Simultaneous T₁ and T₂ Brain Relaxometry in Asymptomatic Volunteers using Magnetic Resonance Fingerprinting

Abstract

Magnetic resonance fingerprinting (MRF) is a method of image acquisition that produces multiple MR parametric maps from a single scan. Here, we describe the normal range and progression of MRF-derived relaxometry values with age in healthy individuals. 56 normal volunteers (ages 11-71 years, M:F 24:32) were scanned. Regions of interest were drawn on T₁ and T₂ maps in 38 areas, including lobar and deep white matter, deep gray nuclei, thalami and posterior fossa structures. Relaxometry differences were assessed using a forward stepwise selection of a baseline model including either gender, age, or both, where variables were included if they contributed significantly (p<0.05). Additionally, differences in regional anatomy, including comparisons between hemispheres and between anatomical subcomponents, were assessed by paired t-tests. Using this protocol, MRF-derived T₁ and T₂ in frontal WM regions were found to increase in with age, while occipital and temporal regions remained relatively stable. Deep gray nuclei, including substantia nigra, were found to have age-related decreases in relaxometry. Gender differences were observed in T₁ and T₂ of temporal regions, cerebellum and pons. Males were also found to have more rapid age-related changes in frontal and parietal WM. Regional differences were identified between hemispheres, between genu and splenium of corpus callosum, and between posteromedial and anterolateral thalami. In conclusion, MRF quantification can measure relaxometry trends in healthy individuals that are in agreement with current understanding of neuroanatomy and neurobiology, and has the ability to uncover additional patterns that have not yet been explored.

Introduction

Physiological aging changes in cerebral gray and white matter (WM) have been well documented in the neurobiology literature. Normal aging is associated with dendritic pruning, axonal loss, demyelination, as well as synaptic and neuronal loss (1-4). Several magnetic resonance imaging (MRI) based metrics such as diffusion tensor imaging (DTI), diffusion, volumetry, and magnetization transfer ratio (MTR) have been utilized to quantify age related changes (5-13). MRI relaxometry techniques have also been used to quantify age-related changes in T₁, T₂ and T₂* relaxation properties in healthy individuals (14-23). All relaxometry studies thus far have utilized separate sequences for quantifying one relaxation property at a time by measuring the signal recovery after spin inversion (T₁) or the decay of the measured MR signal (T₂ or T₂*). Typically, such experiments suffer from long acquisition times and limited accuracy, which can limit utility.

Magnetic resonance fingerprinting (MRF) is a recently introduced method that simultaneously and rapidly measures multiple tissue properties, with initial application in measuring T₁, and T₂. This technique is based on the premise that acquisition parameters can be varied in a pseudorandom manner such that each combination of tissue properties will have a unique signal evolution. Using the Bloch equations, known acquisition parameters and all possible range of values and combinations of the properties of interest, a dictionary of all possible signal evolutions can be created. The actual signal evolution in each voxel can then be compared to the dictionary entry, and the best dictionary match yields the property values for that voxel (24).

With MRF there is now the possibility to observe small changes in multiple tissue relaxation properties simultaneously. However, to date, no study has been performed to describe the normal range and progression of MRF derived relaxometry values in healthy individuals. In this study we present simultaneous quantification of regional brain T1 and T2 relaxation times in healthy volunteers using MRF and assess differences in tissue properties due to age, gender and laterality of hemispheres. We further compare different best-fit options for regression analysis of age and brain relaxometry and also assess effects of age-gender interactions on these findings, and assess these findings in context of the known literature on relaxometry measurements with aging.

Methodology

Participant recruitment

Informed written consent was obtained from all participants according to the protocol approved by the local IRB. Multi-slice MRF data were acquired in 56 healthy volunteers with an age range of 11 to 71 years. There were 24 males (ages 11 to 71 years) and 32 females (ages 18 to 63 years) with an overall median age of 39 years (Fig. 1). Of these, 53 participants were right handed. One of the participants had a remote history of craniotomy for excision of a meningioma; another volunteer had a remote history of surgical correction for Chiari 1 malformation. No other participant had a history of structural neurological disease, or a known psychiatric disease. None of the participants revealed any overt parenchymal abnormalities on clinical T₂ weighted images in the analyzed regions.

Figure 1. Age distribution of all participants in the study.

