Axonal Injury Partially Mediates Associations Between Increased Left Ventricular Mass Index and White Matter Damage.

BACKGROUND AND PURPOSE: Left ventricular (LV) mass index is a marker of subclinical LV remodeling that relates to white matter damage in aging, but molecular pathways underlying this association are unknown. This study assessed if LV mass index related to cerebrospinal fluid (CSF) biomarkers of microglial activation (sTREM2 [soluble triggering receptor expressed on myeloid cells 2]), axonal injury (NFL [neurofilament light]), neurodegeneration (total-tau), and amyloid-ß, and whether these biomarkers partially accounted for associations between increased LV mass index and white matter damage. We hypothesized higher LV mass index would relate to greater CSF biomarker levels, and these pathologies would partially mediate associations with cerebral white matter microstructure. METHODS: Vanderbilt Memory and Aging Project participants who underwent cardiac magnetic resonance, lumbar puncture, and diffusion tensor imaging (n=142, 72±6 years, 37% mild cognitive impairment [MCI], 32% APOE-ε4 positive, LV mass index 51.4±8.1 g/m2, NFL 1070±588 pg/mL) were included. Linear regressions and voxel-wise analyses related LV mass index to each biomarker and diffusion tensor imaging metrics, respectively. Follow-up models assessed interactions with MCI and APOE-ε4. In models where LV mass index significantly related to a biomarker and white matter microstructure, we assessed if the biomarker mediated white matter associations. RESULTS: Among all participants, LV mass index was unrelated to CSF biomarkers (P>0.33). LV mass index interacted with MCI (P=0.01), such that higher LV mass index related to increased NFL among MCI participants. Associations were also present among APOE-ε4 carriers (P=0.02). NFL partially mediated up to 13% of the effect of increased LV mass index on white matter damage. CONCLUSIONS: Subclinical cardiovascular remodeling, measured as an increase in LV mass index, is associated with neuroaxonal degeneration among individuals with MCI and APOE-ε4. Neuroaxonal degeneration partially reflects associations between higher LV mass index and white matter damage. Findings highlight neuroaxonal degeneration, rather than amyloidosis or microglia, may be more relevant in pathways between structural cardiovascular remodeling and white matter damage.

Elevated Aortic Pulse Wave Velocity Relates to Longitudinal Gray and White Matter Changes.

OBJECTIVE: To determine whether baseline aortic stiffness, measured by aortic pulse wave velocity (PWV), relates to longitudinal cerebral gray or white matter changes among older adults. Baseline cardiac magnetic resonance imaging will be used to assess aortic PWV while brain magnetic resonance imaging will be used to assess gray matter and white matter hyperintensity (WMH) volumes at baseline, 18 months, 3 years, 5 years, and 7 years. Approach and Results: Aortic PWV (m/s) was quantified from cardiac magnetic resonance. Multimodal 3T brain magnetic resonance imaging included T1-weighted imaging for quantifying gray matter volumes and T2-weighted fluid-attenuated inversion recovery imaging for quantifying WMHs. Mixed-effects regression models related baseline aortic PWV to longitudinal gray matter volumes (total, frontal, parietal, temporal, occipital, hippocampal, and inferior lateral ventricle) and WMH volumes (total, frontal, parietal, temporal, and occipital) adjusting for age, sex, race/ethnicity, education, cognitive diagnosis, Framingham stroke risk profile, APOE (apolipoprotein E)-ε4 carrier status, and intracranial volume. Two hundred seventy-eight participants (73±7 years, 58% male, 87% self-identified as non-Hispanic White, 159 with normal cognition, and 119 with mild cognitive impairment) from the Vanderbilt Memory & Aging Project (n=335) were followed on average for 4.9±1.6 years with PWV measurements occurring from September 2012 to November 2014 and longitudinal brain magnetic resonance imaging measurements occurring from September 2012 to June 2021. Higher baseline aortic PWV was related to greater decrease in hippocampal (ß=-3.6 [mm3/y]/[m/s]; [95% CI, -7.2 to -0.02] P=0.049) and occipital lobe (ß=-34.2 [mm3/y]/[m/s]; [95% CI, -67.8 to -0.55] P=0.046) gray matter volume over time. Higher baseline aortic PWV was related to greater increase in WMH volume over time in the temporal lobe (ß=17.0 [mm3/y]/[m/s]; [95% CI, 7.2-26.9] P<0.001). All associations may be driven by outliers. CONCLUSIONS: In older adults, higher baseline aortic PWV related to greater decrease in gray matter volume and greater increase in WMHs over time. Because of unmet cerebral metabolic demands and microvascular remodeling, arterial stiffening may preferentially affect certain highly active brain regions like the temporal lobes. These same regions are affected early in the course of Alzheimer disease.

