Ivan I. Kirov

In mild traumatic brain injury, imaging biomarkers are needed to support clinical management. In four antecedent publications, we used two new (sodium and fingerprinting) and two established (spectroscopy and diffusion) MR techniques in a longitudinally followed patient cohort. Here we report final results and combine all data to determine which marker(s) from the four modalities offer the greatest utility for detecting injury and predicting outcomes. We also leverage the independent specificities offered by each modality to explore injury mechanisms. The longitudinal spectroscopy data were analysed to complete a full data set of proton (spectroscopy, fingerprinting, diffusion) and sodium MRI, acquired alongside symptomatic, cognitive, and functional assessments in 27 patients at 1, 3, and 12 months following injury. Twenty-three matched controls were scanned once. Testing for associations between nine MR markers and three outcome measures was standardized across the entire data set, and performed using Spearman correlations and logistic regression. Previously elevated white matter choline and creatine from spectroscopy (markers of membrane turnover and cellular energetics, respectively) normalized to control levels by 3 months, at rates which correlated with the rate of symptom recovery. Sodium and spectroscopy showed findings coinciding in pattern and timepoint, but there were no associations between them, suggesting independent origin. Choline and creatine met the greatest number of biomarker properties, followed by water T₁ from fingerprinting (marker of the cellular microenvironment). We identified independent, dynamic, metabolic and ionic changes, with choline and creatine from spectroscopy fulfilling the most criteria for a clinical biomarker.

Ionic imbalances and sodium channel dysfunction, well-known sequelae of traumatic brain injury (TBI), promote functional impairment in affected subjects. Therefore, non-invasive measurement of sodium concentrations using ²³Na MRI has the potential to detect clinically relevant injury and predict persistent symptoms. Recently, we reported diffusely lower apparent total sodium concentrations (aTSC) in mild TBI patients compared to controls, as well as correlations between lower aTSC and worse clinical outcomes. The main goal of this study was to determine whether these aTSC findings, and their changes over time, predict outcomes at 3- and 12-month from injury. Twenty-seven patients previously studied with ²³Na MRI and outcome measures at 22 ± 10 days (average ± standard deviation) after injury (visit-1, v1) were contacted at 3- (visit-2, v2) and 12-month after injury (visit-3, v3) to complete the Rivermead post-concussion symptoms questionnaire (RPQ), the extended Glasgow outcome scale (GOSE), and the brief test of adult cognition by telephone (BTACT). Follow-up ¹H and ²³Na MRI were additionally scheduled at v2. Linear regression was used to calculate aTSC in global grey and white matters. Six hypotheses were tested in relation to the serial changes in outcome measures and in aTSC, and in relation to the cross-sectional and serial relationships between aTSC and outcome. Twenty patients contributed data at v2 and fifteen at v3. Total RPQ and composite BTACT z-scores differed significantly for v2 and v3 in comparison to v1 (each P < 0.01), reflecting longitudinally reduced symptomatology and improved performance on cognitive testing. No associations between aTSC and outcome were observed at v2. Previously lower grey and white matter aTSC normalized at v2 in comparison to controls, in line with a statistically detectable longitudinal increase in grey matter aTSC between v1 and v2 (P = 0.0004). aTSC values at v1 predicted a subset of future BTACT subtest scores, but not future RPQ scores nor GOSE-defined recovery status. Similarly, aTSC rates of change correlated with BTACT rates of change, but not with those of RPQ. Tissue aTSC, previously shown to be diffusely decreased compared to controls at v1, was no longer reduced by v2, suggesting normalization of the sodium ionic equilibrium. These changes were accompanied by marked improvement in outcome. The results support the notion that early aTSC from ²³Na MRI predicts future BTACT, but not RPQ scores, nor future GOSE status.

