dMRI measures in assessments of white matter, there has been limited work on: 1) examining their covariance along specific pathways; and on 2) combining these different measures to study tissue microstructure. Indeed, it quickly becomes intractable for existing analysis pipelines to process multiple measurements at eac ...[more]

PMID: 31228638

Similar Datasets

Project description:Visualization of anatomical structures using radiological imaging methods is an important tool in medicine to differentiate normal from pathological tissue and can generate large amounts of data for a radiologist to read. Integrating these large data sets is difficult and time-consuming. A new approach uses both supervised and unsupervised advanced machine learning techniques to visualize and segment radiological data. This study describes the application of a novel hybrid scheme, based on combining wavelet transform and nonlinear dimensionality reduction (NLDR) methods, to breast magnetic resonance imaging (MRI) data using three well-established NLDR techniques, namely, ISOMAP, local linear embedding (LLE), and diffusion maps (DfM), to perform a comparative performance analysis.Twenty-five breast lesion subjects were scanned using a 3T scanner. MRI sequences used were T1-weighted, T2-weighted, diffusion-weighted imaging (DWI), and dynamic contrast-enhanced (DCE) imaging. The hybrid scheme consisted of two steps: preprocessing and postprocessing of the data. The preprocessing step was applied for B(1) inhomogeneity correction, image registration, and wavelet-based image compression to match and denoise the data. In the postprocessing step, MRI parameters were considered data dimensions and the NLDR-based hybrid approach was applied to integrate the MRI parameters into a single image, termed the embedded image. This was achieved by mapping all pixel intensities from the higher dimension to a lower dimensional (embedded) space. For validation, the authors compared the hybrid NLDR with linear methods of principal component analysis (PCA) and multidimensional scaling (MDS) using synthetic data. For the clinical application, the authors used breast MRI data, comparison was performed using the postcontrast DCE MRI image and evaluating the congruence of the segmented lesions.The NLDR-based hybrid approach was able to define and segment both synthetic and clinical data. In the synthetic data, the authors demonstrated the performance of the NLDR method compared with conventional linear DR methods. The NLDR approach enabled successful segmentation of the structures, whereas, in most cases, PCA and MDS failed. The NLDR approach was able to segment different breast tissue types with a high accuracy and the embedded image of the breast MRI data demonstrated fuzzy boundaries between the different types of breast tissue, i.e., fatty, glandular, and tissue with lesions (>86%).The proposed hybrid NLDR methods were able to segment clinical breast data with a high accuracy and construct an embedded image that visualized the contribution of different radiological parameters.

Project description:Background MRI performed with echo-planar imaging (EPI) sequences is sensitive to susceptibility artifacts in the presence of metallic objects, which presents a substantial barrier for performing functional MRI and diffusion tensor imaging (DTI) in patients with metallic orthodontic material and other head implants. Purpose To evaluate the ability to reduce susceptibility artifacts in healthy human participants wearing metallic orthodontic braces for two alternative approaches: T2-prepared functional MRI and diffusion-prepared DTI with three-dimensional fast gradient-echo readout. Materials and Methods In this prospective study conducted from February to September 2018, T2-prepared functional MRI and diffusion-prepared DTI were performed in healthy human participants. Removable dental braces with bonding trays were used so that MRI could be performed with braces and without braces in the same participants. Results were evaluated in regions with strong (EPI dropout regions for functional MRI and the inferior fronto-occipital fasciculus for DTI) and minimal (motor cortex for functional MRI and the posterior limb of internal capsule for DTI) susceptibility artifacts. Signal-to-noise ratio (SNR), contrast-to-noise ratio for functional MRI, apparent diffusion coefficient and fractional anisotropy for DTI, and degree of distortion (quantified with the Jaccard index, which measures the similarity of geometric shapes) were compared in regions with strong or minimal susceptibility effects between the current standard EPI sequences and the proposed alternatives by using paired t test. Results Six participants were evaluated (mean age ± standard deviation, 40 years ± 6; three women). In brain regions with strong susceptibility effects from the metallic braces, T2-prepared functional MRI showed significantly higher SNR (37.8 ± 2.4 vs 15.5 ± 5.3; P < .001) and contrast-to-noise ratio (0.83 ± 0.16 vs 0.29 ± 0.10; P < .001), whereas diffusion-prepared DTI showed higher SNR (5.8 ± 1.5 vs 3.8 ± 0.7; P = .03) than did conventional EPI methods. Apparent diffusion coefficient and fractional anisotropy were consistent with the literature. Geometric distortion was substantially reduced throughout the brain with the proposed methods (significantly higher Jaccard index, 0.95 ± 0.12 vs 0.81 ± 0.61; P < .001). Conclusion T2-prepared functional MRI and diffusion-prepared diffusion tensor imaging can acquire functional and diffusion MRI, respectively, in healthy human participants wearing metallic dental braces with less susceptibility artifacts and geometric distortion than with conventional echo-planar imaging. © RSNA, 2019 Online supplemental material is available for this article. See also the editorial by Dietrich in this issue.

Dataset Information

Dimensionality reduction of diffusion MRI measures for improved tractometry of the human brain.

Publications

Dimensionality reduction of diffusion MRI measures for improved tractometry of the human brain.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets