Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Magn Reson Imaging. Author manuscript; available in PMC 2012 November 8.
Published in final edited form as:
PMCID: PMC3492944

Preliminary Evidence of Altered Gray and White Matter Microstructural Development in the Frontal Lobe of Adolescents With Attention-Deficit Hyperactivity Disorder: A Diffusional Kurtosis Imaging Study



To investigate non-Gaussian water diffusion using diffusional kurtosis imaging (DKI) to assess age effects on gray matter (GM) and white matter (WM) microstructural changes in the prefrontal cortex (PFC) of adolescents with attention-deficit hyperactivity disorder (ADHD) compared to typically developing controls (TDC).

Materials and Methods

In this preliminary cross-sectional study, T1-weighted magnetization-prepared rapid gradient echo (MPRAGE) and DKI images were acquired at 3T from TDC (n = 13) and adolescents with ADHD (n = 12). Regression analysis of the PFC region of interest (ROI) was conducted.


TDC show a significant kurtosis increase of WM microstructural complexity from 12 to 18 years of age, particularly in the radial direction, whereas WM microstructure in ADHD is stagnant in both the axial and radial directions. In ADHD, GM microstructure also lacked a significant age-related increase in complexity as seen in TDC; only kurtosis measures were able to detect this difference.


These findings support the prevailing theory that ADHD is a disorder affecting frontostriatal WM. Our study is the first to directly quantify an aberrant age-related trajectory in ADHD within GM microstructure, suggesting that the assessment of non-Gaussian directional diffusion using DKI provides more sensitive and complementary information about tissue microstructural changes than conventional diffusion imaging methods.

Keywords: ADHD, non-Gaussian diffusion MRI, microstructure, frontal lobe

The hallmark behavioral symptoms underlying attention-deficit hyperactivity disorder (ADHD) are abnormally excessive inattention, impulsivity, and hyperactivity. Several neuroimaging studies have begun to pinpoint potential neuroanatomical correlates of these behavioral deficits, with the frontal-striatal circuit being a primary focus due to converging evidence of reduced volume and reduced activation in executive functioning tasks (15). Diffusion tensor imaging (DTI) studies have also reported abnormal white matter (WM) microstructure in this circuit as indexed by fractional anisotropy (FA), with the majority identifying reductions in FA (1,613). However, some conflicting reports of higher FA within this circuit in ADHD have also been published (10,14,15). The goal of this preliminary cross-sectional study is to utilize a novel diffusion imaging method called diffusional kurtosis imaging (DKI) to investigate the effects of age on microstructural changes in the ADHD adolescent brain with measures that complement DTI while extracting more details regarding the previously reported microstructural abnormalities. Specifically, we aim to address the lack of analysis of the microstructure of gray matter (GM) and the paucity of ADHD studies investigating age-related developmental trajectories of tissue microstructure (14,15).

DKI is a natural extension of conventional DTI (1618) that allows for the investigation of the non-Gaussian properties of water diffusion (1922). Non-Gaussian diffusion is believed to arise from diffusion barriers, such as cell membranes and organelles, and is therefore thought to be a sensitive indicator of tissue microstructural integrity. With DKI, one obtains estimates for all the standard DTI diffusion metrics, such as mean diffusivity (MD) and FA, as well as an additional metric related to diffusional non-Gaussianity called the mean kurtosis (MK). Microstructurally more complex tissues generally have more diffusion barriers, thus causing water diffusion to deviate from Gaussianity; this is typically reflected in a higher MK value. In contrast to FA, MK is not limited to anisotropic environments, thus it uniquely permits the quantification of the microstructural integrity of both GM and WM, even in the presence of crossing fibers (21). This sensitivity to GM is of importance in examining microstructural integrity in ADHD, as reduced GM volume has been consistently identified (2).

