Sensitivity of Diffusion MRI to White Matter Pathology: Influence of Diffusion Protocol, Magnetic Field Strength, and Processing Pipeline in Systemic Lupus Erythematosus

Impact of truncating diffusion MRI scans on diffusional kurtosis imaging

OBJECTIVE

Diffusional kurtosis imaging (DKI) extends diffusion tensor imaging (DTI), characterizing non-Gaussian diffusion effects but requires longer acquisition times. To ensure the robustness of DKI parameters, data acquisition ordering should be optimized allowing for scan interruptions or shortening. Three methodologies were used to examine how reduced diffusion MRI scans impact DKI histogram-metrics: 1) the electrostatic repulsion model (OptEEM); 2) spherical codes (OptSC); 3) random (RandomTRUNC).

MATERIALS AND METHODS

Pre-acquired diffusion multi-shell data from 14 female healthy volunteers (29±5 years) were used to generate reordered data. For each strategy, subsets containing different amounts of the full dataset were generated. The subsampling effects were assessed on histogram-based DKI metrics from tract-based spatial statistics (TBSS) skeletonized maps. To evaluate each subsampling method on simulated data at different SNRs and the influence of subsampling on in vivo data, we used a 3-way and 2-way repeated measures ANOVA, respectively.

RESULTS

Simulations showed that subsampling had different effects depending on DKI parameter, with fractional anisotropy the most stable (up to 5% error) and radial kurtosis the least stable (up to 26% error). RandomTRUNC performed the worst while the others showed comparable results. Furthermore, the impact of subsampling varied across distinct histogram characteristics, the peak value the least affected (OptEEM: up to 5% error; OptSC: up to 7% error) and peak height (OptEEM: up to 8% error; OptSC: up to 11% error) the most affected.

CONCLUSION

The impact of truncation depends on specific histogram-based DKI metrics. The use of a strategy for optimizing the acquisition order is advisable to improve DKI robustness to exam interruptions.

Comparison of distortion correction preprocessing pipelines for DTI in the upper limb

PURPOSE

DTI characterizes tissue microstructure and provides proxy measures of nerve health. Echo-planar imaging is a popular method of acquiring DTI but is susceptible to various artifacts (e.g., susceptibility, motion, and eddy currents), which may be ameliorated via preprocessing. There are many pipelines available but limited data comparing their performance, which provides the rationale for this study.

METHODS

DTI was acquired from the upper limb of heathy volunteers at 3T in blip-up and blip-down directions. Data were independently corrected using (i) FSL's TOPUP & eddy, (ii) FSL's TOPUP, (iii) DSI Studio, and (iv) TORTOISE. DTI metrics were extracted from the median, radial, and ulnar nerves and compared (between pipelines) using mixed-effects linear regression. The geometric similarity of corrected b = 0 images and the slice matched T1-weighted (T1w) images were computed using the Sörenson-Dice coefficient.

RESULTS

Without preprocessing, the similarity coefficient of the blip-up and blip-down datasets to the T1w was 0·80 and 0·79, respectively. Preprocessing improved the geometric similarity by 1% with no difference between pipelines. Compared to TOPUP & eddy, DSI Studio and TORTOISE generated 2% and 6% lower estimates of fractional anisotropy, and 6% and 13% higher estimates of radial diffusivity, respectively. Estimates of anisotropy from TOPUP & eddy versus TOPUP were not different but TOPUP reduced radial diffusivity by 3%. The agreement of DTI metrics between pipelines was poor.

CONCLUSIONS

Preprocessing DTI from the upper limb improves geometric similarity but the choice of the pipeline introduces clinically important variability in diffusion parameter estimates from peripheral nerves.

MRI BrainAGE demonstrates increased brain aging in systemic lupus erythematosus patients

Introduction

Systemic lupus erythematosus (SLE) is an autoimmune connective tissue disease affecting multiple organs in the human body, including the central nervous system. Recently, an artificial intelligence method called BrainAGE (Brain Age Gap Estimation), defined as predicted age minus chronological age, has been developed to measure the deviation of brain aging from a healthy population using MRI. Our aim was to evaluate brain aging in SLE patients using a deep-learning BrainAGE model.

Methods

Seventy female patients with a clinical diagnosis of SLE and 24 healthy age-matched control females, were included in this post-hoc analysis of prospectively acquired data. All subjects had previously undergone a 3 T MRI acquisition, a neuropsychological evaluation and a measurement of neurofilament light protein in plasma (NfL). A BrainAGE model with a 3D convolutional neural network architecture, pre-trained on the 3D-T1 images of 1,295 healthy female subjects to predict their chronological age, was applied on the images of SLE patients and controls in order to compute the BrainAGE. SLE patients were divided into 2 groups according to the BrainAGE distribution (high vs. low BrainAGE).

Results

BrainAGE z-score was significantly higher in SLE patients than in controls (+0.6 [±1.1] vs. 0 [±1.0], p = 0.02). In SLE patients, high BrainAGE was associated with longer reaction times (p = 0.02), lower psychomotor speed (p = 0.001) and cognitive flexibility (p = 0.04), as well as with higher NfL after adjusting for age (p = 0.001).

Conclusion

Using a deep-learning BrainAGE model, we provide evidence of increased brain aging in SLE patients, which reflected neuronal damage and cognitive impairment.

Influence of adaptive denoising on Diffusion Kurtosis Imaging at 3T and 7T

BACKGROUND AND OBJECTIVE

Choosing the most appropriate denoising method to improve the quality of diagnostic images maximally is key in pre-processing of diffusion MRI images. Recent advancements in acquisition and reconstruction techniques have questioned traditional noise estimation methods favoring adaptive denoising frameworks, circumventing the need to know a priori information that is hardly available in a clinical setting. In this observational study, we compared two innovative adaptive techniques sharing some features, Patch2Self and Nlsam, through application on reference adult data at 3T and 7T. The primary aim was identifying the most effective method in case of Diffusion Kurtosis Imaging (DKI) data - particularly susceptible to noise and signal fluctuations - at 3T and 7T fields. A side goal consisted of investigating the dependence of kurtosis metrics' variability with respect to the magnetic field on the adopted denoising methodology.

METHODS

For comparison purposes, we focused on qualitative and quantitative analysis of DKI data and related microstructural maps before and after applying the two denoising approaches. Specifically, we assessed computational efficiency, preservation of anatomical details via perceptual metrics, consistency of microstructure model fitting, alleviation of degeneracies in model estimation, and joint variability with varying field strength and denoising method.

RESULTS

Accounting for all these factors, Patch2Self framework has turned out to be specifically suitable for DKI data, with improving performance at 7T. Nlsam method is more robust in alleviating degenerate black voxels while introducing some blurring, which in turn is reflected in an overall loss of image sharpness. Regarding the impact of denoising on field-dependent variability, both methods have been shown to make variations from standard to Ultra-High Field more concordant with theoretical evidence, claiming that kurtosis metrics are sensitive to susceptibility-induced background gradients, directly proportional to the magnetic field strength and sensitive to the microscopic distribution of iron and myelin.

CONCLUSIONS

This study serves as a proof-of-concept stressing the need for an accurate choice of a denoising methodology, specifically tailored for the data under analysis and allowing higher spatial resolution acquisition within clinically compatible timings, with all the potential benefits that improving suboptimal quality of diagnostic images entails.