MRF acquisition

MRF scans were obtained on 3.0 T Siemens scanners (Verio and Skyra; Siemens Healthcare, USA) using standard 20-channel head coils. The acquisition technique has been previously described in detail.²⁴ In MRF the parameters are continuously changed throughout the acquisition to create the desired spatial and temporal incoherence. The flip angle, phase, repetition time (TR), echo time (TE) and sampling patterns are all varied in a pseudorandom fashion.²⁴ The parameters used for MRF acquisition were as follows: field of view: 300 × 300 mm², matrix size: 256 × 256, slice thickness: 5 mm, flip angle: 0 to 60 degrees, TR: 8.7 to 11.6 ms, RF pulse: sinc pulse with duration of 800 μs and time-bandwidth product of 2. In a total acquisition time of 30.8s, 3000 images were acquired for each slice. The echo time (TE) was half of TR and varied with each TR. The MRF acquisition was planned on whole brain clinical standard T₂-weighted images which were acquired with TR: 5650 ms, TE: 94 ms, FOV: 230 mm, slice thickness: 4 mm, flip angle: 150 degrees. Approximately 4-5 two-dimensional MRF slices were acquired through the whole brain for each individual depending on the head position. The entire study for each volunteer including positioning time was about 10 minutes in duration.

Data Processing

Using simulation, a dictionary of signal evolutions that could arise from all possible combinations of materials or system related properties was generated. A total of 287709 signal time courses, each with 3000 time points, with different sets of T₁, T₂ and off-resonance parameters were simulated for the dictionary. The ranges of T₁ and T₂ were chosen according to the typical physiological ranges of the tissues in the brain. T₁ values between 100 and 3000 ms and T₂ values between 10 and 500 ms were included in this dictionary. The off-resonance values included the range between -400 Hz to 400 Hz. The total simulation time was 5.3 minutes. The vector dot product between the measured signal and each dictionary entry was calculated, and the entry yielding the highest dot product was selected as the closest match to the acquired signal.¹⁴ The final output consisted of quantitative T₁, T_2, off-resonance and proton density maps (Fig. 2). MRF-based proton density values are affected by the type of acquisition as well as sensitivity of the receiver coil and thus are not purely tissue specific. Therefore only T₁ and T₂ maps were utilized for further anatomical analysis.

Figure 2. MRF-derived quantitative maps.

(A) T₁, (B) T₂, (C) proton density and (D) off-resonance maps from a single acquisition with duration of 30.8 seconds

Data Analysis

All data processing and analysis was performed on MATLAB (version R2013b, The Mathworks, Natick, MA) and SAS version 9.4 (SAS Institute, Inc, Cary NC). A region of interest (ROI) based analysis was performed on the relaxometry maps as follows. For every subject, a fellowship-trained neuroradiologist manually drew the ROIs from which mean T₁ and T₂ measures were extracted. A total of 38 ROIs (17/hemisphere plus 4 midline) were drawn for each subject (Fig. 3). The selected regions constituted important WM regions, deep gray nuclei as well as posterior fossa structures. Cortical gray matter was not studied to avoid partial volume effects from cerebral spinal fluid (CSF) and WM. T₁ and T₂ maps with narrow window settings and magnified views were used to clearly identify each anatomical region and draw the ROIs. The ROI size depended on the region analyzed, and ranged from 4 to 10 mm². Care was taken to place the ROI in the center of the sampled region with careful separation from adjacent structures to avoid partial volume effects. Regions with gross visible artifacts or distortion were excluded from measurements.

Figure 3. Region of interest (ROI) locations.

(A) 1-superior frontal white matter, 2-centrum semiovale, (B) 3-frontal white matter,(WM) 4-caudate nucleus, 5-putamen, 6-globus pallidus, 7-medial thalamus, 8-lateral thalamus, 9-internal capsule, 10-genu, 11-splenium, 12-parietal WM, (C) 13-substantia nigra, 14-red nucleus, 15-temporal WM, 16-occipital WM, (D) 17-middle cerebellar peduncle, 18-cerebellum, 19-dentate nucleus, 20-vermis, 21-pons.