Furthermore, DKI provides directional information about diffusivity and kurtosis along the axial direction (parallel to the principal diffusion tensor eigenvector) and radial direction (perpendicular to the principal diffusion tensor eigenvector) (22). These measures are called axial diffusivity (D||), radial diffusivity (D[perpendicular]), axial kurtosis (K||) and radial kurtosis (K[perpendicular]). Here we adopt the terms axial and radial which are synonymous to parallel and perpendicular, respectively. Axial diffusivity and kurtosis are believed to reflect axonal integrity, while radial diffusivity and kurtosis are believed to reflect myelin integrity (2325). Although the precise biological relevance of these directional metrics is still being investigated, they have been shown to be useful in better characterizing neuronal tissue microstructure (23).

To explore the sensitivity of DKI metrics in detecting age effects on GM and WM microstructural changes in the ADHD adolescent brain, we used a region of interest (ROI) analysis focusing on the prefrontal cortex (PFC). The PFC was chosen because it is a key executive functioning region of the frontal-striatal circuit known to undergo dynamic changes during normal adolescence (26), while being consistently found to be aberrant in ADHD (2). Besides volume reduction, ADHD-related abnormalities in PFC include the greatest delays in maturation of cortical thickness, reduced activation during attention and inhibition tasks (3,5,2729), and abnormal WM microstructure (1,815). By targeting the PFC, we expected to enhance previous reports of abnormal WM microstructure by examining directional diffusion information, to extend microstructural analyses to GM, and to assess age effects on WM and GM microstructure, thus providing greater insight into the developmental pathophysiology of ADHD.



DKI differs from DTI in that the magnitude of the diffusion-weighted signal in a given diffusion direction as a function of the b-value is fit to the functional form (2022):


where D is the diffusion coefficient and K is the diffusional kurtosis. Thus, the DTI signal model is generalized by adding to the exponent a term quadratic in the b-value. Because of the additional free parameter, at least three distinct b-values must be used (including b = 0). In analogy with the conventional diffusion tensor, a diffusional kurtosis tensor is used to fully characterize the orientational dependence of the kurtosis (2022). This diffusional kurtosis tensor has 15 independent degrees of freedom, and as a consequence, at least 15 independent diffusion directions must be utilized in tissues with anisotropic diffusion properties. Including the dependence on direction, the model of Eq. [1] then has a total of 22 degrees of freedom: 6 for diffusion tensor + 15 for kurtosis tensor + 1 for S(0).

A DKI dataset always contains all the information necessary to construct the diffusion tensor, and hence all the conventional DTI metrics. In this study we will consider the MD, D||, D[perpendicular], and FA, which are defined by:






Here (λ1, λ2, λ3) are the diffusion tensor eigenvalues order so that (λ1 ≥ λ2 ≥ λ3). One may readily show that the MD corresponds to the diffusion coefficient averaged over all possible diffusion directions, whereas D|| is the diffusion coefficient in the direction of the principal diffusion tensor eigenvector and D[perpendicular] is the diffusion coefficient averaged over all diffusion directions perpendicular to the principal diffusion tensor eigenvector.

With DKI, several additional diffusion metrics can be determined that quantify diffusional non-Gaussianity. All of these may be calculated from the diffusion and diffusional kurtosis tensors. For the present work, we utilize the MK, K||, and K[perpendicular], which are kurtosis analogs of the MD, D||, and D[perpendicular]. The MK is defined as the average kurtosis over all diffusion directions, whereas K|| and K[perpendicular] are defined, respectively, as kurtosis in the direction of the principal diffusion tensor eigenvector and kurtosis averaged over all diffusion directions perpendicular to the principal diffusion tensor eigenvector. Further details on the computation of these quantities have been previously described (21,22). The notions of K|| and K[perpendicular] were first introduced by Hui et al (24), although our definition of K[perpendicular] corresponds instead to that of Poot et al (30).

