Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Jul 8.
Published in final edited form as: Neuroimage. 2024 Aug 22;298:120794. doi: 10.1016/j.neuroimage.2024.120794

Quantitative gradient recalled echo (qGRE) MRI enables in vivo measurement of pre-atrophic neurodegeneration in a mouse model of Alzheimer’s disease

Michal R Tomaszewski a,*, Alexander L Sukstanskii b, Hyking Haley a, Xiangjun Meng a, Corin O Miller a, Dmitriy A Yablonskiy b
PMCID: PMC12237143  NIHMSID: NIHMS2092897  PMID: 39173693

Abstract

Robust methods are needed for preclinical evaluation of novel Alzheimer Disease (AD) therapies to accelerate drug discovery. Quantitative Gradient Recalled Echo (qGRE) MRI has shown promise to provide insight into neurodegeneration in AD prior to atrophy development in humans, highlighting areas of low neuronal density. In this study a novel qGRE method (20 echoes, TE=2–40ms) is shown to non-invasively measure the longitudinal neuronal loss in the hippocampus of a mouse model of AD tauopathy Tg4510.

Tg4510 (n=10) and wild type (WT, n=6) mice underwent MRI (7T field strength) at 3–7 months old. 3D qGRE approach was used to generate brain-specific R2* maps free of magnetic field inhomogeneity artifacts. Light-sheet microscopy of the brains stained with NeuN and MBP served to visualize neuronal nuclei and myelin content respectively.

Significant decrease in NeuN staining between 3mo and 5mo was observed in the hippocampus of Tg4510, validating the mouse AD model.

Longitudinal analysis showed clear decreases in R2* metric of qGRE signal in the Tg4510 mice hippocampus undergoing neurodegeneration between 3 and 5 months old. Histogram analysis revealed an upward trend in patterns of low R2* value (Dark Matter, DM), and broadening of R2* distribution. These were quantified as significant increase in both DM Volume Fraction (DMVF) and R2* Standard Deviation (SD) in Tg4510 mice (p=0.004/p=0.016 DMVF/SD) but not in WT controls (p>0.25). Further monotonical increase was also observed in both metrics in time. A significant negative correlation was observed between the DMVF and myelin content (p=0.01, r=−0.76), suggesting sensitivity of the technique to the loss of myelinated axons.

The presented qGRE technique, validated by histological measurements, can be readily applied as in vivo tool in preclinical models of neurodegeneration for pharmacodynamics and mechanism of action assessment.

Keywords: MRI, Alzheimer’s disease, Tg4510, Neurodegeneration, Neuronal density, QGRE

1. Introduction

Alzheimer’s disease (AD) is a debilitating illness affecting more than 6.5 million individuals in the USA alone, causing memory loss and cognitive decline, and may lead to death. As the population ages, the prevalence of AD is also projected to grow rapidly. Significant efforts to develop novel therapies rely on efficient and early diagnosis, staging and treatment response evaluation, including in mouse models to support the active drug research in the field.

In addition to efforts of testing anti-amyloid therapy in AD (Zhang et al., 2023), continued development of tau-targeting therapeutics (Congdon et al., 2023) show promise for deceleration of AD-associated neuronal loss, with a potential direct effect on cognition. Better understanding is still needed of the link between tau burden and its change due to treatment, and neurodegeneration. In particular, the temporal characteristics of the relationship will be essential to inform the choice of target population, dosing, and pharmacodynamic assessment to accelerate the anti-tau drug development process.

Recently proposed by National Institute of Aging and Alzheimer Association (NIA-AA) A/T/N (amyloid/ tau/ neurodegeneration) approach (Jack et al., 2016) is intended to classify different stages of Alzheimer disease by means of AD-related tissue pathology rather than clinical manifestations. While brain amyloid plaques and tau neurofibrillary tangles (NFT) can now be measured in vivo using PET tracers, the neurodegeneration is mostly measured in vivo as tissue atrophy through MRI-based morphological studies (Jack et al., 1999, Jack et al., 1992, Jack et al., 2000, Kesslak et al., 1991, Young et al., 2020) that serve as in vivo biomarkers of neuronal loss (Jack et al., 2016). However, histopathological studies demonstrated that the neuronal loss actually exceeds loss of tissue volume (Price et al., 2001). This result is consistent with the finding of a wide spread reduction of synapses that were more extensive than decreases in gray matter volume (Mecca et al., 2020). Importantly, symptoms of Alzheimer disease appear after sufficient neuronal (Price et al., 2001, Gomez-Isla et al., 1996, Kothapalli et al., 2022) and synaptic (Terry et al., 1991, Selkoe, 2002) losses have occurred. While indirect information about the neuronal tissue structure may be gained from 18F-FDG (Minoshima et al., 2022) and SV2A (Carson et al., 2022) PET, these nuclear medicine approaches do not provide specific insight into neurodegeneration and carry known experimental challenges especially in preclinical applications.