We describe a new comprehensive and open-source MATLAB library, called the Visual Display Interface (VDI), designed to process, model, quantify and visualize magnetic resonance spectroscopy (MRS) and spectroscopic imaging (MRSI) data. The library focuses on three major strengths: promoting reproducible research by creating comprehensive data logging and reporting and identifying outlier datasets; seamlessly combining spectral and spatial processing of both single and multivoxel data; and offering a modern, object-oriented design. VDI handles a wide range of common tasks, including spectral and spatial transforms, built-in spectral fitting, absolute quantification, and running density-matrix simulations for coupled and uncoupled spin systems to generate appropriate basis sets and aid in sequence design. VDI interfaces with the Statistical Parameteric Mapping (SPM) toolbox to carry out tissue segmentation, calculate tissue fractions within voxels, derive mean metabolite values from regions of interest defined by anatomical or functional atlases, and perform linear regression for global white matter and gray matter metabolite concentrations. The library’s workings are demonstrated for two tasks: (1) Pre-processing, fitting and analysis of single-voxel proton MRS data from healthy volunteers; and (2) Extracting region-specific metabolite concentrations from spectroscopic imaging data based on an existing cortical atlas in MNI space, and calculating average gray and white matter global concentrations using linear regression.

Purpose: Proton magnetic resonance spectroscopy (1H MRS) offers biomarkers of metabolic damage after mild traumatic brain injury (mTBI), but a lack of replicability studies hampers clinical translation. In a conceptual replication study design, the results reported in four previous publications were used as the hypotheses (H1-H7), specifically: abnormalities in patients are diffuse (H1), confined to white matter (WM) (H2), comprise low N-acetyl-aspartate (NAA) levels and normal choline (Cho), creatine (Cr) and myo-inositol (mI) (H3), and correlate with clinical outcome (H4); additionally, a lack of findings in regional subcortical WM (H5) and deep gray matter (GM) structures (H6), except for higher mI in patients’ putamen (H7). Methods: 26 mTBI patients (20 female, age 36.5  12.5 [mean  standard deviation] years), within two months from injury and 21 age-, sex-, and education-matched healthy controls were scanned at 3 Tesla with 3D echo-planar spectroscopic imaging. To test H1-H3, global analysis using linear regression was used to obtain metabolite levels of GM and WM in each brain lobe. For H4, patients were stratified into non-recovered and recovered subgroups using the Glasgow Outcome Scale Extended. To test H5-H7, regional analysis using spectral averaging estimated metabolite levels in four GM and six WM structures segmented from T1-weighted MRI. The Mann-Whitney U test and weighted least squares analysis of covariance were used to examine mean group differences in metabolite levels between all patients and all controls (H1-H3, H5- H7), and between recovered and non-recovered patients and their respectively matched controls (H4). Replicability was defined as the support or failure to support the null hypotheses in accordance with the content of H1-H7, and was further evaluated using percent differences, coefficients of variation, and effect size (Cohen’s d). Results: Patients’ occipital lobe WM Cho and Cr levels were 6.0% and 4.6% higher than controls’, respectively (Cho, d = 0.37, p = 0.04; Cr, d = 0.63, p = 0.03). The same findings, i.e., higher patients’ occipital lobe WM Cho and Cr (both p = 0.01), but with larger percent differences (Cho, 8.6%; Cr, 6.3%) and effect sizes (Cho, d = 0.52; Cr, d = 0.88) were found in the comparison of non-recovered patients to their matched controls. For the lobar WM Cho and Cr comparisons without statistical significance (frontal, temporal, parietal), unidirectional effect sizes were observed (Cho, d = 0.07 – 0.37; Cr, d = 0.27 – 0.63). No differences were found in any metabolite in any lobe in the comparison between recovered patients and their matched controls. In the regional analyses, no differences in metabolite levels were found in any GM or WM region, but all WM regions (posterior, frontal, corona radiata, and the genu, body, and splenium of the corpus 3 callosum) exhibited unidirectional effect sizes for Cho and Cr (Cho, d = 0.03 – 0.34; Cr, d = 0.16 – 0.51). Conclusions: We replicated findings of diffuse WM injury, which correlated with clinical outcome (supporting H1-H2, H4). These findings, however, were among the glial markers Cho and Cr, not the neuronal marker NAA (not supporting H3). No differences were found in regional GM and WM metabolite levels (supporting H5-H6), nor in putaminal mI (not supporting H7). Unidirectional effect sizes of higher patients’ Cho and Cr within all WM analyses suggest widespread injury, and are in line with the conclusion from the previous publications, i.e., that detection of WM injury may be more dependent upon sensitivity of the 1H MRS technique than on the selection of specific regions. The findings lend further support to the corollary that clinic-ready 1H MRS biomarkers for mTBI may best be achieved by using high signal-to-noise-ratio single-voxels placed anywhere within WM. The biochemical signature of the injury, however, may differ and therefore absolute levels, rather than ratios may be preferred. Future replication efforts should further test the generalizability of these findings. Funding Information: This work was supported by the United States’ National Institutes of Health (NIH) grant number R01NS097494. GM acknowledges support from grant number R01EB026456, and all authors acknowledge the support of the Center for Advanced Imaging Innovation and Research (CAI2R) at NYU Langone Health under grant number P41EB017183. Declaration of Interests: The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Ethics Approval Statement: All subjects provided written informed consent prior to their examination. This study was approved by the New York University Grossman School of Medicine Institutional Review Board and written informed consent was obtained from all subjects in accordance with the Declaration of Helsinki (FWA 00004952).

