We aimed to identify the white matter (WM) and grey matter (GM) abnormalities in multiple system atrophy (MSA) and assess the utility of longitudinal structural and diffusion changes as surrogate markers for tracking disease progression in MSA.

BACKGROUND:White matter (WM) abnormalities have been implicated in clinically relevant functional decline in multiple system atrophy (MSA). OBJECTIVE:To identify the WM and gray matter (GM) abnormalities in MSA and assess the utility of longitudinal structural and diffusion changes as surrogate markers for tracking disease progression in MSA. METHODS:Twenty-seven participants with early MSA [15 with clinically predominant cerebellar (MSA-C) and 12 with clinically predominant parkinsonian features (MSA-P)] and 14 controls were enrolled as a part of our prospective, longitudinal study of synucleinopathies. Using structural magnetic resonance imaging (MRI) and diffusion MRI (diffusion tensor and neurite orientation and dispersion density imaging), we analyzed whole and regional brain changes in these participants. We also evaluated temporal imaging trajectories based on up to three annual follow-up scans and assessed the impact of baseline diagnosis on these imaging biomarkers using mixed-effect models. RESULTS:MSA patients exhibited more widespread WM changes than GM, particularly in the cerebellum and brainstem, with greater severity in MSA-C. Structural and diffusion measures in the cerebellum WM and brainstem deteriorated with disease progression. Rates of progression of these abnormalities were similar in both MSA subtypes, reflecting increasing overlap of clinical features over time. CONCLUSION:WM abnormalities are core features of MSA disease progression and advance at similar rates in clinical MSA subtypes. Multimodal MRI imaging reveals novel insights into the distribution and pattern of brain abnormalities and their progression in MSA. Selected structural and diffusion measures may be useful for tracking disease progression in MSA clinical trials.

Objective To systematically evaluate structural MRI and diffusion MRI features for cross-sectional discrimination and tracking of longitudinal disease progression in early multiple system atrophy (MSA). Methods In a prospective, longitudinal study of synucleinopathies with imaging on 14 controls and 29 MSA patients recruited at an early disease stage (15 predominant cerebellar ataxia subtype or MSA-C and 14 predominant parkinsonism subtype or MSA-P), we computed regional morphometric and diffusion MRI features. We identified morphometric features by ranking them based on their ability to distinguish MSA-C from controls and MSA-P from controls and evaluated diffusion changes in these regions. For the top performing regions, we evaluated their utility for tracking longitudinal disease progression using imaging from 12-month follow-up and computed sample size estimates for a hypothetical clinical trial in MSA. We also computed these selected morphometric features in an independent validation dataset. Results We found that morphometric changes in the cerebellar white matter, brainstem, and pons can separate early MSA-C patients from controls both cross-sectionally and longitudinally (p < 0.01). The putamen and striatum, though useful for separating early MSA-P patients from control subjects at baseline, were not useful for tracking MSA disease progression. Cerebellum white matter diffusion changes aided in capturing early disease related degeneration in MSA. Interpretation Regardless of clinically predominant features at the time of MSA assessment, brainstem and cerebellar pathways progressively deteriorate with disease progression. Quantitative measurements of these regions are promising biomarkers for MSA diagnosis in early disease stage and potential surrogate markers for future MSA clinical trials.

BACKGROUND:Chronic kidney disease (CKD), a growing public health issue in the elderly, is associated with increased risk of cognitive impairment.OBJECTIVE:To investigate the mechanisms through which CKD impacts brain health using longitudinal imaging.METHODS:We identified 97 participants (74 CKD and 23 non-CKD) from the BRINK (BRain IN Kidney Disease), a longitudinal study of CKD with two MRI scans (baseline and 3-year follow-up). We measured the associations between baseline and change in kidney disease biomarkers of estimated glomerular filtration rate (eGFR) and urinary albumin to creatinine ratio (UACR), considered a measure of microvascular inflammation, and imaging outcomes of cortical thickness and ventricular volume from structural MRI, white matter hyperintensities (WMH) volume from FLAIR images, and fractional anisotropy of the corpus callosum (FACC).RESULTS:There were white matter-specific changes as observed by increased WMH volume and decreased FACC in CKD participants, as well as ventricular volume increase in both CKD and non-CKD groups reflective of aging-related changes. Decline in eGFR was associated with decrease in the FACC, suggesting that subtle early white matter changes due to kidney disease can be captured using DTI. An increase in UACR was associated with increase in ventricular volume.CONCLUSION:Our results support the role of eGFR as a measure of kidney microvascular disease which is associated with concurrent white matter damage in CKD. Future work is needed to investigate the possible link between endothelial microvascular inflammation (as measured by an increased UACR) and ventricular volume increase.