Participant Inclusion and Exclusion Criteria

Twelve adolescents with ADHD (12.9–17.6 years old, mean = 14.4 years, SD = 1.6) and 13 typically developing controls (TDC) (12.6–17.9 years old, mean = 14.8 years, SD = 1.7) were recruited from our institution’s Child Study Center and the local community. Informed consent was obtained as approved by our Institutional Review Board. Estimated full-scale IQ ≥ 75 measured by the Wechsler Abbreviated Scale of Intelligence (WASI) (31) and absence of known neurological or chronic medical diseases were required of all subjects. The DSM-IV diagnosis of ADHD (32) was based on parent interviews using the Schedule of Affective Disorders and Schizophrenia for Children – Present and Lifetime Version (K-SADS-PL) (33). Diagnosis of psychotic, major depressive, conduct, tic, and pervasive developmental disorders were exclusionary. Inclusion as a TDC required T-scores below 60 on the Conners’ Parent Rating Scale-Revised: Long Version DSM-IV Total Score (34). Demographic information (Table 1) and socioeconomic status using the Hollingshead Index of Social Position (35) were also obtained from parents.

Table 1
Clinical Characteristics

Image Acquisition

MRI studies were acquired at 3T (TIM Trio, Siemens Medical Solutions, Erlangen, Germany) with a transmission body coil and an 8-element phase array coil for reception. Whole-brain T1-weighted magnetization-prepared rapid gradient echo (MPRAGE) images were acquired with TR between 180° inversion pulses = 2100 msec, TE = 3.49 msec, PE direction: R > L, flip angle = 12°, 1 slab, slices per slab = 160, slice oversampling = 50%, bandwidth (BW) = 170 Hz/pixel, echo spacing = 7.8 msec, field of view (FOV) (read/phase) = 256 mm/75%, slice thickness = 1 mm, matrix = 256 × 256 in a total time of 3 minutes 47 seconds. DKI data were acquired using a dual-spin-echo echo planar imaging (EPI) diffusion sequence with TR/TE = 2300/109 msec, FOV = 256 × 256 mm2, matrix = 128 × 128, iPAT (GRAPPA = 2), 2 averages, oblique axial slices (number/thickness/gap) 15/2mm/2mm, 30 gradient encoding directions with 6 b values (0–2500 in increments of 500 s/mm2) for each direction in a total time of 11 minutes 17 seconds. Partial DKI brain coverage of the prefrontal cortex and cerebellum was acquired to focus on regions associated with ADHD deficits (15) and to reduce scan time. The acquisition orientation was determined sagittally by aligning the inferior end of the slab to the most inferior point of the hypothalamus and the most inferior point of the cerebellum’s posterior lobe (Fig. 1a). Although in the present study gaps were included to ensure that images were not contaminated by signal crosstalk, we note that a recently published DKI protocol allows for whole-brain coverage without gaps in a scan time of ≈7 minutes per average (22).

Figure 1
DKI slab acquisition and PFC ROI. a: Fifteen-slice DKI slab spanning 58 mm in the z-direction (2 mm gaps) was acquired to include the majority of the PFC and cerebellum. Acquisition angle was determined by aligning the inferior end of the slab to the ...

Image Processing

DKI data were processed using in-house software called a Diffusional Kurtosis Estimator (DKE) running in MatLab (v., 2006; MathWorks, Natick, MA) and the diffusion and kurtosis tensors were calculated on a voxel-by-voxel basis to produce parametric maps for MD (range 0–3 μm2/ms), D|| (range 0–3 μm2/ms), D[perpendicular] (range 0–3 μm2/ms), FA (range 0–1), MK (range 0–2), K|| (range 0–2) and K[perpendicular] (range 0–3) as previously described (21,22,36). We note that the DKE software is made freely available upon request. Data preprocessing with SPM2 (Statistical Parametric Mapping, Wellcome Department of Imaging Neuroscience, University College London, UK) included 3D motion correction with a spatial smoothing Gaussian filter (2.5 mm full-width-half-maximum). Cerebrospinal fluid (CSF) was removed from all parametric maps in MatLab by thresholding each subject’s MD map using a cutoff intensity value of ≥2 μm2/ms to create the applied binary CSF mask.

