Tract-Based Spatial Statistics (TBSS) is a popular software pipeline to coregister sets of diffusion tensor Fractional Anisotropy (FA) images for performing voxel-wise comparisons. It is primarily defined by its skeleton projection step intended to reduce effects of local misregistration. A white matter "skeleton" is computed by morphological thinning of the inter-subject mean FA, and then all voxels are projected to the nearest location on this skeleton. Here we investigate several enhancements to the TBSS pipeline based on recent advances in registration for other modalities, principally based on groupwise registration with the ANTS-SyN algorithm. We validate these enhancements using simulation experiments with synthetically-modified images. When used with these enhancements, we discover that TBSS's skeleton projection step actually reduces algorithm accuracy, as the improved registration leaves fewer errors to warrant correction, and the effects of this projection's compromises become stronger than those of its benefits. In our experiments, our proposed pipeline without skeleton projection is more sensitive for detecting true changes and has greater specificity in resisting false positives from misregistration. We also present comparative results of the proposed and traditional methods, both with and without the skeleton projection step, on three real-life datasets: two comparing differing populations of Alzheimer's disease patients to matched controls, and one comparing progressive supranuclear palsy patients to matched controls. The proposed pipeline produces more plausible results according to each disease's pathophysiology.

The clinical usefulness MRI biomarkers for aging and dementia studies relies on precise brain morphological measurements; however, scanner and/or protocol variations may introduce noise or bias. One approach to address this is post-acquisition scan harmonization. In this work, we evaluate deep learning (neural style transfer, CycleGAN and CGAN), histogram matching, and statistical (ComBat and LongComBat) methods. Participants who had been scanned on both GE and Siemens scanners (cross-sectional participants, known as Crossover (n = 113), and longitudinally scanned participants on both scanners (n = 454)) were used. The goal was to match GE MPRAGE (T1-weighted) scans to Siemens improved resolution MPRAGE scans. Harmonization was performed on raw native and preprocessed (resampled, affine transformed to template space) scans. Cortical thicknesses were measured using FreeSurfer (v.7.1.1). Distributions were checked using Kolmogorov-Smirnov tests. Intra-class correlation (ICC) was used to assess the degree of agreement in the Crossover datasets and annualized percent change in cortical thickness was calculated to evaluate the Longitudinal datasets. Prior to harmonization, the least agreement was found at the frontal pole (ICC = 0.72) for the raw native scans, and at caudal anterior cingulate (0.76) and frontal pole (0.54) for the preprocessed scans. Harmonization with NST, CycleGAN, and HM improved the ICCs of the preprocessed scans at the caudal anterior cingulate (>0.81) and frontal poles (>0.67). In the Longitudinal raw native scans, over- and under-estimations of cortical thickness were observed due to the changing of the scanners. ComBat matched the cortical thickness distributions throughout but was not able to increase the ICCs or remove the effects of scanner changeover in the Longitudinal datasets. CycleGAN and NST performed slightly better to address the cortical thickness variations between scanner change. However, none of the methods succeeded in harmonizing the Longitudinal dataset. CGAN was the worst performer for both datasets. In conclusion, the performance of the methods was overall similar and region dependent. Future research is needed to improve the existing approaches since none of them outperformed each other in terms of harmonizing the datasets at all ROIs. The findings of this study establish framework for future research into the scan harmonization problem.

To develop and validate a tool for Alzheimer's disease (AD) diagnosis in individual subjects using support vector machine (SVM)-based classification of structural MR (sMR) images.

Recent advances in automated face recognition algorithms have increased the risk that de-identified research MRI scans may be re-identifiable by matching them to identified photographs using face recognition. A variety of software exist to de-face (remove faces from) MRI, but their ability to prevent face recognition has never been measured and their image modifications can alter automated brain measurements. In this study, we compared three popular de-facing techniques and introduce our mri_reface technique designed to minimize effects on brain measurements by replacing the face with a population average, rather than removing it. For each technique, we measured 1) how well it prevented automated face recognition (i.e. effects on exceptionally-motivated individuals) and 2) how it altered brain measurements from SPM12, FreeSurfer, and FSL (i.e. effects on the average user of de-identified data). Before de-facing, 97% of scans from a sample of 157 volunteers were correctly matched to photographs using automated face recognition. After de-facing with popular software, 28-38% of scans still retained enough data for successful automated face matching. Our proposed mri_reface had similar performance with the best existing method (fsl_deface) at preventing face recognition (28-30%) and it had the smallest effects on brain measurements in more pipelines than any other, but these differences were modest.

