Keywords: Multiple Sclerosis, High-Field MRI, Transfer learning; Brain Segmentation; Lesion detection

Motivation: Brain segmentation is more challenging at 7T compared to 3T, primarily due to increased bias fields and other artifacts. Generating training data for 7T brain segmentation is tedious, making transfer learning based models a more feasible option.

Goal(s): Brain and lesion segmentation algorithm for use with 7T images in multiple sclerosis. 

Approach: A 3T to-7T transfer learning algorithm (called PLAn) for skull stripping, lesion, and brain segmentation was trained and tested on participants clinically diagnosed with multiple sclerosis. 

Results: In both quantitative and qualitative analysis, PLAn significantly outperformed other segmentation methods including nnU-Net in lesion and brain segmentation. 

Impact: Brain volume is a commonly used marker of disease progression in various neurological and neuropsychiatric diseases; however it is more difficult to implement on 7T images. PLAn, a deep-learning algorithm, can produce fast and reliable whole-brain segmentations.

Keywords: Multiple Sclerosis, Neurodegeneration

Motivation: Microstructure-informed tractography methods face challenges in presence of focal lesions.

Goal(s): This study aims to apply an extension of Myelin Streamline Decomposition (MySD) which accounts for lesions to patients with multiple sclerosis (MS) enrolled in a completed phase 2 clinical trial (SYNERGY) with conventional DTI acquisitions. 

Approach: We applied our novel approach to perform network analysis in SYNERGY’s secondary progressive (SP) and relapsing-remitting (RR) patients.

Results: Using MySD applied to patients with focal pathology, we showed that SPMS patients had increased alterations in myelin-weighted network properties compared to RRMS. Myelin-weighted networks also exhibited correlations with motor and cognitive impairment.

Impact: For the first time, we applied our novel connectomics approach explicitly designed to cope with focal pathology to MS patients in a clinical trial, demonstrating its sensitivity and adaptability to real-world clinical data.

Keywords: Multiple Sclerosis, Diffusion Tensor Imaging, multiple sclerosis; mrf; g-ratio

Motivation: Multiple sclerosis effect both white matter and gray matter. 

Goal(s): Using mIR-MRF with NODDI to generate myelin water fraction map, g-ratio map, and axonal volume fraction gray matter map.

Approach: White matter, gray matter, cerebrospinal fluid (CSF), and myelin water fraction (MWF) map were generated by mIR MRF. NODDI was used to divide each voxel into isotropic (V_iso) and anisotropically restricted volume fraction (V_ic) map. These maps were used to generate the g-ratio and AVF-gm map.

Results: The WM lesions were in consistency within FLAIR and MWF but have a variance in g-ratio maps. AVF-gm map depicted the axonal fiber in GM well.

Impact: The mIR MRF with NODDI could provide insight of demyelination, axonal loss and the change in GM of MS. 

Keywords: Multiple Sclerosis, Neurodegeneration

Motivation: Myelin water fraction (MWF) is a promising method for quantitative evaluation of myelin in the brain. FLORET MR bone imaging using UTE sequences has possibility to show brain myelin structure because myelin also has quite short T2 relaxation times. 

Goal(s): To demonstrate the feasibility of FLORET MyelinVIEW in comparison with MWF.

Approach: We evaluated the FLORET MyelinVIEW and MWF based on the distribution of myelin in the brain structures.

Results: FLORET MyelinVIEW may be useful for evaluating the distribution of myelin.

Impact: FLORET MyelinVIEW showed comparable distribution of brain myelin to MWF, it may be useful to visualize the myelin distribution with the 3D isotropic images.

Keywords: Multiple Sclerosis, Neuroinflammation

Motivation: MT-weighted imaging is important for detecting and diagnosing neuro-degenerative disorders. At 7T, high RF power deposition is challenging for MT-weighted sequences and often limits imaging performance.

Goal(s): We demonstrate a high-resolution, MT-weighted, Zero Time Echo (ZTE) sequence which can produce whole-brain images at clinically useful levels of RF power deposition.

Approach: We demonstrate an MT-weighted ZTE sequence for 7T and use it to scan healthy volunteers and subjects with Multiple Sclerosis

Results: Whole-brain MT-weighted ZTE images of high quality and 0.8mm isotropic imaging were produced with clinically useful SAR levels.

Impact: High-resolution MT-weighted ZTE images could provide important novel information for neurologists and psychiatrists.

Keywords: Multiple Sclerosis, Perfusion, systemic sclerosis

Motivation: Cerebral involvement in systemic sclerosis (SSc) was reported increasingly but lacks evidence of specific imaging biomarkers.

Goal(s): To characterize cerebral structure and perfusion in SSc patients using MR imaging.

Approach: Brain structure and perfusion were characterized using T1w and pCASL MR sequences. The regional volume and cerebral blood flow (CBF) and voxel-wise CBF differences were compared between healthy volunteers, diffuse cutaneous (dcSSc) and limited cutaneous of SSc (lcSSc) patients.