Cerebrospinal fluid biomarkers of neurodegeneration, synaptic dysfunction, and axonal injury relate to atrophy in structural brain regions specific to Alzheimer's disease.

INTRODUCTION: Patterns of atrophy can distinguish normal cognition from Alzheimer's disease (AD), but neuropathological drivers of this pattern are unknown. This study examined associations between cerebrospinal fluid biomarkers of AD pathology, synaptic dysfunction, and neuroaxonal injury with two AD imaging signatures. METHODS: Signatures were calculated using published guidelines. Linear regressions related each biomarker to both signatures, adjusting for demographic factors. Bootstrapped analyses tested if associations were stronger with one signature versus the other. RESULTS: Increased phosphorylated tau (p-tau), total tau, and neurofilament light (P-values <.045) related to smaller signatures (indicating greater atrophy). Diagnosis and sex modified associations between p-tau and neurogranin (P-values<.05) and signatures, such that associations were stronger among participants with mild cognitive impairment and female participants. The strength of associations did not differ between signatures. DISCUSSION: Increased evidence of neurodegeneration, axonopathy, and tau phosphorylation relate to greater AD-related atrophy. Tau phosphorylation and synaptic dysfunction may be more prominent in AD-affected regions in females.

Higher Aortic Stiffness Is Related to Lower Cerebral Blood Flow and Preserved Cerebrovascular Reactivity in Older Adults.

BACKGROUND: Mechanisms underlying the association between age-related arterial stiffening and poor brain health remain elusive. Cerebral blood flow (CBF) homeostasis may be implicated. This study evaluates how aortic stiffening relates to resting CBF and cerebrovascular reactivity (CVR) in older adults. METHODS: Vanderbilt Memory & Aging Project participants free of clinical dementia, stroke, and heart failure were studied, including older adults with normal cognition (n=155; age, 72±7 years; 59% male) or mild cognitive impairment (n=115; age, 73±7 years; 57% male). Aortic pulse wave velocity (PWV; meters per second) was quantified from cardiac magnetic resonance. Resting CBF (milliliters per 100 g per minute) and CVR (CBF response to hypercapnic normoxia stimulus) were quantified from pseudocontinuous arterial spin labeling magnetic resonance imaging. Linear regression models related aortic PWV to regional CBF, adjusting for age, race/ethnicity, education, Framingham Stroke Risk Profile (diabetes mellitus, smoking, left ventricular hypertrophy, prevalent cardiovascular disease, atrial fibrillation), hypertension, body mass index, apolipoprotein E4 ( APOE ε4) status, and regional tissue volume. Models were repeated testing PWV× APOE ε4 interactions. Sensitivity analyses excluded participants with prevalent cardiovascular disease and atrial fibrillation. RESULTS: Among participants with normal cognition, higher aortic PWV related to lower frontal lobe CBF (ß=-0.43; P=0.04) and higher CVR in the whole brain (ß=0.11; P=0.02), frontal lobes (ß=0.12; P<0.05), temporal lobes (ß=0.11; P=0.02), and occipital lobes (ß=0.14; P=0.01). Among APOE ε4 carriers with normal cognition, findings were more pronounced with higher PWV relating to lower whole-brain CBF (ß=-1.16; P=0.047), lower temporal lobe CBF (ß=-1.81; P=0.004), and higher temporal lobe CVR (ß=0.26; P=0.08), although the last result did not meet the a priori significance threshold. Results were similar in sensitivity models. Among participants with mild cognitive impairment, higher aortic PWV related to lower CBF in the occipital lobe (ß=-0.70; P=0.02), but this finding was attenuated when participants with prevalent cardiovascular disease and atrial fibrillation were excluded. Among APOE ε4 carriers with mild cognitive impairment, findings were more pronounced with higher PWV relating to lower temporal lobe CBF (ß=-1.20; P=0.02). CONCLUSIONS: Greater aortic stiffening relates to lower regional CBF and higher CVR in cognitively normal older adults, especially among individuals with increased genetic predisposition for Alzheimer's disease. Central arterial stiffening may contribute to reductions in regional CBF despite preserved cerebrovascular reserve capacity.