AbstractBackgroundWe report the associations between change in renal biomarkers in CKD and non‐ CKD participants and change in structural MRI outcomes over 3 years in the BRain IN Kidney disease (BRINK) study MRI cohort. The imaging outcomes were cortical thickness in Alzheimer’s disease signature regions, white matter hyperintensities (WMH), and fractional anisotropy in the corpus callosum (FA CC), an early marker of cerebrovascular disease.MethodsParticipants were ≥ 45 years with CKD (estimated glomerular filtration rate (eGFR) in mL/min/1.73 m2< 60; non‐ dialysis) or non – CKD (eGFR ≥ 60) in Minneapolis/St. Paul. CKD severity was measured by the eGFR and urine albumin creatinine ratio (UACR‐albuminuria, a measure of microvascular disease). Brain images were obtained on a 1.5 T Phillips Ingenia machine at baseline and 3 years. Linear regression models measured the associations between 3 year change in eGFR and UACR, as predictors of change in MRI measures: AD cortical thickness, WMH, and FA CC over 3 years. Adjusted models included age, gender, AA race, education years, cerebrovascular disease risk factor score (CVD RF), smoker/alcohol use, and pulse pressure as covariates, and TIV for change in WMH models.ResultsAt baseline, mean age of the 97 participants was 67 years, and mean eGFR was 37 for the CKD group vs. 83.2 for non‐ CKD (Table 1). Mean change in eGFR and UACR in the CKD group were ‐3.6 and +212, respectively compared with ‐ 8.5 and +1.7 in the non‐ CKD group. Mean WMH volume increase (+5.8 cc) and mean FA CC decrease (‐0.012) were significant in CKD (p < 0.001) for both (Table 2). In adjusted models both a 1% decrease and a ≥ 3 unit/year decline in eGFR were associated with a decline in FA CC over 3 years in the CKD group (β (SE) =‐0.020 (0.007); p = 0.004) and ‐0.0097 (0.0042); p = 0.025, respectively). Cortical thickness models were not significant. Unexpectedly, change in UACR was not associated with change in any MRI outcomes.ConclusionsDecreased eGFR was associated with decreased FA CC in CKD, suggesting that CKD progression exerts its influence largely through cerebrovascular mechanisms.

Elevated systolic blood pressure (SBP) and diastolic blood pressure (DBP) are risk factors for white matter (WM) injury associated with cognitive decline and dementia. White matter hyperintensities (WMH) on MRI, markers of WM injury, increase with aging and are more common in women than in men after the age of 60. We examined SBP and DBP in relation to WMH volume and diffusion MRI measures of WM integrity in postmenopausal women.

The Alzheimer's Disease Neuroimaging Initiative began its third phase, ADNI3, in late 2016. As with the ADNI1 to 2 transition, ADNI3 updates the study with recent MRI advances. ADNI2 included DTI, but only for GE scanners. ADNI3 includes diffusion for all manufacturers involved (GE, Philips, and Siemens), and emphasizes consistency across scanners. The resolution has been improved from 2.7 to 2.0mm (isotropic), and, where simultaneous multislice (SMS) acceleration is available, acquires three diffusion shells, supporting advanced diffusion analyses such as NODDI1 and DKI2. To serve the general dementia research community, the diffusion protocol was designed to support three different analyses: DTI, with an emphasis on cross-sectional and longitudinal consistency; tractography, calling for more directions at higher diffusion weighting (b); and multi-compartment analyses such as NODDI and DKI, which require > 2 b values. Each depends on spatial resolution and SNR, but the time available for diffusion in ADNI3 is limited to around 8 minutes, and most scanners will not have SMS capability until partway through ADNI3. This strongly favors a HARDI-type acquisition with two protocols, a basic b=1000 s/mm2 shell, and for SMS-capable scanners, the addition of shells at b = 500 and 2000s/mm2. The directions were evenly spaced using an electrostatic repulsion algorithm3. Figure 1 shows the ADNI3 diffusion parameters.