Our approach to quantifying neurodegeneration relies on the quantitative Gradient Recalled Echo (qGRE) MRI technique (Ulrich and Yablonskiy, 2016). This approach provides quantitative in vivo high resolution 3D measurements of several brain-tissue-specific relaxation properties (qGRE metrics) of the gradient recalled echo (GRE) MRI signal. qGRE allows separation of commonly used R2* relaxation rate parameter into tissue-specific R2t* metric that depends on the molecular constituents present in the brain (Zhao et al., 2016, Wen et al., 2018, Yablonskiy and Sukstanskii, 2024) and brain hemodynamic properties-specific R2’ metric (Yablonskiy and Haacke, 1994, Yablonskiy, 1998). Hence, these qGRE metrics can serve as surrogate markers of tissue alterations that reflect the integrity of brain cellular structure and disease-related tissue damage.

We have demonstrated that qGRE metrics substantially change with distinct patterns in AD (Kothapalli et al., 2022, Zhao et al., 2017), multiple sclerosis (MS) (Sati et al., 2010, Luo et al., 2014, Luo et al., 2012, Wen et al., 2014, Wen et al., 2015, Patel et al., 2015, Xiang et al., 2019, Xiang et al., 2019, Xiang et al., 2020, Xiang et al., 2020, Levasseur et al., 2022, Xiang et al., 2022), and psychiatric diseases (Mamah et al., 2015). qGRE data obtained from a well-characterized cohort of participants from the Knight Alzheimer Disease Research Center revealed the existence of brain regions with low qGRE R2t* values (Zhao et al., 2017), termed Dark Matter, as they appear dark on R2t* maps, representing tissue essentially devoid of neurons (Kothapalli et al., 2022). These data show that Dark Matter can be identified already in people with preclinical stages of AD (amyloid positive but without clinical symptoms) and has a predictive power of future AD progression. Importantly, Dark Matter represents pre-atrophic neurodegeneration that is not detected on T1W images widely used to measure tissue atrophy; hence, Dark Matter can serve as a biomarker of early preclinical AD.

Preclinical models, including transgenic tau mice are widely utilized to simulate human disease development (Jankowsky and Z., 2017). A prominent example is the Tg4510 mouse, a well-established and well-characterized model of early AD tauopathy, known to show gradual decrease in neuronal density with age (Ramsden et al., 2005). The changes in the temporal development of this pathology can be measured in response to treatment, to characterize potential phenotype rescue and provide a powerful pharmacodynamics readout for novel drug candidates. However, significant heterogeneity observed between animals in the pathology dynamics in this model highlights the need for repeated longitudinal measurements. Overall, a substantial body of literature available for imaging characterization of tauopathy mouse models, highlighting their complexity (Ni, 2022). Prior studies have looked at structure (Sahara et al., 2014), metabolism (Wells et al., 2015), and neuronal function (Degiorgis et al., 2020), showing broad differences to wild type mice, including in the context of magnetic susceptibility contrast (O’Callaghan et al., 2017). However, understanding of the longitudinal development and early neurodegenerative pathology dynamics within the models remain less explored. A significant unmet need exists for development and application of in vivo methods for longitudinal evaluation of changes in brain neuronal density to aid preclinical evaluation of anti-tau compounds and accelerate drug discovery in the field.

In this paper, qGRE is adapted and applied to quantify the longitudinal neuronal loss of Tg4510 mice. Previous cross-correlation studies (Wen et al., 2018) have established relationship between qGRE metrics and brain tissue neuronal structure defined through the genetic information obtained from the Allen Human Brain Atlas. Herein, for the first time, we demonstrate direct relationship between qGRE metrics and histologically defined neuronal loss. We show that qGRE can be used for longitudinal measurements of changes in neuronal density in mouse brain, and hence readily applied to evaluation of treatment response to novel anti-AD drug candidates.