A deep neural network estimation of brain age is sensitive to cognitive impairment and decline.

The greatest known risk factor for Alzheimer's disease (AD) is age. While both normal aging and AD pathology involve structural changes in the brain, their trajectories of atrophy are not the same. Recent developments in artificial intelligence have encouraged studies to leverage neuroimaging-derived measures and deep learning approaches to predict brain age, which has shown promise as a sensitive biomarker in diagnosing and monitoring AD. However, prior efforts primarily involved structural magnetic resonance imaging and conventional diffusion MRI (dMRI) metrics without accounting for partial volume effects. To address this issue, we post-processed our dMRI scans with an advanced free-water (FW) correction technique to compute distinct FW-corrected fractional anisotropy (FAFWcorr) and FW maps that allow for the separation of tissue from fluid in a scan. We built 3 densely connected neural networks from FW-corrected dMRI, T1-weighted MRI, and combined FW+T1 features, respectively, to predict brain age. We then investigated the relationship of actual age and predicted brain ages with cognition. We found that all models accurately predicted actual age in cognitively unimpaired (CU) controls (FW: r=0.66, p=1.62x10-32; T1: r=0.61, p=1.45x10-26, FW+T1: r=0.77, p=6.48x10-50) and distinguished between CU and mild cognitive impairment participants (FW: p=0.006; T1: p=0.048; FW+T1: p=0.003), with FW+T1-derived age showing best performance. Additionally, all predicted brain age models were significantly associated with cross-sectional cognition (memory, FW: ß=-1.094, p=6.32x10-7; T1: ß=-1.331, p=6.52x10-7; FW+T1: ß=-1.476, p=2.53x10-10; executive function, FW: ß=-1.276, p=1.46x10-9; T1: ß=-1.337, p=2.52x10-7; FW+T1: ß=-1.850, p=3.85x10-17) and longitudinal cognition (memory, FW: ß=-0.091, p=4.62x10-11; T1: ß=-0.097, p=1.40x10-8; FW+T1: ß=-0.101, p=1.35x10-11; executive function, FW: ß=-0.125, p=1.20x10-10; T1: ß=-0.163, p=4.25x10-12; FW+T1: ß=-0.158, p=1.65x10-14). Our findings provide evidence that both T1-weighted MRI and dMRI measures improve brain age prediction and support predicted brain age as a sensitive biomarker of cognition and cognitive decline.

MidRISH: Unbiased harmonization of rotationally invariant harmonics of the diffusion signal.

Data harmonization is necessary for removing confounding effects in multi-site diffusion image analysis. One such harmonization method, LinearRISH, scales rotationally invariant spherical harmonic (RISH) features from one site ("target") to the second ("reference") to reduce confounding scanner effects. However, reference and target site designations are not arbitrary and resultant diffusion metrics (fractional anisotropy, mean diffusivity) are biased by this choice. In this work we propose MidRISH: rather than scaling reference RISH features to target RISH features, we project both sites to a mid-space. We validate MidRISH with the following experiments: harmonizing scanner differences from 37 matched patients free of cognitive impairment, and harmonizing acquisition and study differences on 117 matched patients free of cognitive impairment. We find that MidRISH reduces bias of reference selection while preserving harmonization efficacy of LinearRISH. Users should be cautious when performing LinearRISH harmonization. To select a reference site is to choose diffusion metric effect-size. Our proposed method eliminates the bias-inducing site selection step.

Perivascular space burden interacts with APOE-ε4 status on cognition in older adults.