Phase three of the Alzheimer's Disease Neuroimaging Initiative (ADNI-3) began in late 2016. MRI for ADNI-1 focused on consistent longitudinal structural imaging on 1.5T scanners. ADNI-2 imaging was performed at 3T with structural imaging similar to ADNI-1, adding 2D FLAIR and T2*-weighted imaging at all sites. Advanced imaging (diffusion imaging, resting state fMRI, or arterial spin labeling) was included depending on scanner manufacturer. ADNI-3 implements these three advanced sequences from ADNI-2 across all sites. To maximize protocol utility for other studies, only product sequences are used. MR sequence parameters are summarized in Table 1. Consistency across scanner models is important to support pooling of data across the 57 scanners used by 59 ADNI enrollment sites. Completely consistent protocols would, however, be limited by the least capable scanner. As a compromise, two sets of protocols (“basic” and “advanced”) were developed. The basic and advanced protocols share several sequences. Whole brain structure is assessed with T1-weighted inversion recovery and 3D FLAIR images. Higher spatial resolution images are acquired over a limited field of view centered on the hippocampus. T2-weighted images are acquired for assessment of cerebral microbleeds. Arterial spin labeling (ASL) images are acquired via 3D pCASL, 2D PASL or 3D PASL sequences, depending on availability. The protocols diverge for resting state fMRI and diffusion imaging. The bifurcation is driven largely by the availability of simultaneous multi-slice (SMS) accelerated imaging. Diffusion imaging is done using a single b=1000 s/mm2 shell in the basic protocol and with three shells (b=500, 1000 and 2000 s/mm2) in the advanced protocol. Resting state data are acquired with TR ∼600 ms in the advanced protocol and 3000 ms in the basic protocol. The breakdown of scanner manufacturers and models for the 57 scanners is shown in Table 2. The commercial availability of SMS in the U.S. is new and still increasing; initially all scanners have been outfitted with the basic protocol. As scanners are upgraded, it is estimated that approximately 30–45% of scanners upgrade to an advanced protocol by 2019. ADNI-3 data will be collected using basic and advanced protocols. Structural imaging is implemented consistently across the protocols.

Phase three of the Alzheimer's Disease Neuroimaging Initiative (ADNI-3) began in late 2016. MRI for ADNI-1 focused on consistent longitudinal structural imaging on 1.5T scanners. ADNI-2 imaging was performed at 3T with structural imaging similar to ADNI-1, adding 2D FLAIR and T2*-weighted imaging at all sites. Advanced imaging (diffusion imaging, resting state fMRI, or arterial spin labeling) was included depending on scanner manufacturer. ADNI-3 implements these three advanced sequences from ADNI-2 across all sites. To maximize protocol utility for other studies, only product sequences are used. MR sequence parameters are summarized in Table 1. Consistency across scanner models is important to support pooling of data across the 57 scanners used by 59 ADNI enrollment sites. Completely consistent protocols would, however, be limited by the least capable scanner. As a compromise, two sets of protocols ("basic" and "advanced") were developed. The basic and advanced protocols share several sequences. Whole brain structure is assessed with T1-weighted inversion recovery and 3D FLAIR images. Higher spatial resolution images are acquired over a limited field of view centered on the hippocampus. T2*-weighted images are acquired for assessment of cerebral microbleeds. Arterial spin labeling (ASL) images are acquired via 3D pCASL, 2D PASL or 3D PASL sequences, depending on availability. The protocols diverge for resting state fMRI and diffusion imaging. The bifurcation is driven largely by the availability of simultaneous multi-slice (SMS) accelerated imaging. Diffusion imaging is done using a single b=1000 s/mm2shell in the basic protocol and with three shells (b=500, 1000 and 2000 s/mm2) in the advanced protocol. Resting state data are acquired with TR ∼600 ms in the advanced protocol and 3000 ms in the basic protocol. The breakdown of scanner manufacturers and models for the 57 scanners is shown in Table 2. The commercial availability of SMS in the U.S. is new and still increasing; initially all scanners have been outfitted with the basic protocol. As scanners are upgraded, it is estimated that approximately 30–45% of scanners upgrade to an advanced protocol by 2019. ADNI-3 data will be collected using basic and advanced protocols. Structural imaging is implemented consistently across the protocols.