Objectives To assess whether MR fingerprinting (MRF)–based relaxation properties exhibit cross-sectional and prospective correlations with patient outcome and compare the results with those from DTI. Methods Clinical imaging, MRF, and DTI were acquired in patients (24 ± 10 days after injury (timepoint 1) and 90 ± 17 days after injury (timepoint 2)) and once in controls. Patient outcome was assessed with global functioning, symptom profile, and neuropsychological testing. ADC and fractional anisotropy (FA) from DTI and T₁ and T₂ from MRF were compared in 12 gray and white matter regions with Mann–Whitney tests. Bivariate associations between MR measures and outcome were assessed using the Spearman correlation and logistic regression. Results Data from 22 patients (38 ± 12 years; 17 women) and 18 controls (32 ± 8 years; 12 women) were analyzed. Fourteen patients (37 ± 12 years; 11 women) returned for timepoint 2, while two patients provided only timepoint 2 clinical outcome data. At timepoint 1, there were no differences between patients and controls in T₁, T₂, and ADC, while FA was lower in mTBI frontal white matter. T₁ at timepoint 1 and the change in T₁ exhibited more (n = 18) moderate to strong correlations (|r|= 0.6–0.85) with clinical outcome at timepoint 2 than T₂ (n = 3), FA (n = 7), and ADC (n = 2). High T₁ at timepoint 1, and serially increasing T₁, accounted for five of the six MR measures with the highest utility for identification of non-recovered patients at timepoint 2 (AUC > 0.80). Conclusion T₁ derived from MRF was found to have higher utility than T₂, FA, and ADC for predicting 3-month outcome after mTBI. Key Points • In a region-of-interest approach, FA, ADC, and T₁ and T₂ all showed limited utility in differentiating patients from controls at an average of 24 and 90 days post-mild traumatic brain injury. • T₁ at 24 days, and the serial change in T₁, revealed more and stronger predictive correlations with clinical outcome at 90 days than did T₂, ADC, or FA. • T₁ showed better prospective identification of non-recovered patients at 90 days than ADC, T₂, and FA.

The hippocampus is one of the most challenging brain regions for proton MR spectroscopy (MRS) applications. Moreover, quantification of J‐coupled species such as myo‐inositol (m‐Ins) and glutamate + glutamine (Glx) is affected by the presence of macromolecular background. While long echo time (TE) MRS eliminates the macromolecules, it also decreases the m‐Ins and Glx signal and, as a result, these metabolites are studied mainly with short TE. Here, we investigate the feasibility of reproducibly measuring their concentrations at a long TE of 120 ms, using a semi‐adiabatic localization by adiabatic selective refocusing (sLASER) sequence, as this sequence was recently recommended as a standard for clinical MRS. Gradient offset‐independent adiabatic refocusing pulses were implemented, and an optimal long TE for the detection of m‐Ins and Glx was determined using the T₂ relaxation times of macromolecules. Metabolite concentrations and their coefficients of variation (CVs) were obtained for a 3.4‐mL voxel centered on the left hippocampus on 3‐T MR systems at two different sites with three healthy subjects (aged 32.5 ± 10.2 years [mean ± standard deviation]) per site, with each subject scanned over two sessions, and with each session comprising three scans. Concentrations of m‐Ins, choline, creatine, Glx and N‐acetyl‐aspartate were 5.4 ± 1.5, 1.7 ± 0.2, 5.8 ± 0.3, 11.6 ± 1.2 and 5.9 ± 0.4 mM (mean ± standard deviation), respectively. Their respective mean within‐session CVs were 14.5% ± 5.9%, 6.5% ± 5.3%, 6.0% ± 3.4%, 10.6% ± 6.2% and 3.5% ± 1.4%, and their mean within‐subject CVs were 17.8% ± 18.2%, 7.5% ± 6.3%, 7.4% ± 6.4%, 12.4% ± 5.3% and 4.8% ± 3.0%. The between‐subject CVs were 25.0%, 12.3%, 5.3%, 10.7% and 6.4%, respectively. Hippocampal long‐TE sLASER single voxel spectroscopy can provide macromolecule‐independent assessment of all major metabolites including Glx and m‐Ins.