Since tau PET tracers were introduced, investigators have quantified them using a wide variety of automated methods. As longitudinal cohort studies acquire second and third time points of serial within-person tau PET data, determining the best pipeline to measure change has become crucial. We compared a total of 415 different quantification methods (each a combination of multiple options) according to their effects on a) differences in annual SUVR change between clinical groups, and b) longitudinal measurement repeatability as measured by the error term from a linear mixed-effects model. Our comparisons used MRI and Flortaucipir scans of 97 Mayo Clinic study participants who clinically either: a) were cognitively unimpaired, or b) had cognitive impairments that were consistent with Alzheimer's disease pathology. Tested methods included cross-sectional and longitudinal variants of two overarching pipelines (FreeSurfer 6.0, and an in-house pipeline based on SPM12), three choices of target region (entorhinal, inferior temporal, and a temporal lobe meta-ROI), five types of partial volume correction (PVC) (none, two-compartment, three-compartment, geometric transfer matrix (GTM), and a tau-specific GTM variant), seven choices of reference region (cerebellar crus, cerebellar gray matter, whole cerebellum, pons, supratentorial white matter, eroded supratentorial WM, and a composite of eroded supratentorial WM, pons, and whole cerebellum), two choices of region masking (GM or GM and WM), and two choices of statistic (voxel-wise mean vs. median). Our strongest findings were: 1) larger temporal-lobe target regions greatly outperformed entorhinal cortex (median sample size estimates based on a hypothetical clinical trial were 520-526 vs. 1740); 2) longitudinal processing pipelines outperformed cross-sectional pipelines (median sample size estimates were 483 vs. 572); and 3) reference regions including supratentorial WM outperformed traditional cerebellar and pontine options (median sample size estimates were 370 vs. 559). Altogether, our results favored longitudinally SUVR methods and a temporal-lobe meta-ROI that includes adjacent (juxtacortical) WM, a composite reference region (eroded supratentorial WM + pons + whole cerebellum), 2-class voxel-based PVC, and median statistics.

The clinical diagnosis of Alzheimer's disease (AD) does not exactly match the pathological findings at autopsy in every subject. Therefore, in-vivo imaging measures, such as Magnetic Resonance Imaging (MRI) that reflect underlying pathology, would be clinically useful independent supplementary measures of disease stage. We have developed an algorithm that extracts atrophy information from individual patient's 3D MRI scans and assigns a STructural Abnormality iNDex (STAND)-score to the scan based on the degree of atrophy in comparison to patterns extracted from a large library of clinically well characterized AD and CN (cognitively normal) subject's MRI scans. STAND-scores can be adjusted for demographics to give adjusted-STAND (aSTAND)-scores which are >0 for subjects with brains identified as abnormal by the algorithm. Since histopathological findings are considered to represent the "ground truth", our objective was to assess the sensitivity of aSTAND-scores to pathological AD staging. This was done by comparing antemortem MRI based aSTAND-scores with postmortem grading of disease severity in 101 subjects who had both antemortem MRI and postmortem Braak neurofibrillary tangle (NFT) staging. We found a rank correlation of 0.62 (p<0.0001) between Braak NFT stage and aSTAND-scores. The results show that optimally extracted information from MRI scans such as STAND-scores accurately capture the severity of neuronal pathology and can be used as an independent approximate surrogate marker for in-vivo pathological staging as well as for early identification of AD in individual subjects.

It is well known that de-identified research brain images from MRI and CT can potentially be re-identified using face recognition; however, this has not been examined for PET images. We generated face reconstruction images of 182 volunteers using amyloid, tau, and FDG PET scans, and we measured how accurately commercial face recognition software (Microsoft Azure's Face API) automatically matched them with the individual participants' face photographs. We then compared this accuracy with the same experiments using participants' CT and MRI. Face reconstructions from PET images from PET/CT scanners were correctly matched at rates of 42% (FDG), 35% (tau), and 32% (amyloid), while CT were matched at 78% and MRI at 97-98%. We propose that these recognition rates are high enough that research studies should consider using face de-identification ("de-facing") software on PET images, in addition to CT and structural MRI, before data sharing. We also updated our mri_reface de-identification software with extended functionality to replace face imagery in PET and CT images. Rates of face recognition on de-faced images were reduced to 0-4% for PET, 5% for CT, and 8% for MRI. We measured the effects of de-facing on regional amyloid PET measurements from two different measurement pipelines (PETSurfer/FreeSurfer 6.0, and one in-house method based on SPM12 and ANTs), and these effects were small: ICC values between de-faced and original images were > 0.98, biases were <2%, and median relative errors were < 2%. Effects on global amyloid PET SUVR measurements were even smaller: ICC values were 1.00, biases were <0.5%, and median relative errors were also <0.5%.