2. Methods

Female Tg4510 and wild type (WT) mice were used in the study (Taconic, Germantown, NY), 3 months of age at the first scan time-point. All protocols for the use of these animals were approved by the Institutional Animal Care and Use Committee of Merck & Co., Inc., Rahway, NJ. Mice were anesthetized using isoflurane (3% induction and 1–2% maintenance to ensure stable respiration rate 40–50 bpm), mixed in 100% O2 to maximize venous blood oxygen content, therefore minimizing BOLD contributions to the R2* MR images (He et al., 2008). This allows using R2* qGRE metric instead of R2t* (R2* = R2t* +R2’, with R2’ proportional to venous blood deoxygenation level (Yablonskiy and Haacke, 1994, Yablonskiy, 1998)). Animal core temperature was maintained at 37C using a circulating water blanket (SA Instruments). The mice underwent MRI at following age time-points: 3mo (10/6 Tg4510/WT), 5mo (8/4 Tg4510/WT), 6mo (4/0 Tg4510/WT) and 7mo (4/0 Tg4510/WT). Some animals were sacrificed at each time-point for ex vivo analysis as described below, one scan at 5mo was excluded due to poor shim. MRI (Bruker, 7T field strength) was performed with a 2 × 2 mouse brain array receive coil and 72mm whole body volume coil for RF transmission. 2nd order map shim was performed in an ellipsoid including the hippocampus and surrounding areas. When linewidth in the region following the shim was measured to exceed 35Hz, the procedure was repeated to achieve more homogeneous field. Balanced Steady State Free Precession (bSSFS/TrueFISP) was acquired for anatomical reference: 90 × 192 × 192 (slice-read-phase) points, TR/TE=3.2/1.6ms, FOV: (16mm)3. qGRE data for R2* mapping were acquired using 3D gradient-recalled echo sequence with the same FOV and 20 gradient-recalled echoes, first TE=2ms, ΔTE=2ms, TR=50ms, FA=10°, matrix size 90 × 105 × 105 (slice-read-phase), 2 averages, 18 minutes scan time, total time in the MRI ~28 minutes per mouse per session.

Details of qGRE data analysis were provided in (Ulrich and Yablonskiy, 2016). In brief, after data acquisition, raw k-space data are read into MATLAB (MathWorks) for post-processing. First, we apply FFT to get images. The multi-channel data are combined using the following algorithm (Luo et al., 2012):

Sn(TE)=∑ch=1Mλch⋅S¯nchTE1⋅Snch(TE);λch=1M⋅εch2∑ch′=1Mεch′2 (1)

where the sum is taken over all M channels (ch), S¯ denotes complex conjugate of S, λch are weighting parameters and εch are noise amplitudes (r.m.s.). Index n corresponds to the voxel position (n=x,y,z). As we demonstrated previously, this algorithm allows for the optimal estimation of quantitative parameters (Luo et al., 2012), and also removes the initial phase incoherence among the channels (Luo et al., 2012). Standard R2* = 1/T2* values are estimated by fitting the following biophysical model to experimental data:

S(TE)=A0⋅exp−R2*⋅TE+i⋅2π⋅Δf⋅TE⋅F(TE) (2)

where TE is the gradient echo time, Δf is the frequency shift (dependent on tissue structure and also macroscopic magnetic field created mostly by tissue/air interfaces), and function F(TE) describes the effects of macroscopic magnetic field inhomogeneities. We use a voxel spread function (VSF) method (Yablonskiy et al., 2013) for calculating F(TE). Regions with especially strong field inhomogeneities (defined as F (TE10)<0.5) experience significant signal losses and were excluded from further analysis.