Results: The SSc patients showed significantly lower voxel-wize CBF than healthy volunteers, and the dcSSc patients showed significantly lower CBF in both region-wise and voxel-wise level than the lcSSc patients.

Impact: Our study demonstrates the regional-wise and voxel-wise cerebral blood flow decrease in SSc patients, which might provide the new insights into the central nervous system involvement and related clinical manifestations in SSc.

Keywords: Multiple Sclerosis, Arterial spin labelling, exercise, fatigue, celebral blood flow

Motivation: While fatigue is a common symptom in neurological conditions such as Multiple Sclerosis and Parkinson's, there is currently a lack of MRI studies that have investigated the CBF response to fatigue. 

Goal(s): To use ASL MRI to capture dynamic changes in CBF as muscles fatigued during maximal isometric effort.

Approach: ASL CBF images were acquired pre, during, and post exercise on 7 healthy volunteers.

Results: Primary motor cortex, supplementary motor cortex, putamen, and anterior cerebellum showed an increase in CBF followed by a gradual return to the baseline with a delay between the start of the decrease in CBF and the fatigue onset.

Impact: As the use of MRI in neurological diseases involving fatigue continues to grow, the concurrent measurement of CBF alongside other physiological parameters holds the potential to yield valuable insights that could contribute to advancement of effective treatments and preventive strategies.

Keywords: Multiple Sclerosis, Multiple Sclerosis

Motivation: Though periventricular gradient (PVG) mapping has been demonstrated to be useful, previous studies have employed 2D magnetization transfer ratios and T1 maps, which are not specific to myelin. 

Goal(s): This study investigates the use of synthetic MRI myelin maps to assess the periventricular gradient (PVG) of myelin volume fraction (MVF) in multiple sclerosis (MS) patients.

Approach: This study includes 22 healthy individuals, 84 patients with relapsing-remitting MS (RRMS), and 17 patients with secondary progressive MS (SPMS). 

Results: PVGs of RRMS and SPMS were higher than that of healthy individuals, with SPMS showing higher PVG than RRMS. 

Impact: In conclusion, the influence of demyelination in MS appears to be more prominent in the vicinity of the ventricles.

Keywords: Multiple Sclerosis, Multiple Sclerosis, Relaxometry, T1 mapping, T2 mapping, T1w/T2w ratio

Motivation: T1w/T2w is a marker of tissue integrity in multiple sclerosis (MS), but is derived from (normalised) signal intensities, not quantitative measures.

Goal(s): To determine whether longitudinal quantitative T1 and T2 and T1.T2_product (directionally equivalent to T1w/T2w) measures are sensitive to disability-relevant tissue damage in early MS.

Approach: Voxelwise T1.T2_product maps were estimated in people recently diagnosed with relapsing-remitting MS (n=62). Longitudinal change in normal-appearing white matter (NAWM), grey matter (GM) and white matter lesions (WML), and relationship with disability, were assessed.

Results: Greater increases in T1.T2_product in NAWM and GM are associated with disability worsening.

Impact: T1, T2 and derived T1.T2_product measures are sensitive to disability-relevant tissue damage in recently diagnosed relapsing-remitting MS, and may provide greater insights into underlying pathological change than T1w/T2w intensity ratios.

Keywords: Multiple Sclerosis, Machine Learning/Artificial Intelligence, Quantitative MRI; qMRI; Contrast enhanced imaging; active lesions

Motivation: The gold standard way for assessing Multiple sclerotic (MS) disease activity is by identifying new active lesions using contrast enhanced imaging. The repeated use of gadolinium injections for MS patients constitute a major concern due to long-term accumulation and even breakdown of this agent in the brain and body without efficient clearance.

Goal(s): Classify active vs. inactive MS lesions using quantitative MRI (qMRI) without the need for contrast-enhanced imaging.

Approach: Machine learning classifier trained on qMRI features of MS lesions.

Results: qMRI profiling has the potential to classify MS lesions into active/inactive state with accuracy of 81.7 ± 10 %.

Impact: Multiple sclerosis disease activity is assessed using contrast-enhanced MRI. Recently, concerns have been raised regarding the long-term accumulation and breakdown of contrast agents in the brain. This study introduces a qMRI-based and contrast-free approach for assessing multiple sclerosis disease activity.

Keywords: Multiple Sclerosis, Multiple Sclerosis, Chronic lesions, diffusion MRI

Motivation: Slowly expanding lesions (SEL) have gained significant attention as a biomarker of chronic active multiple sclerosis (MS) lesions, however, one study¹ suggests that all MS lesions tend to shrink over a long period of time.

Goal(s): The objective of this work is to investigate the microstructure of expanding lesions (EL), shrinking lesions (SL), and stable lesions.