Enlarged perivascular spaces (ePVS) may adversely affect cognition. Little is known about how basal ganglia ePVS interact with apolipoprotein (APOE)-ε4 status. Vanderbilt Memory and Aging Project participants (n = 326, 73 ± 7, 59% male) underwent 3 T brain MRI at baseline to assess ePVS and longitudinal neuropsychological assessments. The interaction between ePVS volume and APOE-ε4 carrier status was related to baseline outcomes using ordinary least squares regressions and longitudinal cognition using linear mixed-effects regressions. ePVS volume interacted with APOE-ε4 status on cross-sectional naming performance (ß = -0.002, p = 0.002), and executive function excluding outliers (ß = 0.001, p = 0.009). There were no significant longitudinal interactions (p-values>0.10) except for Coding excluding outliers (ß = 0.002, p = 0.05). While cross-sectional models stratified by APOE-ε4 status indicated greater ePVS related to worse cognition mostly in APOE-ε4 carriers, longitudinal models stratified by APOE-ε4 status showed greater ePVS volume related to worse cognition among APOE-ε4 non-carriers only. Results indicated that greater ePVS volume interacts with APOE-ε4 status on cognition cross-sectionally. Longitudinally, the association of greater ePVS volume and worse cognition appears stronger in APOE-ε4 non-carriers, possibly due to the deleterious effects of APOE-ε4 on cognition across the lifespan.

Variants in the MS4A cluster interact with soluble TREM2 expression on biomarkers of neuropathology.

Recent evidence suggests that Alzheimer's disease (AD) genetic risk variants (rs1582763 and rs6591561) of the MS4A locus are genome-wide significant regulators of soluble TREM2 levels such that the minor allele of the protective variant (rs1582763) is associated with higher sTREM2 and lower AD risk while the minor allele of (rs6591561) relates to lower sTREM2 and higher AD risk. Our group previously found that higher sTREM2 relates to higher Aß40, worse blood-brain barrier (BBB) integrity (measured with the CSF/plasma albumin ratio), and higher CSF tau, suggesting strong associations with amyloid abundance and both BBB and neurodegeneration complicate interpretation. We expand on this work by leveraging these common variants as genetic tools to tune the interpretation of high CSF sTREM2, and by exploring the potential modifying role of these variants on the well-established associations between CSF sTREM2 as well as TREM2 transcript levels in the brain with AD neuropathology. Biomarker analyses leveraged data from the Vanderbilt Memory & Aging Project (n = 127, age = 72 ± 6.43) and were replicated in the Alzheimer's Disease Neuroimaging Initiative (n = 399, age = 73 ± 7.39). Autopsy analyses were performed leveraging data from the Religious Orders Study and Rush Memory and Aging Project (n = 577, age = 89 ± 6.46). We found that the protective variant rs1582763 attenuated the association between CSF sTREM2 and Aß40 (ß = -0.44, p-value = 0.017) and replicated this interaction in ADNI (ß = -0.27, p = 0.017). We did not observe this same interaction effect between TREM2 mRNA levels and Aß peptides in brain (Aß total ß = -0.14, p = 0.629; Aß1-38, ß = 0.11, p = 0.200). In contrast to the effects on Aß, the minor allele of this same variant seemed to enhance the association with blood-brain barrier dysfunction (ß = 7.0e-4, p = 0.009), suggesting that elevated sTREM2 may carry a much different interpretation in carriers vs. non-carriers of this allele. When evaluating the risk variant (rs6591561) across datasets, we did not observe a statistically significant interaction against any outcome in VMAP and observed opposing directions of associations in ADNI and ROS/MAP on Aß levels. Together, our results suggest that the protective effect of rs1582763 may act by decoupling the associations between sTREM2 and amyloid abundance, providing important mechanistic insight into sTREM2 changes and highlighting the need to incorporate genetic context into the analysis of sTREM2 levels, particularly if leveraged as a clinical biomarker of disease in the future.

DeepN4: Learning N4ITK Bias Field Correction for T1-weighted Images.

T1-weighted (T1w) MRI has low frequency intensity artifacts due to magnetic field inhomogeneities. Removal of these biases in T1w MRI images is a critical preprocessing step to ensure spatially consistent image interpretation. N4ITK bias field correction, the current state-of-the-art, is implemented in such a way that makes it difficult to port between different pipelines and workflows, thus making it hard to reimplement and reproduce results across local, cloud, and edge platforms. Moreover, N4ITK is opaque to optimization before and after its application, meaning that methodological development must work around the inhomogeneity correction step. Given the importance of bias fields correction in structural preprocessing and flexible implementation, we pursue a deep learning approximation / reinterpretation of the N4ITK bias fields correction to create a method which is portable, flexible, and fully differentiable. In this paper, we trained a deep learning network "DeepN4" on eight independent cohorts from 72 different scanners and age ranges with N4ITK-corrected T1w MRI and bias field for supervision in log space. We found that we can closely approximate N4ITK bias fields correction with naïve networks. We evaluate the peak signal to noise ratio (PSNR) in test dataset against the N4ITK corrected images. The median PSNR of corrected images between N4ITK and DeepN4 was 47.96 dB. In addition, we assess the DeepN4 model on eight additional external datasets and show the generalizability of the approach. This study establishes that incompatible N4ITK preprocessing steps can be closely approximated by naïve deep neural networks, facilitating more flexibility. All code and models are released at https://github.com/MASILab/DeepN4 .

