Abstract
Accurate and automated reconstruction of the in vivo human cerebral cortical surface from anatomical magnetic resonance (MR) images facilitates the quantitative analysis of cortical ...structure. Anatomical MR images with sub-millimeter isotropic spatial resolution improve the accuracy of cortical surface and thickness estimation compared to the standard 1-millimeter isotropic resolution. Nonetheless, sub-millimeter resolution acquisitions require averaging multiple repetitions to achieve sufficient signal-to-noise ratio and are therefore long and potentially vulnerable to subject motion. We address this challenge by synthesizing sub-millimeter resolution images from standard 1-millimeter isotropic resolution images using a data-driven supervised machine learning-based super-resolution approach achieved via a deep convolutional neural network. We systematically characterize our approach using a large-scale simulated dataset and demonstrate its efficacy in empirical data. The super-resolution data provide improved cortical surfaces similar to those obtained from native sub-millimeter resolution data. The whole-brain mean absolute discrepancy in cortical surface positioning and thickness estimation is below 100 μm at the single-subject level and below 50 μm at the group level for the simulated data, and below 200 μm at the single-subject level and below 100 μm at the group level for the empirical data, making the accuracy of cortical surfaces derived from super-resolution sufficient for most applications.
Purpose
The goal of this study is to leverage an advanced fast imaging technique, wave‐controlled aliasing in parallel imaging (Wave‐CAIPI), and a generative adversarial network (GAN) for denoising ...to achieve accelerated high‐quality high‐signal‐to‐noise‐ratio (SNR) volumetric magnetic resonance imaging (MRI).
Methods
Three‐dimensional (3D) T2‐weighted fluid‐attenuated inversion recovery (FLAIR) image data were acquired on 33 multiple sclerosis (MS) patients using a prototype Wave‐CAIPI sequence (acceleration factor R = 3 × 2, 2.75 min) and a standard T2‐sampling perfection with application‐optimized contrasts by using flip angle evolution (SPACE) FLAIR sequence (R = 2, 7.25 min). A hybrid denoising GAN entitled “HDnGAN” consisting of a 3D generator and a 2D discriminator was proposed to denoise highly accelerated Wave‐CAIPI images. HDnGAN benefits from the improved image synthesis performance provided by the 3D generator and increased training samples from a limited number of patients for training the 2D discriminator. HDnGAN was trained and validated on data from 25 MS patients with the standard FLAIR images as the target and evaluated on data from eight MS patients not seen during training. HDnGAN was compared to other denoising methods including adaptive optimized nonlocal means (AONLM), block matching with 4D filtering (BM4D), modified U‐Net (MU‐Net), and 3D GAN in qualitative and quantitative analysis of output images using the mean squared error (MSE) and Visual Geometry Group (VGG) perceptual loss compared to standard FLAIR images, and a reader assessment by two neuroradiologists regarding sharpness, SNR, lesion conspicuity, and overall quality. Finally, the performance of these denoising methods was compared at higher noise levels using simulated data with added Rician noise.
Results
HDnGAN effectively denoised low‐SNR Wave‐CAIPI images with sharpness and rich textural details, which could be adjusted by controlling the contribution of the adversarial loss to the total loss when training the generator. Quantitatively, HDnGAN (λ = 10–3) achieved low MSE and the lowest VGG perceptual loss. The reader study showed that HDnGAN (λ = 10–3) significantly improved the SNR of Wave‐CAIPI images (p < 0.001), outperformed AONLM (p = 0.015), BM4D (p < 0.001), MU‐Net (p < 0.001), and 3D GAN (λ = 10–3) (p < 0.001) regarding image sharpness, and outperformed MU‐Net (p < 0.001) and 3D GAN (λ = 10–3) (p = 0.001) regarding lesion conspicuity. The overall quality score of HDnGAN (λ = 10–3) (4.25 ± 0.43) was significantly higher than those from Wave‐CAIPI (3.69 ± 0.46, p = 0.003), BM4D (3.50 ± 0.71, p = 0.001), MU‐Net (3.25 ± 0.75, p < 0.001), and 3D GAN (λ = 10–3) (3.50 ± 0.50, p < 0.001), with no significant difference compared to standard FLAIR images (4.38 ± 0.48, p = 0.333). The advantages of HDnGAN over other methods were more obvious at higher noise levels.
Conclusion
HDnGAN provides robust and feasible denoising while preserving rich textural detail in empirical volumetric MRI data. Our study using empirical patient data and systematic evaluation supports the use of HDnGAN in combination with modern fast imaging techniques such as Wave‐CAIPI to achieve high‐fidelity fast volumetric MRI and represents an important step to the clinical translation of GANs.
Purpose
To develop an efficient MR technique for ultra‐high resolution diffusion MRI (dMRI) in the presence of motion.
Methods
gSlider is an SNR‐efficient high‐resolution dMRI acquisition technique. ...However, subject motion is inevitable during a prolonged scan for high spatial resolution, leading to potential image artifacts and blurring. In this study, an integrated technique termed Motion Corrected gSlider (MC‐gSlider) is proposed to obtain high‐quality, high‐resolution dMRI in the presence of large in‐plane and through‐plane motion. A motion‐aware reconstruction with spatially adaptive regularization is developed to optimize the conditioning of the image reconstruction under difficult through‐plane motion cases. In addition, an approach for intra‐volume motion estimation and correction is proposed to achieve motion correction at high temporal resolution.