Hippocampus Regions of Interest (ROIs) were segmented manually in the bSSFS anatomical scan for all mice and time-points. Automated registration and segmentation were not performed given the significant longitudinal deformation of the brain characteristic for the Tg4510 mouse model. These ROIs were transferred into R2* maps. The data analysis is based on R2* histograms. We first define a normal R2* distribution characterizing a physiological tissue variability based on hippocampus R2* distribution in 3mo healthy control mice. We characterize this distribution by its mean value (R2*-mean) and standard deviation (STD). The hippocampal regions with low R2* values

R2*<R2mean*−1.96⋅STD (3)

(i.e. below 95% confidence interval) are considered as abnormal. A single R2* threshold was thus defined and used in quantification of all mice and scans. This is an analog of Dark Matter introduced for human studies (Kothapalli et al., 2022). The difference in these approaches is due to R2* dependence on magnetic field strength (7T in current study vs. 3T in human study). The relevant metrics for all mice – histogram Standard Deviation (STD) and Dark Matter Volume Fraction (DMVF) were quantified based on the threshold. DMVF was defined as a fraction of hippocampus voxels satisfying the criterion in Eq. (3).

Mice were sacrificed at the 3mo (n=2/2 Tg4510/WT) and 5mo (n=4/2 Tg4510/WT) time-points and brains harvested for ex vivo analysis of myelin content (LifeCanvas Technologies, Cambridge, MA). A separate cohort of 3 and 5mo Tg4510 mice (n=3/3) was used for ex vivo validation of the neuronal loss in the model, and did not undergo MRI. The same tissue collection protocol was used for both cohorts. Animals were euthanized by CO2 inhalation, followed by cardiac perfusion with ice cold saline and subsequently with 10% formalin. The brains were harvested, fixed for 24h in 10% formalin and processed according to vendor protocols: preserved with using SHIELD reagents (Park et al., 2019), delipidated and immunolabeled with NeuN (Cell Signaling Technology #24307) and Myelin Basic Protein (EnCor Biotechnology #MCA-7G7) antibodies for visualization of neuronal nuclei and myelin content respectively. After staining and refractive index matching (EasyIndex, LifeCanvas Technologies, RI=1.52) the whole samples were imaged in 3D using a light sheet microscope, affine registration performed to Allen Brain Atlas (https://portal.brain-map.org/), and nuclear density (for NeuN) and mean fluorescent intensity (MBP) quantified in the hippocampus. Additionally, cells are counted in the left and right hippocampus region separately and the count divided by the region volume. Segmentation of the nuclei was performed by LifeCanvas with a well-established analysis pipeline. The cell detection was based on automated intensity thresholding of NeuN staining and size and shape filtering.

Statistical significance of longitudinal changes in MRI metrics was assessed with paired two sided t-test on the animals with repeated scans (Microsoft Excel 365), and the correlation between histological and MR metrics was assessed with Pearson correlation analysis (MATLAB R2021a, MathWorks).

3. Results

Initial validation of the mouse model was performed with ex vivo quantification of neuronal density suing light-sheet microscopy. The study focuses on the hippocampus, given its robust and confirmed longitudinal neurodegeneration in the Tg4510 model, and its relevance in AD. The NeuN staining revealed a significant decrease in neuronal nuclei count in the hippocampus of Tg4510 mice between 3mo and 5mo cohorts. These results indicate a decrease of over 30% in neuronal density (p<0.001, Fig. 1), comparable to previously reported value (Santacruz et al., 2005), confirming our understanding of the pathology development in the mouse model used.

Fig. 1.

Fig. 1.

Open in a new tab

Histological analysis reveals neurodegeneration in Tg4510 mice. Example NeuN staining pattern indicative of neuronal density at 5mo is shown in (A, left) together with anatomical True-FISP MRI. Quantification of mean neuronal densities in hippocampus of Tg4510 mice (B) reveals a decrease between 3 and 5 months of age (n=6 hemispheres from n=3 brains at each time-point). Green lines indicate atlas regions overlaid over the image. Grayscale presentation was chosen to facilitate visual comparison with MRI and emphasize the more binary black/white negative/positive relevance of the cellular staining, especially in the case of NeuN.

Next, R2* value distributions in the Tg4510 hippocampus were compared longitudinally between the 3 and 5 months of age to understand the associated imaging signature of the pathology (Fig. 2). Beyond the anatomical changes as expected in the model, clear changes were observed in the R2* maps of remaining tissue. Increased heterogeneity of R2* in the hippocampus of the older mice was coupled with emergence of pronounced low R2* value areas.

Fig. 2.

Fig. 2.