Statistical Analysis

T₁ and T₂ values extracted from the MRF data were analyzed based on review of prior literature. Previous relaxometry studies have utilized either a linear or a polynomial regression model to assess the relationship between age and relaxometry (14-23). For this study, age and gender effects were first examined using forward stepwise selection to select a baseline model including either gender, age, or both, where variables were included at each step if they were significant with p-value less than 0.05. For regions where the baseline model included age, we then tested whether adding a quadratic term to the model significantly (p<0.05) improved fit. Also, for regions with significant linear age effects, effects in males and females were compared using test of equality between slopes to assess for age and gender interaction. Based on the slopes and intercepts, age-gender interplay was categorized as either ‘age+gender’ effect or ‘age*gender’ effect. Age+gender effect included regions where males and females had similar slopes with respect to age, but different intercepts. Age*gender effect included regions where each gender had significantly different slopes and intercepts with respect to age. Thus, for each brain region, we evaluated changes of MRF-based T₁ and T₂ with age using linear and quadratic models, differences between genders, and differences in the trajectory of age effects between genders.

To test for differences between right and left hemispheres, regional relaxometry data from only right-handed participants (n=53) were used. In this sub-analysis a paired t-test was performed to compare relaxometry measures for each region across hemispheres. A paired t-test was also used to compare different components within a region, specifically between the medial and lateral thalami and between the genu and splenium of corpus callosum. For this subgroup analysis pooled data from right and left handed subjects were analyzed.

For statistical analysis, all comparisons with p-value of less than 0.05 before correction for multiple comparisons were considered significant results and discussed. This was done with an intention of describing all identifiable trends that may have physiological implications. However, correction for multiple comparisons testing using the Bonferroni method was also utilized, and outcomes that were statistically significant overall were identified.

Results

All regions with field inhomogeneity and susceptibility artifacts were excluded from analysis. The largest number of field inhomogeneity and banding artifacts was seen in the region of genu of the corpus callosum (n=15). T₂ maps were more susceptible to field inhomogeneity artifacts compared to T₁. While best attempts were made to include all ROIs in the collected slices, slight variations in slice placement during imaging resulted in omission of some regions, most commonly the splenium of corpus callosum (n=8).

Aging progression

When examining T₁, positive linear correlations with age were observed in three frontal WM regions and genu of corpus callosum. Negative linear correlations were seen in the left substantia nigra (SN) (Table 1, Fig. 4A). Quadratic trends were observed in three frontoparietal WM regions and the right SN, with the latter showing an overall decline in T₁ with age (Table 2, Fig. 4B). When examining T₂, positive linear correlations were seen in left frontal WM and medial left thalamus, while negative linear correlations with age were detected in bilateral SN (Table 1, Fig. 4A). Quadratic relationships with age were observed in right frontal WM and left dentate nucleus, with additional gender effect in right frontal WM described further in the following section (Table 2, Fig. 4B).

Table 1. Regions showing significant linear relationship between relaxation parameters and age without any gender effects.

Region Name	Intercept	Slope	p-value	R2
T1 Relaxometry
Right Superior Frontal WM	821.98	0.848	0.045	0.07
Right Centrum Semiovale	843.14	1.179	0.010	0.11
Left Centrum Semiovale	850.49	1.081	0.029	0.08
Corpus Callosum Genu	743.60	1.086	0.029	0.12
Left Substantia Nigra	976.9	-2.893	0.0004*	0.24
T2 Relaxometry
Left Frontal WM	58.64	0.1840	0.0002*	0.24
Left Thalamus (medial)	59.21	0.1071	0.0299	0.09
Right Substantia Nigra	46.21	-0.1090	0.015	0.13
Left Substantia Nigra	47.29	-0.1321	0.011	0.13

^*Statistically significant after correcting for multiple comparison testing using Bonferroni method

Figure 4. Regions with significant T₁, T₂ correlation with ag.e.

(A) Regions with significant linear relationship between T₁, T₂ and age. (B) Regions with significant quadratic relationship between relaxation parameters and age

Table 2. Regions showing significant quadratic relationship between relaxation parameters and age.

Region Name	Intercept	Age	Age2	p(Age)2	R2
T1 Relaxometry
Right Frontal WM	811.8	-3.3619	0.0559	0.045	0.21
Left Frontal WM	848.3	-4.4117	2.3849	0.021	0.19
Left Parietal WM	905.5	-5.5573	0.0928	0.001*	0.42
Right Substantia Nigra	1093.8	-10.287	0.0965	0.008*	0.39
T2 Relaxometry
Right Frontal WMa	67.18	-0.6168	0.0088	0.013	0.32
- Female	67.18	-0.6168	0.0088	0.013	0.32
- Male	71.52	-0.6168	0.0088
Left Dentate Nucleus	74.62	-0.6124	0.0060	0.046	0.16

^aFor Right Frontal WM, the quadratic model also included a term for gender, which was statistically significant (p=0.023), indicating a difference in intercepts between males and females. Results are displayed as separate regressions for males and females, having different intercepts but the same linear and quadratic terms for age.