BACKGROUNDChronic kidney disease (CKD) studies have reported variable prevalence of brain pathologies, in part due to low inclusion of participants with moderate to severe CKD.OBJECTIVETo measure the association between kidney function biomarkers and brain MRI findings in CKD.METHODSIn the BRINK (BRain IN Kidney Disease) study, MRI was used to measure gray matter volumes, cerebrovascular pathologies (white matter hyperintensity (WMH), infarctions, microhemorrhages), and microstructural changes using diffusion tensor imaging (DTI). We performed regression analyses with estimated glomerular filtration rate (eGFR) and urine albumin to creatinine ratio (UACR) as primary predictors, and joint models that included both predictors, adjusted for vascular risk factors.RESULTSWe obtained 240 baseline MRI scans (150 CKD with eGFR <45 in ml/min/1.73 m2; 16 mild CKD: eGFR 45-59; 74 controls: eGFR≥60). Lower eGFR was associated with greater WMH burden, increased odds of cortical infarctions, and worsening diffusion changes throughout the brain. In eGFR models adjusted for UACR, only cortical infarction associations persisted. However, after adjusting for eGFR, higher UACR provided additional information related to temporal lobe atrophy, increased WMH, and whole brain microstructural changes as measured by increased DTI mean diffusivity.CONCLUSIONSBiomarkers of kidney disease (eGFR and UACR) were associated with MRI brain changes, even after accounting for vascular risk factors. UACR adds unique additional information to eGFR regarding brain structural and diffusion biomarkers. There was a greater impact of kidney function biomarkers on cerebrovascular pathologies and microstructural brain changes, suggesting that cerebrovascular etiology may be the primary driver of cognitive impairment in CKD.

Background:White matter hyperintensities (WMH), a marker of small vessel disease, are common in Alzheimer’s disease (AD) and global measures of WMH have been associated with increased neurodegeneration (MRI atrophy, FDG-PET hypometabolism) and cognitive function. However, the impact of local WMH onto neural network function in AD remains unclear. The current study aimed to test whether the association between WMH within fiber tracts of the default mode network (DMN) are associated with reduced functional connectivity (FC), independent of amyloid-beta (Ab) in Alzheimer’s disease (AD). We hypothesized that higher WMH volume in a particular DMN fiber tract is associated with reduced functional MRI (fMRI) assessed FC and cognition between those brain areas that the fiber tract connects. Methods: The study included subjects recruited from ADNI with AD dementia (n1⁄422), mild cognitive impairment (n1⁄442), and healthy cognition (n1⁄414), all with high Ab burden (global (g)AV-45 PET binding > 1.11), and a control group with low Ab burden (n 1⁄4 24; global (g)AV-45 PET binding < 1.11). WMH were classified on FLAIR images, superimposed on fiber tract maps of the DMN, and tractspecific ratio of WMH volume vs. tract volume (WMHr) were calculated. The resting state fMRI based DMN maps, determined for each subject by independent component analysis, were superimposed on the DMN fiber tract map to extract FC values at the projection zone of the fiber tracts. The associations between tract-specific WMHr and FC ROI values and cognition (executive function or memory composite scores) were tested using robust regression analysis, controlling for age, gAV-45 PET, diagnosis and education. Results:Higher WMHr within the inferior fronto-occipital fasciculus (IFOF) was associated with decreased FC in the projection areas of the IFOF (p1⁄40.02), independent of gAV-45 PET binding. Independently from WMHr, greater gAV-45 PET binding was associated with decreased FC within the projection areas of the IFOF, cingulum-hippocampal and the superior longitudinal fasciculus tracts (p<0.05). A marginal association was observed between IFOF WMHr and executive function (p1⁄40.06). Conclusions: The results suggest that WMH disrupt functional connectivity in a fiber-tract specific way, independent from Ab. WMHmay thus contribute to dysfunction of neural networks in AD. O1-08-05 ASSOCIATION OF NEURODEGENERATIVE AND CEREBROVASCULAR IMAGING WITH KIDNEY FUNCTION: BASELINE BRINK MRI