Approach: EL and SL were computed using deformation-based volumetric MRI and microstructure was investigated using quantitative T1 and multi-shell diffusion MRI.

Results: SL showed a more destructive phenotype at baseline when compared to EL, while stable lesions were considerably less destructive.

Impact: This preliminary study underlies the necessity of considering the full spectrum of multiple sclerosis (MS) lesions, especially MRI-evolving lesions, whether shrinking or expanding, in MS research to extend our knowledge of the disease pathophysiology.

Keywords: Multiple Sclerosis, Multiple Sclerosis, Myelin Water Imaging

Motivation: ViSTa offers improved myelin-weighted image. However, the presence of flow artifacts was seen when scanning a subject with a high flow speed.

Goal(s): This study aimed to construct a novel 3D ViSTa sequence by implementing a series of flow saturation pulses.

Approach: Three flow saturation pulses were placed as evenly as possible between the 1^st inversion pulse and the excitation pulse to effectively mitigate the presence of flow with various speeds.

Results: The results from the proposed ViSTa sequence reveal a significant reduction in flow artifacts.
When we conduct MS patient scans, lesions have a significantly diminished signal level, providing a distinct demarcation.

Impact: The proposed method provides improved whole-brain covered 3D myelin-weighted images in a clinically reasonable scan time (< 7 min). Moreover, it shows good sensitivity to MS lesions. These features make the proposed method appealing for clinical neuroimaging applications.

Keywords: Multiple Sclerosis, Multiple Sclerosis

Motivation: Differentiation of Paramagnetic Rim Lesions, based on T1 values, could allow clinical stratification of patients.

Goal(s): To investigate PRL heterogeneity with quantitative T1 and normalised MPRAGE and the resulting impact on clinical outcomes.

Approach: White Matter lesions were segmented and categorised based on visual confidence of paramagnetic rim visibility and their T1 values. RF analyses was used to reveal impact of lesion types on clinical disability.

Results: Visibility of Rim+ lesions is correlated with increased T1 values. In a Random Forest Model, T1-derived long-T1 Rim+ lesions and number of short-T1 Rim+ lesions could predict clinical disability and clinical progression in next 12 months.

Impact: Characterisation of PRL heterogeneity, not only with advanced-quantitative imaging techniques but also with clinically available sequences, allows a better understanding of a patient's current clinical disability and progression over the next twelve months.

Keywords: Multiple Sclerosis, Brain

Motivation: Multiple sclerosis (MS) is characterized by demyelination of the central nervous system. Ultra High Field (UHF) MRI can reveal CNS pathology that is not apparent on conventional MRI in MS patients, but options for measuring myelin density are limited.

Goal(s): Calculation of MWF using ViSTa map requires intensity normalization. We evaluate normalization approaches for visualization of short transverse relaxation time component (ViSTa), a semiquantitative measure of apparent myelin water fraction (aMWF).

Approach: We compare two types of normalization scan - a turboflash (TFL) and a gradient recalled echo (GRE) on MS patients.

Results: We find that normalization with GRE provides better conspicuity of lesions.

Impact: Characterization of myelin density has the potential to supplement UHF approaches to evaluating MS.

Keywords: Multiple Sclerosis, Multiple Sclerosis

Motivation: Magnetic susceptibility is influenced by myelin concentration, playing a significant role in the pathogenesis of MS as a demyelinating disease.

Goal(s): This study aims to investigate normal-appearing-white-matter in MS patients using QSM, focusing on the cortico-spinal tract and optic radiation, to find non-invasive biomarkers of pre-clinical inflammatory activity.

Approach: The automated implemented pipeline relies on the acquisition of multiple MR sequences. Several susceptibility histogram properties were considered and correlated with disability scores.

Results: A decrease in myelin concentration was detected in MS group, consistently with the pathophysiology. Correlations between susceptibility and clinical disability occur, distinguishing clinical phenotypes and levels of motor impairment.

Impact: Studying the normal-appearing-white-matter tracts using QSM reveals decrease in myelin concentration within cortico-spinal tract and optic radiation in MS patients. Variations were observed between different clinical phenotypes and  various levels of motor impairment, suggesting biomarkers for  early diagnosis and prognosis.

Keywords: Multiple Sclerosis, Multiple Sclerosis

Motivation: To understand the different BOLD activation pattern in multiple sclerosis patients

Goal(s): To Investigate the neuro-vascular uncoupling of MS patients to be compared with healthy controls

Approach: During the visual task and hypercapnia challenge, BOLD and CBF are measured simultaneously with EEG in patients of MS (n=11) and health controls (n=15).

Results: We find that, for healthy controls, only models that included neuronal activity, i.e. EEG power change during task performance, were significantly predictive of BOLD response

Impact: We investigate BOLD activation of MS patients and compare with controls using EEG simultaneously during a visual task and hypercapnia challenge. We find that the BOLD response in MS does not directly depend on the degree of underlying neuronal activity.