Open in a new tab

Changes observed longitudinally in R2* maps in Tg4510 mice. Differences in both brain anatomy (TrueFISP scan, left) and R2* relaxation maps (right) are shown for a representative Tg4510 mouse brain between 3 (top) and 5 (bottom) months of age. Black arrows indicate example areas of low R2* which appear in 5mo but not 3mo brain.

R2* histogram from a representative animal illustrates these observations. A pattern of increase in low R2* value incidence and broadening of R2* distribution is revealed (Fig. 3A). This is in line with clinical observations (Kothapalli et al., 2022) of the presence of “Dark Matter” (DM), low R2* regions in qGRE scans in patients at more advanced AD stages. For DM Volume Fraction (DMVF) quantification, R2* mean and standard deviation in hippocampus of healthy 3mo mice were measured to be 22.5 s−1 and 3.1 s−1 respectively. While mean R2* showed a near-significant trend (Fig. 3b), both the DMVF, and the R2* standard deviation (SD), a measure of value heterogeneity, showed significant increase from 3 to 5mo in Tg4510 mice (Fig. 3c, p=0.004/p=0.016 DMVF/SD), but not in WT controls (p=0.25). As expected for this model, significant, but highly variable hippocampus atrophy was also observed (volume decrease 12±3%, p=0.005). Importantly, further monotonical increase was also observed in both DMVF and SD (Fig. 4) at later ages, providing insight into the dynamic range of these measurements and metrics chosen.

Fig. 3.

Fig. 3.

Open in a new tab

Quantification of R2* changes (Paired comparison of n=7). Evolution of the R2* value distribution in a representative Tg4510 hippocampus is shown in (A), reflected in a trend towards lower mean R2* (B) and significantly increased incidence of low R2* values, quantified in (C) as Dark Matter Volume Ratio (DMVF), as well as increased spatial R2* heterogeneity quantified in (D) as standard deviation. Note that bright red color represents overlap of 3 months old and 5 months old histograms.

Fig. 4.

Fig. 4.

Open in a new tab

Temporal evolution of R2* metrics in Tg4510 mice. Monotonic increase past 5 months of age is shown both for the Dark Matter Volume Fraction (A) and R2* standard deviation (B). n=10/7/4/4 at 3/5/6/7 months old. Error bars indicate standard error of the mean.

To better understand the biological basis of the observed R2* metrics changes, further histological analysis for myelin protein (MBP) abundance was performed. Importantly in the pooled analysis of all brains stained, a significant negative correlation was observed between myelin content and the DMVF (Fig. 5, p=0.01, r=−0.76), suggesting that the loss of myelinated axons, and associated breakdown of the layered membranes drives the local decrease in tissue R2*. Spatial comparison between the MRI and MBP images did not yield notable shared patterns, given the low local heterogeneity in hippocampal MBP.

Fig. 5.

Fig. 5.

Open in a new tab

Brain R2* changes are associated with de-myelination. (A) a representative image of Myelin Basic Protein staining in the Tg4510 brain, (B) relationship between hippocampal R2*-defined Dark Matter Volume Fraction (DMVF) and the average intensity of MBP signal show negative correlation (p=0.01, B). Each point represents data from one mouse. Green lines indicate atlas regions overlaid over the image. WT – wild type.

4. Discussion

Pre-atrophic neurodegeneration is an important hallmark of Alzheimer’s Disease, appearing before volumetric brain changes and clinical symptoms (Price et al., 2001, Gomez-Isla et al., 1996, Kothapalli et al., 2022). Its direct in vivo measurement method is urgently needed to provide an early AD biomarker. This is particularly relevant for rodent models, such as the well-established Tg4510 used in this study. Cognitive and behavioral markers can be used, but do not provide information on neurodegeneration and suffer from low reproducibility, low throughput, and controversial translational value. Instead, a quantitative, longitudinal insight into neuronal loss is desired. In this study a method is proposed for using R2* relaxometry as a validated biomarker of the neuronal loss. Brain regions with low R2* volume fraction (DMVF) and the heterogeneity of R2*, quantified as histogram standard deviation, both show high sensitivity for capturing longitudinal neurodegeneration.