The goal of this cross-sectional study was to compare neurodegenerative and cerebrovascular disease imaging in subjects with and without Chronic Kidney Disease (CKD) using MRI, given previous evidence for pathophysiological interactions between brain and kidney. We used i) cortical thickness using structural MRI as a marker of change in gray matter, ii) fractional anisotropy (FA) using diffusion tensor MRI (DTI) as a marker of changes in white matter integrity, and iii) white matter hyperintensity (WMH) burden, brain lesion and microhemorrhage assessment using FLAIR and GRE MRI as a marker of cerebrovascular disease. We studied 245 subjects (155 CKD, 15 mild-CKD and 75 controls) from BRINK (BRain IN Kidney Disease), an NIH funded longitudinal study of cognitive impairment and stroke in CKD patients who had baseline MRIs. We computed cortical thickness and hippocampal volumes using Freesurfer v5.3 on structural MRIs and computed FA, WMH, brain lesion, and microhemorrhage assessments using in-house tools. We performed two analyses – 1) dichotomous analyses (t-test or Poisson regression) for each of the MRI measures to test group differences between CKD and controls, and 2) Pearson's correlation with Glomerular filtration rate (GFR – a measure of kidney function) specifically for WMH and Structural MRI measures after controlling for age, sex, education, diabetes, and race. There was significant thinning of the temporal and frontal cortices, and lower hippocampal volume in CKD patients (p<0.05). There was reduced FA in deep gray and white matter, brainstem, periventricular parietal, subcortical parietal, and subcortical frontal regions in CKD patients (p<0.05). There were larger WMH volumes, higher numbers of cortical infarctions, subcortical infarctions, and microhemorrhages (p<0.05) in CKD patients compared to controls. There was a significant association between frontal cortex thickness and GFR (r=0.15 and p=0.02) and WMH volumes and GFR (r=-0.17 and p=0.008). There appears to be a stronger association of kidney disease with cerebrovascular than neurodegenerative etiology given the significantly larger numbers of microhemorrhages, cortical infarctions and subcortical infarcts; and also larger WMH volume and cortical thinning in the frontal cortices. This study provides evidence that cognitive impairment may be driven by both cerebrovascular and neurodegenerative changes in CKD patients.

Imaging in neuroscience, clinical research and pharmaceutical trials often employs the 3D magnetisation-prepared rapid gradient-echo (MPRAGE) sequence to obtain structural T1-weighted images with high spatial resolution of the human brain. Typical research and clinical routine MPRAGE protocols with ~1mm isotropic resolution require data acquisition time in the range of 5-10min and often use only moderate two-fold acceleration factor for parallel imaging. Recent advances in MRI hardware and acquisition methodology promise improved leverage of the MR signal and more benign artefact properties in particular when employing increased acceleration factors in clinical routine and research. In this study, we examined four variants of a four-fold-accelerated MPRAGE protocol (2D-GRAPPA, CAIPIRINHA, CAIPIRINHA elliptical, and segmented MPRAGE) and compared clinical readings, basic image quality metrics (SNR, CNR), and automated brain tissue segmentation for morphological assessments of brain structures. The results were benchmarked against a widely-used two-fold-accelerated 3T ADNI MPRAGE protocol that served as reference in this study. 22 healthy subjects (age=20-44yrs.) were imaged with all MPRAGE variants in a single session. An experienced reader rated all images of clinically useful image quality. CAIPIRINHA MPRAGE scans were perceived on average to be of identical value for reading as the reference ADNI-2 protocol. SNR and CNR measurements exhibited the theoretically expected performance at the four-fold acceleration. The results of this study demonstrate that the four-fold accelerated protocols introduce systematic biases in the segmentation results of some brain structures compared to the reference ADNI-2 protocol. Furthermore, results suggest that the increased noise levels in the accelerated protocols play an important role in introducing these biases, at least under the present study conditions.