Proton (¹H) magnetic resonance spectroscopy provides a non-invasive and quantitative measure of brain metabolites. Traumatic brain injury impacts cerebral metabolism and a number of research groups have successfully used this technique as a biomarker of injury and/or outcome in both pediatric and adult TBI populations. However, this technique is underutilized, with studies being performed primarily at centers with access to MR research support. In this paper we present a technical introduction to the acquisition and analysis of in vivo ¹H magnetic resonance spectroscopy and review ¹H magnetic resonance spectroscopy findings in different injury populations. In addition, we propose a basic ¹H magnetic resonance spectroscopy data acquisition scheme (Supplemental Information) that can be added to any imaging protocol, regardless of clinical magnetic resonance platform. We outline a number of considerations for study design as a way of encouraging the use of ¹H magnetic resonance spectroscopy in the study of traumatic brain injury, as well as recommendations to improve data harmonization across groups already using this technique.

The pathological cascade of tissue damage in mild traumatic brain injury is set forth by a perturbation in ionic homeostasis. However, whether this class of injury can be detected in vivo and serve as a surrogate marker of clinical outcome is unknown. We employ sodium MRI to test the hypotheses that regional and global total sodium concentrations: (i) are higher in patients than in controls and (ii) correlate with clinical presentation and neuropsychological function. Given the novelty of sodium imaging in traumatic brain injury, effect sizes from (i), and correlation types and strength from (ii), were compared to those obtained using standard diffusion imaging metrics. Twenty-seven patients (20 female, age 35.9 ± 12.2 years) within 2 months after injury and 19 controls were scanned with proton and sodium MRI at 3 Tesla. Total sodium concentration, fractional anisotropy and apparent diffusion coefficient were obtained with voxel averaging across 12 grey and white matter regions. Linear regression was used to obtain global grey and white matter total sodium concentrations. Patient outcome was assessed with global functioning, symptom profiles and neuropsychological function assessments. In the regional analysis, there were no statistically significant differences between patients and controls in apparent diffusion coefficient, while differences in sodium concentration and fractional anisotropy were found only in single regions. However, for each of the 12 regions, sodium concentration effect sizes were uni-directional, due to lower mean sodium concentration in patients compared to controls. Consequently, linear regression analysis found statistically significant lower global grey and white matter sodium concentrations in patients compared to controls. The strongest correlation with outcome was between global grey matter sodium concentration and the composite z-score from the neuropsychological testing. In conclusion, both sodium concentration and diffusion showed poor utility in differentiating patients from controls, and weak correlations with clinical presentation, when using a region-based approach. In contrast, sodium linear regression, capitalizing on partial volume correction and high sensitivity to global changes, revealed high effect sizes and associations with patient outcome. This suggests that well-recognized sodium imbalances in traumatic brain injury are (i) detectable non-invasively; (ii) non-focal; (iii) occur even when the antecedent injury is clinically mild. Finally, in contrast to our principle hypothesis, patients’ sodium concentrations were lower than controls, indicating that the biological effect of traumatic brain injury on the sodium homeostasis may differ from that in other neurological disorders. Note: This figure has been annotated.