Direct biological interpretation of MR parameters is challenging, yet a significant correlation observed between the R2*-based Dark Matter and myelin content in the mouse hippocampus suggests the measured changes may be caused by disruption of cell membranes and concomitant loss of myelinated axons due to neuronal cell death. These conclusions are supported by prior research. Experiments in Tg4510 mice model showed correlated changes between myelin and mean T2* (O’Callaghan et al., 2017), albeit at a later age. Human studies (Wen et al., 2018) used genetic information available from the Allen Human Brain Atlas (Allen Institute) to establish a proportionality relationship between neuronal density and R2t* metric of qGRE signal. Since R2t* is a tissue-cellular-specific sub-component of R2* (R2* = R2t* +R2’), our findings of progressive reduction in R2* are in agreement with progressive reduction in neuronal density. This reduction is also in agreement with previous human studies (Kothapalli et al., 2022) reported increase Dark Matter (reduced R2t*) in people with Alzheimer disease. Additional studies will be performed to elucidate the detailed biological mechanism underlying the observed relationship, including spatial co-registration of the MRI and ex vivo images.

There are some limitations to the discussed work. Most tauopathy mouse models, including Tg4510, are characterized by rapid progression of the neurodegeneration. Capturing of neuronal density loss prior to brain atrophy is therefore very challenging. Instead, the two effects were observed in parallel.

Given the severe brain deformation characteristic for the model, automated registration and morphometry was not performed. Instead, manual segmentation was performed to minimize errors. In future studies a Tg4510-specific atlas may be assembled and used for automated segmentation. In addition, the signal-to-noise ratio obtained in mouse brain experiment did not allow for separation of R2* components into tissue cellular specific (R2t*) and vascular (R2’) as proposed clinically (Kothapalli et al., 2022). To ensure minimal contribution of the vascular BOLD effect to the measured signal, isoflurane anesthesia mixed in 100% O2 was used in the study. Future work will explore increasing the signal-to-noise to allow for R2* component separation using a cryocooled RF coil.

The findings of this study can be readily translated into rodent drug discovery studies, comparing the temporal rates of change in R2* Dark Matter abundance and R2* histogram heterogeneity between the treatment groups, to provide a histologically-proved biomarkers of neuronal loss in the brain, complementary to traditional atrophy quantification. Rescue of the neurodegenerative phenotype characteristic for the model may aid pharmacodynamics and mechanism of action evaluation of a drug candidate, and ultimately offer valuable in vivo proof of efficacy to support further development.

Acknowledgement

D.A.Y. and A.L.S. supported by NIH grants RF1 AG077658 and RF1 AG082030.

Footnotes

CRediT authorship contribution statement

Michal R. Tomaszewski: Writing – review & editing, Writing – original draft, Visualization, Investigation, Formal analysis, Data curation, Conceptualization. Alexander L. Sukstanskii: Validation, Methodology, Formal analysis, Data curation, Conceptualization. Hyking Haley: Project administration, Investigation, Data curation. Xiangjun Meng: Project administration, Investigation, Data curation. Corin O. Miller: Writing – review & editing, Supervision, Funding acquisition, Conceptualization. Dmitriy A. Yablonskiy: Writing – review & editing, Supervision, Resources, Project administration, Methodology, Funding acquisition, Formal analysis, Conceptualization.

Declaration of competing interest

M.R.T., X.M., H.H. and C.O.M. are employed by and hold stock of Merck & Co., Inc. Other authors have no conflicts to declare.

Data availability

Analysis code and processed data is available upon reasonable request following establishment of a data sharing agreement.