The NIA BRINK (BRain IN Kidney) Disease study is a longitudinal study of cognitive impairment and stroke in chronic kidney disease (CKD) patients. Our goal was to compare baseline structural MRI abnormalities between CKD and non-CKD participants (ppts) in the BRINK study. We classified ppts as CKD if their estimated glomerular filtration rate (eGFR) was < 45 mL/min/1.73 m 2 and not on dialysis; non-CKD as eGFR ≥ 60. There were 109 CKD and 23 non-CKD ppts with usable baseline MRI scans. Gray matter volume regions of interest (ROIs) from Freesurfer, white matter hyperintensity (WMH) fraction, cortical (> 1 cm) and subcortical infarcts and micro-hemorrhages were assessed from structural MRI scans (T1, FLAIR, and GRE sequences on 1.5T). We tested for differences between CKD and non-CKD using linear regression for WMH fraction, Fisher's exact test for infarcts, and negative binomial regression for counts of definite micro-hemorrhages. Adjusted models controlled for age, race, sex, education, and diabetes. Participant characteristics are described in Table 1. Gray matter volume was lower in CKD vs. non-CKD ppts in ROIs that are associated with cerebrovascular disease (frontal lobes) and Alzheimer's disease (temporoparietal regions) (p<0.05). Global cortical volume was also significantly lower in CKD vs. non-CKD ppts (p<0.05). We also found higher mean WMH fraction in CKD participants (unadjusted p=0.03, adjusted p=0.17). Ten (9%) CKD ppts had > 1 small cortical or > 1 subcortical infarct vs. none in non-CKD (p=0.14). Twenty-three CKD ppts had 57 definite micro-hemorrhages, compared to 9 in 3 non-CKD ppts, but this difference was notstatistically significant. We found significantly lower gray matter volume in ROIs associated with both AD and vascular disease and greater WMH burden in CKD vs. non-CKD ppts. Prevalence of ppts with > 1 cortical or > 1 subcortical infarct and prevalence of ppts with micro-hemorrhages were also higher in CKD vs. non-CKD ppts, although these differences were not statistically significant. These results suggest that CKD is associated with structural changes in the brain that reflect both neurodegenerative and vascular etiologies. Further sutdy is needed to understand the correlation of the structural abnormalities with cognitive function. Sponsor:NIA AG037551-01A3.