Empirical assessment of the assumptions of ComBat with diffusion tensor imaging.

Purpose: Diffusion tensor imaging (DTI) is a magnetic resonance imaging technique that provides unique information about white matter microstructure in the brain but is susceptible to confounding effects introduced by scanner or acquisition differences. ComBat is a leading approach for addressing these site biases. However, despite its frequent use for harmonization, ComBat's robustness toward site dissimilarities and overall cohort size have not yet been evaluated in terms of DTI. Approach: As a baseline, we match N=358 participants from two sites to create a "silver standard" that simulates a cohort for multi-site harmonization. Across sites, we harmonize mean fractional anisotropy and mean diffusivity, calculated using participant DTI data, for the regions of interest defined by the JHU EVE-Type III atlas. We bootstrap 10 iterations at 19 levels of total sample size, 10 levels of sample size imbalance between sites, and 6 levels of mean age difference between sites to quantify (i) ßAGE, the linear regression coefficient of the relationship between FA and age; (ii) Î³/f*, the ComBat-estimated site-shift; and (iii) Î´/f*, the ComBat-estimated site-scaling. We characterize the reliability of ComBat by evaluating the root mean squared error in these three metrics and examine if there is a correlation between the reliability of ComBat and a violation of assumptions. Results: ComBat remains well behaved for ßAGE when N>162 and when the mean age difference is less than 4 years. The assumptions of the ComBat model regarding the normality of residual distributions are not violated as the model becomes unstable. Conclusion: Prior to harmonization of DTI data with ComBat, the input cohort should be examined for size and covariate distributions of each site. Direct assessment of residual distributions is less informative on stability than bootstrap analysis. We caution use ComBat of in situations that do not conform to the above thresholds.

Evaluation of Mean Shift, ComBat, and CycleGAN for Harmonizing Brain Connectivity Matrices Across Sites.

Connectivity matrices derived from diffusion MRI (dMRI) provide an interpretable and generalizable way of understanding the human brain connectome. However, dMRI suffers from inter-site and between-scanner variation, which impedes analysis across datasets to improve robustness and reproducibility of results. To evaluate different harmonization approaches on connectivity matrices, we compared graph measures derived from these matrices before and after applying three harmonization techniques: mean shift, ComBat, and CycleGAN. The sample comprises 168 age-matched, sex-matched normal subjects from two studies: the Vanderbilt Memory and Aging Project (VMAP) and the Biomarkers of Cognitive Decline Among Normal Individuals (BIOCARD). First, we plotted the graph measures and used coefficient of variation (CoV) and the Mann-Whitney U test to evaluate different methods' effectiveness in removing site effects on the matrices and the derived graph measures. ComBat effectively eliminated site effects for global efficiency and modularity and outperformed the other two methods. However, all methods exhibited poor performance when harmonizing average betweenness centrality. Second, we tested whether our harmonization methods preserved correlations between age and graph measures. All methods except for CycleGAN in one direction improved correlations between age and global efficiency and between age and modularity from insignificant to significant with p-values less than 0.05.

Sex, racial, and APOE-ε4 allele differences in longitudinal white matter microstructure in multiple cohorts of aging and Alzheimer's disease.

INTRODUCTION: The effects of sex, race, and Apolipoprotein E (APOE) - Alzheimer's disease (AD) risk factors - on white matter integrity are not well characterized. METHODS: Diffusion MRI data from nine well-established longitudinal cohorts of aging were free-water (FW)-corrected and harmonized. This dataset included 4,702 participants (age=73.06 ± 9.75) with 9,671 imaging sessions over time. FW and FW-corrected fractional anisotropy (FAFWcorr) were used to assess differences in white matter microstructure by sex, race, and APOE-ε4 carrier status. RESULTS: Sex differences in FAFWcorr in association and projection tracts, racial differences in FAFWcorr in projection tracts, and APOE-ε4 differences in FW limbic and occipital transcallosal tracts were most pronounced. DISCUSSION: There are prominent differences in white matter microstructure by sex, race, and APOE-ε4 carrier status. This work adds to our understanding of disparities in AD. Additional work to understand the etiology of these differences is warranted.