Results
Theoretical SNR and resolution analysis validated the efficiency of MC‐gSlider with regularization, and aided in selection of reconstruction parameters. Simulations and in vivo experiments further demonstrated the ability of MC‐gSlider to mitigate motion artifacts and recover detailed brain structures for dMRI at 860 μm isotropic resolution in the presence of motion with various ranges.
Conclusion
MC‐gSlider provides motion‐robust, high‐resolution dMRI with a temporal motion correction sensitivity of 2 s, allowing for the recovery of fine detailed brain structures in the presence of large subject movements.
Diffusion‐weighted magnetic resonance imaging (DW‐MRI) has evolved to provide increasingly sophisticated investigations of the human brain's structural connectome in vivo. Restriction spectrum ...imaging (RSI) is a method that reconstructs the orientation distribution of diffusion within tissues over a range of length scales. In its original formulation, RSI represented the signal as consisting of a spectrum of Gaussian diffusion response functions. Recent technological advances have enabled the use of ultra‐high b‐values on human MRI scanners, providing higher sensitivity to intracellular water diffusion in the living human brain. To capture the complex diffusion time dependence of the signal within restricted water compartments, we expand upon the RSI approach to represent restricted water compartments with non‐Gaussian response functions, in an extended analysis framework called linear multi‐scale modeling (LMM). The LMM approach is designed to resolve length scale and orientation‐specific information with greater specificity to tissue microstructure in the restricted and hindered compartments, while retaining the advantages of the RSI approach in its implementation as a linear inverse problem. Using multi‐shell, multi‐diffusion time DW‐MRI data acquired with a state‐of‐the‐art 3 T MRI scanner equipped with 300 mT/m gradients, we demonstrate the ability of the LMM approach to distinguish different anatomical structures in the human brain and the potential to advance mapping of the human connectome through joint estimation of the fiber orientation distributions and compartment size characteristics.
Linear Multi‐scale Modeling (LMM) for diffusion weighted MRI enables a detailed microstructural tissue characterization by separating orientation distributions of restricted and hindered diffusion water compartments over a range of length scales. We demonstrate the ability of LMM to estimate volume fractions, compartment sizes and orientation distributions utilizing both simulations as well as empirical data from healthy subjects using a human 3T MRI scanner equipped with a 300 mT/m gradient system.
Conventional diffusion-weighted MR imaging techniques provide limited specificity in disentangling disease-related microstructural alterations involving changes in both axonal density and ...myelination. By simultaneously probing multiple diffusion regimens, multi-shell diffusion MRI is capable of increasing specificity to different tissue sub-compartments and hence separate different contributions to changes in diffusion-weighted signal attenuation. Advanced multi-shell diffusion models impose significant requirements on the amount of diffusion weighting (i.e. gradient coil performance) and angular resolution (i.e. in-scanner subject time), which commonly limits their applicability in a clinical setting. In this paper, we apply a high-b-value, high angular resolution multi-shell diffusion MRI protocol to a population of early multiple sclerosis (MS) patients and healthy controls. Through the Composite Hindered and Restricted Model of Diffusion (CHARMED) model, we extract indices for axonal density as well as parameters sensitive to myelin. We demonstrate increased sensitivity to microstructural changes in normal appearing white matter and in lesions in MS as compared to traditional models like DTI. These changes appear to be predominantly in axonal density, pointing towards the existence of axonal damage mechanisms in early MS.
With the advent of neuroimaging techniques, especially functional MRI (fMRI), studies have mapped brain regions that are associated with good and poor reading, most centrally a region within the left ...occipito-temporal/fusiform region (L-OT/F) often referred to as the visual word form area (VWFA). Despite an abundance of fMRI studies of the VWFA, research about its structural connectivity has just started. Provided that the VWFA may be connected to distributed regions in the brain, it remains unclear how this network is engaged in constituting a well-tuned reading circuitry in the brain. Here we used diffusion MRI to study the structural connectivity patterns of the putative VWFA and surrounding areas within the L-OT/F in children with typically developing (TD) reading ability and with word recognition deficits (WRD; sometimes referred to as dyslexia). We found that L-OT/F connectivity varied along a posterior- anterior gradient, with specific structural connectivity patterns related to reading ability in the ROIs centered upon the putative VWFA. Findings suggest that the architecture of the VWFA connectivity is fundamentally different between TD and WRD, with TD showing greater connectivity to linguistic regions than WRD, and WRD showing greater connectivity to visual and parahippocampal regions than TD. Findings thus reveal clear structural abnormalities underlying the functional abnormalities in the VWFA in WRD.
Deep Brain Stimulation (DBS) is a neurosurgical procedure that can reduce symptoms in medically intractable obsessive-compulsive disorder (OCD). Conceptually, DBS of the ventral capsule/ventral ...striatum (VC/VS) region targets reciprocal excitatory connections between the orbitofrontal cortex (OFC) and thalamus, decreasing abnormal reverberant activity within the OFC-caudate-pallidal-thalamic circuit. In this study, we investigated these connections using diffusion magnetic resonance imaging (dMRI) on human connectome datasets of twenty-nine healthy young-adult volunteers with two-tensor unscented Kalman filter based tractography. We studied the morphology of the lateral and medial orbitofrontothalamic connections and estimated their topographic variability within the VC/VS region. Our results showed that the morphology of the individual orbitofrontothalamic fibers of passage in the VC/VS region is complex and inter-individual variability in their topography is high. We applied this method to an example OCD patient case who underwent DBS surgery, formulating an initial proof of concept for a tractography-guided patient-specific approach in DBS for medically intractable OCD. This may improve on current surgical practice, which involves implanting all patients at identical stereotactic coordinates within the VC/VS region.