References

  1. Carson RE, et al. , 2022. Imaging of synaptic density in neurodegenerative disorders. J. Nuclear Med 63 (Supplement 1), 60S–67S. [DOI] [PubMed] [Google Scholar]
  2. Congdon EE, et al. , 2023. Tau-targeting therapies for Alzheimer disease: current status and future directions. Nature Rev. Neurol 19 (12), 715–736. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Degiorgis L, et al. , 2020. Brain network remodelling reflects tau-related pathology prior to memory deficits in Thy-Tau22 mice. Brain J. Neurol 143 (12). [DOI] [PubMed] [Google Scholar]
  4. Gomez-Isla T, et al. , 1996. Profound loss of layer II entorhinal cortex neurons occurs in very mild Alzheimer’s disease. J. Neurosci 16 (14), 4491–4500. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. He X, Zhu M, Yablonskiy DA, 2008. Validation of oxygen extraction fraction measurement by qBOLD technique. Magn. Reson. Med 60 (4), 882–888. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Jack C, et al. , 2000. Rates of hippocampal atrophy correlate with change in clinical status in aging and AD. Neurology 55 (4), 484–490. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Jack CR, et al. , 1992. MR-based hippocampal volumetry in the diagnosis of Alzheimer’s disease. Neurology 42 (1), 183–183. [DOI] [PubMed] [Google Scholar]
  8. Jack CR, et al. , 1999. Prediction of AD with MRI-based hippocampal volume in mild cognitive impairment. Neurology 52 (7), 1397–1397. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Jack CR, et al. , 2016. A/T/N: An unbiased descriptive classification scheme for Alzheimer disease biomarkers. Neurology 87 (5), 539–547. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Jankowsky J, Z H, 2017. Practical considerations for choosing a mouse model of Alzheimer’s disease. Mol. Neurodegener 12 (1). [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Kesslak JP, Nalcioglu O, Cotman CW, 1991. Quantification of magnetic resonance scans for hippocampal and parahippocampal atrophy in Alzheimer’s disease. Neurology 41 (1), 51–51. [DOI] [PubMed] [Google Scholar]
  12. Kothapalli S, et al. , 2022. Quantitative gradient echo MRI identifies dark matter as a new imaging biomarker of neurodegeneration that precedes tisssue atrophy in early alzheimer’s disease. J. Alzheimers. Dis 85 (2), 905–924. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Levasseur VA, et al. , 2022. Stronger Microstructural damage revealed in multiple sclerosis lesions with central vein sign by quantitative gradient echo MRI. J. Cent. Nerv. Syst. Dis 14, 11795735221084842. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Luo J, et al. , 2012. Gradient echo plural contrast imaging–signal model and derived contrasts: T2*, T1, phase, SWI, T1f, FST2*and T2*-SWI. Neuroimage 60 (2), 1073–1082. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Luo J, et al. , 2014. Gradient echo magnetic resonance imaging correlates with clinical measures and allows visualization of veins within multiple sclerosis lesions. Mult. Scler 20 (3), 349–355. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Mamah D, et al. , 2015. Subcomponents of brain T2* relaxation in schizophrenia, bipolar disorder and siblings: A Gradient Echo Plural Contrast Imaging (GEPCI) study. Schizophr. Res [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Mecca AP, et al. , 2020. In vivo measurement of widespread synaptic loss in Alzheimer’s disease with SV2A PET. Alzheimer’s Dementia 16 (7), 974–982. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Minoshima S, et al. , 2022. 18F-FDG PET imaging in neurodegenerative dementing disorders: insights into subtype classification, emerging disease categories, and mixed dementia with copathologies. J. Nuclear Med 63 (Supplement 1), 2S–12S. [DOI] [PubMed] [Google Scholar]
  19. Ni R, 2022. Magnetic resonance imaging in tauopathy animal models. Front. Aging Neurosci 13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. O’Callaghan J, et al. , 2017. Tissue magnetic susceptibility mapping as a marker of tau pathology in Alzheimer’s disease. Neuroimage 159. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Park Y-G, et al. , 2019. Protection of tissue physicochemical properties using polyfunctional crosslinkers. Nat. Biotechnol 37 (1), 73–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Patel KR, et al. , 2015. Detection of cortical lesions in multiple sclerosis: A new imaging approach. Multiple Sclerosis J. Exper., Trans. Clin 1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Price JL, et al. , 2001. Neuron number in the entorhinal cortex and CA1 in preclinical Alzheimer disease. Arch. Neurol 58 (9), 1395–1402. [DOI] [PubMed] [Google Scholar]