The parametric maps were segmented into GM and WM using SPM5 implemented in MatLab. For each subject, the MPRAGE and b = 0 images were centered at the anterior commissure. The MPRAGE was then coregistered to the b = 0 using the spm_coreg function, which stores the affine transformation matrix into the MPRAGE header (37). The coregistered MPRAGE image, in its original resolution, was then segmented into GM, WM, and CSF using the spm_pre-proc and spm_prep2sn functions (38). These segmented images were then resliced into the b = 0 space using the spm_reslice function which also applies the registration parameters in the header to the image. Lastly, the GM and WM resliced images were converted to binary masks and applied to the parametric maps (Fig. 1c,d).

Image Analysis

ImageJ (v. 1.14o, National Institutes of Health, Bethesda, MD) was used to manually delineate the PFC ROI and to overlay the ROIs onto the segmented parametric maps. A trained rater (V.A.) used a standardized protocol based on anatomical landmarks to manually delineate the PFC in five consecutive axial slices using the b = 0 images to avoid bias (see Fig. 1b for a more detailed description). Five slices were chosen to sample the PFC because these slices were the most anatomically consistent between subjects and were minimally affected by EPI distortion.

For each subject, GM and WM means from each PFC ROI were obtained for MK and MD. Directional metric means (axial and radial values) and FA means were taken only from WM segmented maps and not GM maps because GM is nearly isotropic. Statistical analyses were performed using SPSS (v. 17.0.0; Chicago, IL). Although exponential fits from nonlinear regression have been found to best describe developmental trajectories for diffusion metrics between the ages of 5 to 30 years (39), linear regression analysis investigating age effects on each metric was conducted for the TDC and ADHD group due to the narrow age-range of our cohort. Group GM and WM means for each metric were also calculated and compared with two-sample t-tests (2-tailed).


Clinical Group Characteristics

Seven of the 12 adolescents with ADHD met current criteria for combined type ADHD (ADHD-C) and the remaining five for predominantly inattentive type ADHD (ADHD-I). One ADHD-C and one ADHD-I subject were also diagnosed with anxiety disorders (one with Generalized Anxiety Disorder and Social Phobia, one with Obsessive Compulsive Disorder, respectively). Three ADHD-C and two ADHD-I subjects were medication-naïve at the time of scanning and one ADHD-C subject had discontinued stimulants for 1 month prior to scanning after several years of treatment. The remaining ADHD-I (n = 3) and ADHD-C (n = 3) subjects were being treated with methylphenidate or amphetamines for periods ranging from 5 months to 7 years. One adolescent with ADHD-C on a low dose of amphetamine (5 mg/day) was also prescribed 0.5 mg/day risperidone. The two groups did not statistically differ on estimated IQ, age, sex distribution, socioeconomic status distribution (62% of TDC and 75% of ADHD were from the two highest SES classes), or parent-identified ethnicity (Caucasian 38% and 42%, African-American 23% and 17%, Latino 8% and 42%, others including Asian and mixed 31% and 0%, in TDC and ADHD, respectively). Symptom ratings differed significantly, as expected (Table 1).

Prefrontal Cortex ROI Analysis

Linear regression analysis of individual MK means versus age for TDC (Fig. 2a) show significant age-related increases for both GM (r2 = 0.648, P = 0.001, black dots, black regression line, left side black y-axis) and WM (r2 = 0.510, P = 0.006, white dots, gray regression line, right side gray y-axis) with a higher rate of change in the WM (MK white solid bar vs. gray solid bar, Fig. 3). Individuals with ADHD show no significant increase in MK means with age (Fig. 2b) for either GM (r2 < 0.001, P = 0.978) or WM (r2 = 0.135, P = 0.239) as both GM and WM in ADHD have higher MK means around ages 12 to 14 years old compared to TDC and seem to not change significantly through 18 years of age. These trends are also reflected in the linear regression analysis of individual K|| and K[perpendicular] means versus age. In TDC, both the axial and radial WM kurtoses increase significantly with age: r2 = 0.489 for K||, P = 0.008, and r2 = 0.463 for K[perpendicular], P = 0.010 (Table 2), but with about twice the rate of change in the radial direction (K[perpendicular] vs. K|| white solid bars, Fig. 3). In the ADHD group, neither the axial nor the radial kurtoses in WM displayed any significant changes with age: r2 = 0.044 for K||, P = 0.514, and r 2 = 0.185 for K[perpendicular], P = 0.162 (Table 2).