^*Statistically significant after correcting for multiple comparison testing using Bonferroni correction technique

Gender differences

Differences in MRF-derived relaxometry between genders were observed in the absence of significant correlation with age. In the analysis of T₁, of the 38 regions examined, left temporal WM, bilateral cerebellar hemispheres, and pons showed differences between gender with higher T₁ in males as compared to females and no significant change with age. In the analysis of T₂, significant difference between genders was detected in the right lentiform nucleus.

Differences between genders with age effects were categorized as either ‘age+gender’ effect (males and females had similar slopes with respect to age, but different intercepts) or ‘age*gender’ effect (each gender had significantly different slopes and intercepts). Recall, age effects could be fit with a linear or quadratic model. In T₁ analysis, left superior frontal and right parietal WM showed a linear age+gender effect (Fig. 5A). In T₂ analysis, age*gender interaction was seen in bilateral superior frontal, parietal WM and centrum semiovale. Linear age+gender effect was observed in right superior frontal WM, and quadratic age+gender effect was observed in right frontal WM and right dentate nucleus (Fig. 5B).

Figure 5. Regions with significant age and gender effects.

(A) Regions with significant linear age+gender effects; in these models, the slope of linear regression on age for males and females is similar but the intercepts are significantly different. (B) Regions with significant age*gender effect on T₂ relaxometry; in this model, the slope of linear regression on age between males and females is statistically significant.

Of all the gender differences measured, after adjusting for multiple comparisons testing, only T₁ variations in the right parietal WM (p<0.0001, R²=0.30) and T₂ differences in right superior frontal WM (p<0.0001, R²=0.30) remained statistically significant.

Regional differences

Only right-handed individuals (n=53) were included in this analysis and 34-paired regions were studied. Several regions with T₁ and T₂ differences between right and left hemispheres were identified (Table 3). In the analysis within regions, splenium of corpus callosum had significantly higher T₁ but a lower T₂ as compared to the genu. The medial components of bilateral thalami showed higher T₁ and T₂ values as compared to the lateral components.

Table 3. Relaxometry differences across hemispheres and within regions.

Differences across hemispheres (includes only right handed participants)
	T1 (Right minus Left)				T2 (Right minus Left)
Region	N	Mean	SD	p-value	N	Mean	SD	p-value
Superior Frontal WM	53	-6.29	25.97	0.083	52	-1.35	3.81	0.013
Centrum semiovale	52	-2.98	28.03	0.446	52	-1.63	3.81	0.003
Frontal WM	52	-13.04	39.42	0.020	52	-5.01	7.68	<0.0001*
Caudate nucleus	49	-4.52	48.29	0.515	45	2.78	7.72	0.020
Putamen	52	-17.00	40.32	0.003	51	-1.98	5.22	0.009
Parietal WM	52	1.13	42.98	0.850	52	-3.92	5.34	<0.0001*
Internal capsule	50	-14.52	33.02	0.003	48	2.99	7.21	0.006
Occipital WM	48	-17.98	56.58	0.032	48	-3.52	5.88	0.0001*
Temporal WM	45	33.02	58.97	0.0005*	46	-0.83	6.33	0.377
Middle Cerebellar peduncle	48	20.08	51.92	0.010	48	-1.58	5.05	0.035
Dentate nucleus	47	-6.44	48.71	0.369	45	-4.11	4.96	<0.0001*
Cerebellum	49	2.72	69.67	0.785	47	-3.48	7.02	0.001*
Differences within structures (includes right and left handed participants)
	T1				T2
Region	N	Mean	SD	p-value	N	Mean	SD	p-value
Rt thalamus (medial-lateral)	52	91.48	54.76	<0.0001*	51	4.55	5.35	<0.0001*
Lt thalamus (medial-lateral)	51	110.91	52.77	<0.0001*	51	6.72	4.40	<0.0001*
C Callosum (genu-splenium)	35	-68.00	55.63	<0.0001*	35	4.37	7.34	0.0012*

^*Statistically significant after correcting for multiple comparison testing using Bonferroni correction technique

Discussion

This is the first in vivo use of magnetic resonance fingerprinting at 3.0 T for measuring tissue properties of multiple brain regions in healthy human subjects across different age groups.