Multiphasic modification of intrinsic functional connectivity of the rat brain during increasing levels of propofol.

The dose-dependent effects of anesthetics on brain functional connectivity are incompletely understood. Resting-state functional magnetic resonance imaging (rsfMRI) is widely used to assess the functional connectivity in humans and animals. Propofol is an anesthetic agent with desirable characteristics for functional neuroimaging in animals but its dose-dependent effects on rsfMRI functional connectivity have not been determined. Here we tested the hypothesis that brain functional connectivity undergoes specific changes in distinct neural networks at anesthetic depths associated with loss of consciousness. We acquired spontaneous blood oxygen level-dependent (BOLD) signals simultaneously with electroencephalographic (EEG) signals from rats under steady-state, intravenously administered propofol at increasing doses from light sedation to deep anesthesia (20, 40, 60, 80, and 100 mg/kg/h IV). Power spectra and burst suppression ratio were calculated from the EEG to verify anesthetic depth. Functional connectivity was determined from the whole brain correlation of BOLD data in regions of interest followed by a segmentation of the correlation maps into anatomically defined regional connectivity. We found that propofol produced multiphasic, dose dependent changes in functional connectivity of various cortical and subcortical networks. Cluster analysis predicted segregation of connectivity into two cortical and two subcortical clusters. In one cortical cluster (somatosensory and parietal), the early reduction in connectivity was followed by transient reversal; in the other cluster (sensory, motor and cingulate/retrosplenial), this rebound was absent. The connectivity of the subcortical cluster (brainstem, hippocampal and caudate) was strongly reduced, whereas that of another (hypothalamus, medial thalamus and n. basalis) did not. Subcortical connectivity increased again in deep anesthesia associated with EEG burst suppression. Regional correlation analysis confirmed the breakdown of connectivity within and between specific cortical and subcortical networks with deepening propofol anesthesia. Cortical connectivity was suppressed before subcortical connectivity at a critical propofol dose associated with loss of consciousness.

A deep neural network estimation of brain age is sensitive to cognitive impairment and decline.

The greatest known risk factor for Alzheimer's disease (AD) is age. While both normal aging and AD pathology involve structural changes in the brain, their trajectories of atrophy are not the same. Recent developments in artificial intelligence have encouraged studies to leverage neuroimaging-derived measures and deep learning approaches to predict brain age, which has shown promise as a sensitive biomarker in diagnosing and monitoring AD. However, prior efforts primarily involved structural magnetic resonance imaging and conventional diffusion MRI (dMRI) metrics without accounting for partial volume effects. To address this issue, we post-processed our dMRI scans with an advanced free-water (FW) correction technique to compute distinct FW-corrected fractional anisotropy (FAFWcorr) and FW maps that allow for the separation of tissue from fluid in a scan. We built 3 densely connected neural networks from FW-corrected dMRI, T1-weighted MRI, and combined FW+T1 features, respectively, to predict brain age. We then investigated the relationship of actual age and predicted brain ages with cognition. We found that all models accurately predicted actual age in cognitively unimpaired (CU) controls (FW: r=0.66, p=1.62×10-32; T1: r=0.61, p=1.45×10-26, FW+T1: r=0.77, p=6.48×10-50) and distinguished between CU and mild cognitive impairment participants (FW: p=0.006; T1: p=0.048; FW+T1: p=0.003), with FW+T1-derived age showing best performance. Additionally, all predicted brain age models were significantly associated with cross-sectional cognition (memory, FW: ß=-1.094, p=6.32×10-7; T1: ß=-1.331, p=6.52×10-7; FW+T1: ß=-1.476, p=2.53×10-10; executive function, FW: ß=-1.276, p=1.46×10-9; T1: ß=-1.337, p=2.52×10-7; FW+T1: ß=-1.850, p=3.85×10-17) and longitudinal cognition (memory, FW: ß=-0.091, p=4.62×10-11; T1: ß=-0.097, p=1.40×10-8; FW+T1: ß=-0.101, p=1.35×10-11; executive function, FW: ß=-0.125, p=1.20×10-10; T1: ß=-0.163, p=4.25×10-12; FW+T1: ß=-0.158, p=1.65×10-14). Our findings provide evidence that both T1-weighted MRI and dMRI measures improve brain age prediction and support predicted brain age as a sensitive biomarker of cognition and cognitive decline.