  24. Ramsden M, et al. , 2005. Age-dependent neurofibrillary tangle formation, neuron loss, and memory impairment in a mouse model of human tauopathy (P301L). J. Neurosci 25 (46), 10637–10647. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Sahara N, et al. , 2014. Age-related decline in white matter integrity in a mouse model of tauopathy: an in vivo diffusion tensor magnetic resonance imaging study. Neurobiol. Aging 35 (6). [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Santacruz K, et al. , 2005. Tau suppression in a neurodegenerative mouse model improves memory function. Science 309 (5733). [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Sati P, et al. , 2010. In vivo quantitative evaluation of brain tissue damage in multiple sclerosis using gradient echo plural contrast imaging technique. Neuroimage 51 (3), 1089–1097. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Selkoe DJ, 2002. Alzheimer’s disease is a synaptic failure. Science (1979) 298 (5594), 789–791. [DOI] [PubMed] [Google Scholar]
  29. Terry RD, et al. , 1991. Physical basis of cognitive alterations in Alzheimer’s disease: synapse loss is the major correlate of cognitive impairment. Ann. Neurol 30 (4), 572–580. [DOI] [PubMed] [Google Scholar]
  30. Ulrich X, Yablonskiy DA, 2016. Separation of cellular and BOLD contributions to T2* signal relaxation. Magn. Reson. Med 75 (2), 606–615. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Wells J, et al. , 2015. In vivo imaging of tau pathology using multi-parametric quantitative MRI. Neuroimage 111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Wen J, et al. , 2015. Detection and quantification of regional cortical gray matter damage in multiple sclerosis utilizing gradient echo MRI. NeuroImage: Clin. 9, 164–175. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Wen J, et al. , 2018. Genetically defined cellular correlates of the baseline brain MRI signal. Proc. Natl. Acad. Sci. U S. A 115 (41), E9727–E9736. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Wen J, Cross AH, Yablonskiy DA, 2014. On the role of physiological fluctuations in quantitative gradient echo MRI: implications for GEPCI, QSM, and SWI. Magn. Reson. Med [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Xiang B, et al. , 2019. Single scan quantitative gradient recalled echo MRI for evaluation of tissue damage in lesions and normal appearing gray and white matter in multiple sclerosis. J. Magn. Reson. Imaging 49 (2), 487–498. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Xiang B, et al. , 2019. Evaluation of myelin damage in multiple sclerosis with smart MRI. Annals of Neurology. WILEY, NJ USA, 111 RIVER ST, HOBOKEN 07030–5774. [Google Scholar]
  37. Xiang B, et al. , 2020. In vivo evolution of biopsy-proven inflammatory demyelination quantified by R2t* mapping. Ann. Clin. Transl. Neurol 7 (6), 1055–1060. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Xiang B, et al. , 2020. Quantitative assessment of multiple sclerosis tissue damage and partial repair in a biopsy proven demyelinating brain lesion using gradient recalled echo imaging. Multiple Sclerosis J. 26 (1_ SUPPL), 93–94. [Google Scholar]
  39. Xiang B, et al. , 2022. Tissue damage detected by quantitative gradient echo MRI correlates with clinical progression in non-relapsing progressive MS. Multiple Sclerosis J. 28 (10), 1515–1525. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Yablonskiy DA, 1998. Quantitation of intrinsic magnetic susceptibility-related effects in a tissue matrix. Phantom study. Magn. Reson. Med 39 (3), 417–428. [DOI] [PubMed] [Google Scholar]
  41. Yablonskiy DA, et al. , 2013. Voxel spread function method for correction of magnetic field inhomogeneity effects in quantitative gradient-echo-based MRI. Magn. Reson. Med 70 (5), 1283–1292. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Yablonskiy DA, Haacke EM, 1994. Theory of NMR signal behavior in magnetically inhomogeneous tissues: the static dephasing regime. Magn. Reson. Med 32 (6), 749–763. [DOI] [PubMed] [Google Scholar]
  43. Yablonskiy DA, Sukstanskii AL, 2024. Quantum dipole interactions and transient hydrogen bond orientation order in cells, cellular membranes and myelin sheath: Implications for MRI signal relaxation, anisotropy, and T1 magnetic field dependence. Magn. Reson. Med n/a(n/a). [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Young PNE, et al. , 2020. Imaging biomarkers in neurodegeneration: current and future practices. Alzheimers. Res. Ther 12 (1), 49. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Zhang Y, et al. , 2023. Amyloid β-based therapy for Alzheimer’s disease: challenges, successes and future. Signal. Transduct. Target. Ther 8 (1), 248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Zhao Y, et al. , 2016. On the relationship between cellular and hemodynamic properties of the human brain cortex throughout adult lifespan. Neuroimage 133, 417–429. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Zhao Y, et al. , 2017. In vivo detection of microstructural correlates of brain pathology in preclinical and early Alzheimer disease with magnetic resonance imaging. Neuroimage 148, 296–304. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Analysis code and processed data is available upon reasonable request following establishment of a data sharing agreement.

RESOURCES