Target Audience: MRI physicists and gradient coil designers. Background: Standard spatial encoding in MRI assumes uniform gradients, but in practice, implementing gradient linearity over the entire imaging field-of-view (FOV) is usually not feasible. Gradient non-linearity, if not properly compensated, causes geometric image distortion. Standard strategies to correct gradient non-uniformity on commercial systems (e.g., “GradWarp” or gradient distortion correction) are based on a parameterization of the gradient field and typically contain up to 5 order terms in the expansion. The correction coefficients are predetermined and the same set of coefficients is applied to all systems in the field. The purpose of this work is to develop a simple method to measure and fit the gradient correction on a per-system basis using the Alzheimer’s Disease Neuroimaging Initiative (ADNI) phantom. A 3D spherical harmonic approximation of the distortion is then determined and used for correction. Methods: The ADNI phantom contains 160 fiducial spheres (with a diameter of 1.0 or 1.5 cm) that are distributed within a 20-cm diameter spherical shell. The spatial positions of the spheres can be tracked using the AQUAL analysis software that is available with the phantom (see Ref. 6 for details). Differences between the expected locations of the fiducials and their positions measured from the distorted image can be expressed using a spherical harmonic model as in Eq.1, where and are measured and expected location in Cartesian coordinates, respectively, r, , and are polar coordinates, is the associated Legendre polynomial, N is the approximation order, and Anm and Bnm are model coefficients. Denoting H as the spherical harmonic basis and c the corresponding coefficient vector form, the right-hand side of Eq. 1 can be stated as Hc, and c can be estimated via the constrained optimization process in Eq. 2, where the constraint represents optional a priori knowledge of gradient system design. For example, if the gradient system design is symmetric about isocenter, we expect the even order terms to be zero. Such a constraint can be represented by setting ck=0, where ck is the coefficient of the kth null term. To test the proposed strategy, an 3D MP-RAGE sequence was performed on 3T GE Signa HDxt system (acquisition plane:sagittal, Nx=Ny=256, Nz=196, Δx=1.05mm, Δz=1.3mm) using the whole-body gradient with maximum gradient amplitude and slew rates of 40mT/m and 200mT/m/msec, respectively. The phantom was placed close to the scanner isocenter, and the slight shifting and rotation from the isocenter was estimated and corrected by rigid transformation. The spherical harmonic model was then constructed and the model coefficients were solved by the Nesterov’s optimal gradient method, with an iteration number 50. The symmetry of the gradient system was taken into account by constraining the even-order coefficients to be zero. Finally, image distortion was corrected by cubic spline interpolation using the estimated coefficients, and the corrected images were compared against images acquired without Gradwarp. Results: Examples of images before and after (1/2 row) the proposed gradient nonlinear correction with spherical harmonics of order N=5 are shown in Fig 1. Fig. 2 shows the error in fiducial displacement versus displacement along the three magnet axes, before and after correction. The geometrical distortion, which is most apparent in the sagittal and coronal planes, is successfully corrected (Fig. 1). The displacements of the fiducials due to nonlinear gradients can be effectively reduced by the proposed method, as illustrated in Fig. 2. The radial residual mean squared error (RMSE) was reduced from 3.27 mm to 0.32 mm after correction. Discussion: Ref. 6 reported that correction with scanner vendor-provided gradient warping method reduced the RMSE to about 0.3 mm, and we typically use 0.35 mm as cutoff point during quality control tests in the ADNI study. Comparably, the proposed method is able to achieve a similar degree of correction as vendors’ 3D correction. We emphasize however, that neither a vendor correction nor knowledge of the vendor’s correction coefficients are required for the proposed method, suggesting its potential use as an independent characterization and correction for the gradient system. The method is also flexible and can account for non-zero, even-order correction terms that are expected to be present in asymmetric, highperformance gradient coils, which are of interest for a head-only system. The 20-cm diameter of the ADNI phantom is well-matched to a typical head scan FOV. We also expect the method can be readily extended to include higher-order terms (i.e., N>5), if needed. Conclusions: In this work, we have demonstrated that geometrical distortion due to gradient nonlinearity can be successfully measured from image data acquired with the ADNI phantom and corrected via spherical harmonics fitting. The proposed method does not require direct measurement of the magnetic field or detailed knowledge of gradient coil design, and thus can be used in a variety of settings. While feasibility of the method was demonstrated on a symmetric gradient system whose even-order correction coefficients are zero, the method is expected to be flexible enough to handle an asymmetric gradients, such as those used in a head-only system. References: [1] L. Schad et al., MRI 10:609-21, 1992; [2] S. Doran et al., Phys Med Biol 50:1343-61, 2005; [3] G. Glover et al., U.S. Patent 4591789, 1986; [4] A. Janke et al., MRM 52(1):115-22, 2004; [5] C. Jack, Jr., et al., JMRI 27(4):685-91, 2008; [6] J. Gunter et al., Med Phys 36(6):2193-2205, 2009; [7] Y. Nesterov et al., Sov Math Dokl 27:372-6, 1983; [8] B. O’Donoghue et al., Found Comput Math. In press. [9] J. Mathieu et al., ISMRM 2013: 2708. Acknowledgements: This work was supported in part by the NIH grant 5R01EB010065. Sagittal Coronal Axial