MidRISH: Unbiased harmonization of rotationally invariant harmonics of the diffusion signal.

Objective: Data harmonization is necessary for removing confounding effects in multi-site diffusion image analysis. One such harmonization method, LinearRISH, scales rotationally invariant spherical harmonic (RISH) features from one site ("target") to the second ("reference") to reduce confounding scanner effects. However, reference and target site designations are not arbitrary and resultant diffusion metrics (fractional anisotropy, mean diffusivity) are biased by this choice. In this work we propose MidRISH: rather than scaling reference RISH features to target RISH features, we project both sites to a mid-space. Methods: We validate MidRISH with the following experiments: harmonizing scanner differences from 37 matched patients free of cognitive impairment, and harmonizing acquisition and study differences on 117 matched patients free of cognitive impairment. Conclusion: MidRISH reduces bias of reference selection while preserving harmonization efficacy of LinearRISH. Significance: Users should be cautious when performing LinearRISH harmonization. To select a reference site is to choose diffusion metric effect-size. Our proposed method eliminates the bias-inducing site selection step.

Enlarged perivascular space burden associations with arterial stiffness and cognition.

Enlarged perivascular spaces (ePVS) are difficult to quantify, and their etiologies and consequences are poorly understood. Vanderbilt Memory and Aging Project participants (n = 327, 73 ± 7 years) completed 3T brain MRI to quantify ePVS volume and count, longitudinal neuropsychological assessment, and cardiac MRI to quantify aortic stiffness. Linear regressions related (1) PWV to ePVS burden and (2) ePVS burden to cross-sectional and longitudinal neuropsychological performance adjusting for key demographic and medical factors. Higher aortic stiffness related to greater basal ganglia ePVS volume (ß = 7.0×10-5, p = 0.04). Higher baseline ePVS volume was associated with worse baseline information processing (ß = -974, p = 0.003), executive function (ß = -81.9, p < 0.001), and visuospatial performances (ß = -192, p = 0.02) and worse longitudinal language (ß = -54.9, p = 0.05), information processing (ß = -147, p = 0.03), executive function (ß = -10.9, p = 0.03), and episodic memory performances (ß = -10.6, p = 0.02). Results were similar for ePVS count. Greater arterial stiffness relates to worse basal ganglia ePVS burden, suggesting cardiovascular aging as an etiology. ePVS burden is associated with adverse cognitive trajectory, emphasizing the clinical relevance of ePVS.

Comparing voxel- and feature-wise harmonization of complex graph measures from multiple sites for structural brain network investigation of aging.

Complex graph theory measures of brain structural connectomes derived from diffusion weighted images (DWI) provide insight into the network structure of the brain. Further, as the number of available DWI datasets grows, so does the ability to investigate associations in these measures with major biological factors, like age. However, one key hurdle that remains is the presence of scanner effects that can arise from different DWI datasets and confound multisite analyses. Two common approaches to correct these effects are voxel-wise and feature-wise harmonization. However, it is still unclear how to best leverage them for graph-theory analysis of an aging population. Thus, there is a need to better characterize the impact of each harmonization method and their ability to preserve age related features. We investigate this by characterizing four complex graph theory measures (modularity, characteristic path length, global efficiency, and betweenness centrality) in 48 participants aged 55 to 86 from Baltimore Longitudinal Study of Aging (BLSA) and Vanderbilt Memory and Aging Project (VMAP) before and after voxel- and feature-wise harmonization with the Null Space Deep Network (NSDN) and ComBat, respectively. First, we characterize across dataset coefficients of variation (CoV) and find the combination of NSDN and ComBat causes the greatest reduction in CoV followed by ComBat alone then NSDN alone. Second, we reproduce published associations of modularity with age after correcting for other covariates with linear models. We find that harmonization with ComBat or ComBat and NSDN together improves the significance of existing age effects, reduces model residuals, and qualitatively reduces separation between datasets. These results reinforce the efficiency of statistical harmonization on the feature-level with ComBat and suggest that harmonization on the voxel-level is synergistic but may have reduced effect after running through the multiple layers of the connectomics pipeline. Thus, we conclude that feature-wise harmonization improves statistical results, but the addition of biologically informed voxel-based harmonization offers further improvement.