At a microstructural level, brain aging is characterized by loss of myelinated fibers, myelin pallor, ballooning and redundant myelination; macroscopically there is loss of grey and WM volume and expansion of CSF spaces (25-28). Increase in free water and decrease in water bound to macromolecules (such as myelin) is reflected by a lower MTR in older age groups (29, 30). The increase in gliosis, free water content, loss of myelination and other aging changes also result in longer T₁ and T₂ relaxation times in WM. Although the published literature varies in types of statistical modeling employed and regional predilection of findings, all studies agree that there is an overall increase in T₁ (1/R₁) and/or T₂ (1/R₂) in various WM regions/tracts with increasing age (31-33).

A recent study has measured the R₁ of various WM tracts over age and found that R₁ increased from childhood up to age of about 40 years and then decreases to the 8-year-old levels between ages 70-80 (13). In this study comparable trends are seen in T₁ of bilateral frontal and left parietal WM, with a dip in T₁ values between 30-50 years followed by an increase in the later decades (Fig. 4B, Table 2). Various volumetry and DTI studies have consistently demonstrated a frontal predilection for age-related changes (34-37). DeCarli et al. also showed that the volumes of bilateral temporal lobes stayed stable across the human lifespan (34). These findings support the results shown here, which demonstrate that age effects on WM relaxometry are significant in frontal and parietal regions whereas occipital and temporal relaxometry values stay relatively stable. Also the fact that the quadratic age model is a significantly better fit for certain frontal and parietal white regions over a linear age model alludes to a dynamic state of tissue turnover in these regions throughout the adult life.

White matter in the genu of the corpus callosum (CC) also demonstrated increased T₁ with age in this study. Prior DTI and relaxometry studies exploring the effects of aging on CC microstructure have found that the anterior portions of the corpus callosum (including the genu) are more susceptible to age-dependent changes as compared to the splenium (38-40). More specifically, DTI studies showed greater decreases in fractional anisotropy in the genu, what was explained by increases in free water content and demyelination in the CC with age. Such microstructural changes would also cause an increase in T₁ relaxometry (Table 2).

With age, deep gray nuclei show drops in T₂ and less frequently T₁ values secondary to increasing mineralization and iron deposition (13,32,33,41,42). We identified similar trends in left dentate nucleus and bilateral SN, the latter being statistically significant. T₂ shortening in SN can be explained by increasing iron deposition as a part of physiological aging process, and has been extensively reported in the literature (43-47). On the other hand, the age-dependent decrease in T₁ of SN has not been as extensively explored; recent study assessing the relationship between R₁ of SN and age showed findings similar to our results (48). Histopathological studies of SN have shown that there is nearly a 10% decrease in the number of neuromelanin containing neurons per decade in neurologically intact individuals (49). As neuromelanin inherently has a T₁ shortening effect, in theory this loss should manifest as T₁ lengthening with age, but the data indicate a different effect to be dominant. The findings seen here may be an outcome of combination of iron deposition and extraneuronal melanin deposition that are also seen with normal aging, both of which are expected to shorten T₁ (44,49-51). In this study, T₁ and T₂ in SN were determined to decrease with age in a linear or quadratic pattern.

Currently, there is no consensus in the neuroimaging literature on whether a linear or quadratic model is the best fit for regression analysis of age and relaxometry. Also there is no physiologic reason to assume that the entire brain should conform to one model uniformly over the other. Our results suggest that for the more dynamically changing frontal WM regions, the quadratic model may be a better fit than the linear model, especially for T₁ [Fig. 4B].

Two major differences in gender relaxometry were seen in this study, the first being different effects of aging on certain WM regions for males and females. In older age groups, males were observed to have higher relaxation time measurements in frontal and parietal WM as compared to females. A few studies looking at age and gender interactions in the past have shown that frontotemporal volume loss with age is more prominent in males, although a few other imaging studies have shown no such interaction (8,52-55). Coffey et al. found that there was greater age-related increase in sulcal and Sylvian CSF volumes with lower size of parieto-occipital regions in males as compared to females (56). The gender effects on aging seen in our study are an additional piece of evidence that could reflect the greater predilection of males towards neurodegenerative processes and neurocognitive decline that become more prominent with age (56-58). The second major difference in gender relaxometry that was identified in this study was in mean relaxometry of temporal regions, cerebellum and pons. Similar gender effects seen previously have been attributed to sexual dimorphism arising from effects of sex steroids on microscopic processes such as glial proliferation, myelination, presence of paramagnetic substances as well as macrostructural phenotypes of gray and WM volumes (7,8,32,54,59-61).