When using imaging to predict time to progression from mild cognitive impairment (MCI) to Alzheimer's disease (AD), time-to-event statistical methods account for varying lengths of follow-up times among subjects whereas two-sample t-tests in voxel-based morphometry (VBM) do not. Our objectives were to apply a time-to-event voxel-based analytic method to identify regions on MRI where atrophy is associated with significantly increased risk of future progression to AD in subjects with MCI and to compare it to traditional voxel-level patterns obtained by applying two-sample methods. We also compared the power required to detect an association using time-to-event methods versus two-sample approaches.

Altered iron metabolism has been hypothesized to be associated with Alzheimer's disease pathology, and prior work has shown associations between iron load and beta amyloid plaques. Quantitative susceptibility mapping (QSM) is a recently popularized MR technique to infer local tissue susceptibility secondary to the presence of iron as well as other minerals. Greater QSM values imply greater iron concentration in tissue. QSM has been used to study relationships between cerebral iron load and established markers of Alzheimer's disease, however relationships remain unclear. In this work we study QSM signal characteristics and associations between susceptibility measured on QSM and established clinical and imaging markers of Alzheimer's disease. The study included 421 participants (234 male, median age 70 years, range 34-97 years) from the Mayo Clinic Study of Aging and Alzheimer's Disease Research Center; 296 (70%) had a diagnosis of cognitively unimpaired, 69 (16%) mild cognitive impairment, and 56 (13%) amnestic dementia. All participants had multi-echo gradient recalled echo imaging, PiB amyloid PET, and Tauvid tau PET. Variance components analysis showed that variation in cortical susceptibility across participants was low. Linear regression models were fit to assess associations with regional susceptibility. Expected increases in susceptibility were found with older age and cognitive impairment in the deep and inferior gray nuclei (pallidum, putamen, substantia nigra, subthalamic nucleus) (betas: 0.0017 to 0.0053 ppm for a 10 year increase in age, p = 0.03 to <0.001; betas: 0.0021 to 0.0058 ppm for a 5 point decrease in Short Test of Mental Status, p = 0.003 to p<0.001). Effect sizes in cortical regions were smaller, and the age associations were generally negative. Higher susceptibility was significantly associated with higher amyloid PET SUVR in the pallidum and putamen (betas: 0.0029 and 0.0012 ppm for a 20% increase in amyloid PET, p = 0.05 and 0.02, respectively), higher tau PET in the basal ganglia with the largest effect size in the pallidum (0.0082 ppm for a 20% increase in tau PET, p<0.001), and with lower cortical gray matter volume in the medial temporal lobe (0.0006 ppm for a 20% decrease in volume, p = 0.03). Overall, these findings suggest that susceptibility in the deep and inferior gray nuclei, particularly the pallidum and putamen, may be a marker of cognitive decline, amyloid deposition, and off-target binding of the tau ligand. Although iron has been demonstrated in amyloid plaques and in association with neurodegeneration, it is of insufficient quantity to be reliably detected in the cortex using this implementation of QSM.

It is now widely known that research brain MRI, CT, and PET images may potentially be re-identified using face recognition, and this potential can be reduced by applying face-deidentification ("de-facing") software. However, for research MRI sequences beyond T1-weighted (T1-w) and T2-FLAIR structural images, the potential for re-identification and quantitative effects of de-facing are both unknown, and the effects of de-facing T2-FLAIR are also unknown. In this work we examine these questions (where applicable) for T1-w, T2-w, T2*-w, T2-FLAIR, diffusion MRI (dMRI), functional MRI (fMRI), and arterial spin labelling (ASL) sequences. Among current-generation, vendor-product research-grade sequences, we found that 3D T1-w, T2-w, and T2-FLAIR were highly re-identifiable (96-98%). 2D T2-FLAIR and 3D multi-echo GRE (ME-GRE) were also moderately re-identifiable (44-45%), and our derived T2* from ME-GRE (comparable to a typical 2D T2*) matched at only 10%. Finally, diffusion, functional and ASL images were each minimally re-identifiable (0-8%). Applying de-facing with mri_reface version 0.3 reduced successful re-identification to ≤8%, while differential effects on popular quantitative pipelines for cortical volumes and thickness, white matter hyperintensities (WMH), and quantitative susceptibility mapping (QSM) measurements were all either comparable with or smaller than scan-rescan estimates. Consequently, high-quality de-facing software can greatly reduce the risk of re-identification for identifiable MRI sequences with only negligible effects on automated intracranial measurements. The current-generation echo-planar and spiral sequences (dMRI, fMRI, and ASL) each had minimal match rates, suggesting that they have a low risk of re-identification and can be shared without de-facing, but this conclusion should be re-evaluated if they are acquired without fat suppression, with a full-face scan coverage, or if newer developments reduce the current levels of artifacts and distortion around the face.