Baseline grey matter volumes and white matter hyperintensities predict decline in functional activities in older adults over a 5-year follow-up period.

INTRODUCTION: Functional independence is an essential predictor of quality of life in aging, yet few accessible predictors of functional decline have been identified. This study examined associations between baseline structural neuroimaging markers and longitudinal functional status. METHODS: Linear mixed effects models with follow-up time interaction terms related baseline grey matter volume and white matter hyperintensities (WMHs) to functional trajectory, adjusting for demographic and medical covariates. Subsequent models assessed interactions with cognitive status and apolipoprotein E (APOE) ε4 status. RESULTS: Smaller baseline grey matter volumes, particularly in regions commonly affected by Alzheimer's disease (AD), and greater baseline WMHs were associated with faster functional decline over a mean 5-year follow-up. Effects were stronger in APOE-ε4 carriers on grey matter variables. Cognitive status interacted with most MRI variables. DISCUSSION: Greater atrophy in AD-related regions and higher WMH burden at study entry were associated with faster functional decline, particularly among participants at increased risk of AD.

Leveraging longitudinal diffusion MRI data to quantify differences in white matter microstructural decline in normal and abnormal aging.

Introduction: It is unclear how rates of white matter microstructural decline differ between normal aging and abnormal aging. Methods: Diffusion MRI data from several well-established longitudinal cohorts of aging (Alzheimer's Disease Neuroimaging Initiative [ADNI], Baltimore Longitudinal Study of Aging [BLSA], Vanderbilt Memory & Aging Project [VMAP]) were free-water corrected and harmonized. This dataset included 1723 participants (age at baseline: 72.8 ± 8.87 years, 49.5% male) and 4605 imaging sessions (follow-up time: 2.97 ± 2.09 years, follow-up range: 1-13 years, mean number of visits: 4.42 ± 1.98). Differences in white matter microstructural decline in normal and abnormal agers was assessed. Results: While we found a global decline in white matter in normal/abnormal aging, we found that several white matter tracts (e.g., cingulum bundle) were vulnerable to abnormal aging. Conclusions: There is a prevalent role of white matter microstructural decline in aging, and future large-scale studies in this area may further refine our understanding of the underlying neurodegenerative processes. HIGHLIGHTS: Longitudinal data were free-water corrected and harmonized.Global effects of white matter decline were seen in normal and abnormal aging.The free-water metric was most vulnerable to abnormal aging.Cingulum free-water was the most vulnerable to abnormal aging.

DeepN4: Learning N4ITK Bias Field Correction for T1-weighted Images.

T1-weighted (T1w) MRI has low frequency intensity artifacts due to magnetic field inhomogeneities. Removal of these biases in T1w MRI images is a critical preprocessing step to ensure spatially consistent image interpretation. N4ITK bias field correction, the current state-of-the-art, is implemented in such a way that makes it difficult to port between different pipelines and workflows, thus making it hard to reimplement and reproduce results across local, cloud, and edge platforms. Moreover, N4ITK is opaque to optimization before and after its application, meaning that methodological development must work around the inhomogeneity correction step. Given the importance of bias fields correction in structural preprocessing and flexible implementation, we pursue a deep learning approximation / reinterpretation of the N4ITK bias fields correction to create a method which is portable, flexible, and fully differentiable. In this paper, we trained a deep learning network "DeepN4" on eight independent cohorts from 72 different scanners and age ranges with N4ITK-corrected T1w MRI and bias field for supervision in log space. We found that we can closely approximate N4ITK bias fields correction with naïve networks. We evaluate the peak signal to noise ratio (PSNR) in test dataset against the N4ITK corrected images. The median PSNR of corrected images between N4ITK and DeepN4 was 47.96 dB. In addition, we assess the DeepN4 model on eight additional external datasets and show the generalizability of the approach. This study establishes that incompatible N4ITK preprocessing steps can be closely approximated by naïve deep neural networks, facilitating more flexibility. All code and models are released at https://github.com/MASILab/DeepN4.