In right-handed subjects, several areas of hemispheric asymmetry were identified in frontal, parietal and temporal WM, internal capsule region, and dentate nuclei. These regional differences hint at underlying microstructural distinctions arising out of asymmetry in motor cortex and WM connectivity (62). Prior attempts to evaluate cerebral laterality with techniques such as morphometry, DTI, functional MRI have shown that several subtle macro-structural as well as microstructural differences in cerebral hemispheres can be identified although there is no single predominant pattern that has emerged (63-67).

In this study, the genu of the CC showed significantly lower T₁ and higher T₂ values as compared to the splenium. Previous DTI studies have shown higher fractional anisotropy in the splenium of the CC as compared to the genu region (68,69). Thus these two regions of the CC are known to have measurable differences on diffusion MRI. Several factors such as axonal fiber density, diameter of fibers, orientation, degree of myelination, and overall microstructural integrity that affect the diffusion metrics could also have an effect on the relaxometry characteristics of CC and explain our findings, although the exact relationship between these factors remains unexplored.

We also found interesting regional variation in relaxometry of thalami. For this analysis, it was not possible to anatomically segment the thalami into the component nuclei. Rather than analyze each thalamus in its entirety we divided it into posteromedial and anterolateral components. The posteromedial segment approximately included the regions of pulvinar and medial nuclei, whereas the anterolateral segment included the anterior and lateral regions. For both hemispheres, the T₁ and T₂ of posteromedial thalami were higher by approximately 100 and 5 ms respectively, as compared to the lateral portions. The exact cause of these differences is unclear, although a differential in gray-white matter composition, unique nuclear arrangement and differences in associated WM pathways may explain some of these findings (70). Several relaxometry studies have been attempted in normal subjects and in patients with multiple sclerosis (MS) (41,71,72). Because thalami are frequently studied in MS, our findings could have implications in designing future relaxometry studies in patients, as it may be necessary to analyze the medial and lateral portions of the thalami separately.

This study utilizes the original MRF technique with 2-D acquisitions and an in-plane resolution of 1.2 mm (24). Lack of 3-D whole brain data limited ability to select brain regions and necessitated analysis using the time intensive ROI method. Future iterations of MRF acquisitions seek to address these limitations with improved in-plane resolution and 3-D acquisition capabilities while improving processing speeds and patient comfort (73,74). Relaxometry measurements from certain regions such as the genu of the corpus callosum are limited by the presence of field inhomogeneity and banding artifacts. These artifacts are more typical for all types of balanced SSFP based sequences and are commonly seen near air-tissue interface, where large field inhomogeneity is introduced. The incidence of these artifacts could be considerably reduced in future studies by using a FISP based MRF acquisition technique (75).

Limitations of this study include lack of details about study participant medical history that may affect brain anatomy and microstructure, including history of caffeine and alcohol intake, smoking, and diseases such as diabetes mellitus, hypertension, endocrinopathies or current medications, and these factors could potentially alter relaxation parameters. No mini mental state examination or psychological testing was administered to the participants as a part of this study, though all participants demonstrated understanding of the consent form. Our ROIs included deep gray nuclei and WM regions; cortical gray matter was not analyzed.

In conclusion, this pilot study introduces magnetic resonance fingerprinting as a rapid multiparametric in vivo quantitation tool in normative brain imaging and demonstrates that MRF can identify and quantify differences in brain parenchyma related to age, gender, hemisphere, and anatomy. This T₁ and T₂ normative database can be used as a reference for future MRF studies in various disease states. Dedicated efforts to improve in-plane resolution, facilitate 3-D coverage and reduce inhomogeneity artifacts are underway to develop an efficient and powerful quantitation tool for applications in neuroimaging and beyond.

Abbreviations

MRI

Magnetic resonance imaging

MRF

magnetic resonance fingerprinting

WM

white matter

DTI

diffusion tensor imaging

TR

repetition time

TE

echo time

ROI

region of interest

CSF

cerebral spinal fluid

SN

substantia nigra

CC

corpus callosum