Figure 2
PFC kurtosis and diffusivity means with age. a: TDC individual MK means with age for GM and WM. b: ADHD individual MK means with age for GM and WM. c: TDC individual MD means with age for GM and WM. d: ADHD individual MD means with age for GM and WM. ...
Figure 3
Linear regression slopes of age versus PFC means. TDCs are solid bars and ADHD are checkered bars. Gray bars represent GM and white bars represent WM. Linear regression slopes and their standard errors are shown. MK slope: 1/year, K|| slope: 1/year, K ...
Table 2
Linear Regression of PFC Metric Means With Age

In the TDC group, no significant age-related diffusivity changes were detected with diffusivity in the GM of the PFC (r2 = 0.011 for MD, P = 0.730, Fig. 2c) but all three diffusivity metrics were able to detect significant age-related decreases in means within the WM: r2 = 0.412 for MD, P = 0.018 (Fig. 2c), r2 = 0.430 for D||, P = 0.015 and r2 = 0.368 for D[perpendicular], P = 0.028 (Table 2). Similar rates of change were observed in the axial and radial directions (D|| vs. D[perpendicular] white solid bars, Fig. 3). In the ADHD group, no significant age-related changes were observed in either GM or WM for all diffusivity measures: r2 = 0.311 for MD of GM, P = 0.059 and r2 = 0.059 for MD of WM, P = 0.448 (Fig. 2d), r2 = 0.001 for D|| of WM, P = 0.917 and r2 = 0.121 for D[perpendicular] of WM, P = 0.268 (Table 2).

FA was unable to detect any significant age-related changes in WM means for either TDC or ADHD groups: r2 = 0.244 for TDC FA, P = 0.086, and r2 = 0.191 for ADHD FA, P = 0.156 (Table 2). A summary of all linear regression analyses is shown in Table 2 and all regression slopes with standard errors are summarized in Fig. 3.

Two-sample t-test analysis alone did not detect a statistically significant difference between TDC and ADHD group means for all GM or WM metrics.


Adolescence marks a period of dynamic cognitive and behavioral changes concurrent with maturation of brain regions involved in higher-order functions, particularly in the prefrontal cortex (26). Although the neural alterations are subtle compared to the dramatic changes that occur in the first 10 years of life (39), capturing the intricacies of normal development during this critical period is important for investigating aberrant trajectories in patient populations. Conventional imaging methods have identified little change in total brain volume during adolescence but increases in WM volume and decreases in GM volume (39,40) imply microstructural changes.

In TDC, our regression analyses of kurtosis measures in the PFC from ages 12 to 18 years corroborated these expected changes in microstructure not only in WM but also in GM (Fig. 2a). Increases in WM kurtosis may reflect increased myelination, denser packing of axons and fiber bundles, and changes in axonal membrane permeability, while increases in GM kurtosis may indicate an increase in basal dendrites, synaptic refinement, and changes in cell packing density. These observations are known to occur in associative regions during this period and are correlated with more efficient cognitive functioning (26). The higher rate of WM microstructural complexity changes, depicted in the higher regression slope of WM versus GM (MK white solid bar vs. gray solid bar, Fig. 3), may signify the prominent increase in WM volume and complexity versus the concomitant reduction in GM volume and cortical thickness reported in the normal developmental literature (39). Examining directional kurtosis in WM revealed that significant increases in microstructural complexity occurred in both the axial and radial directions (Table 2). These changes seem to be dominated by a more rapid radial increase (K[perpendicular] vs. K|| white solid bars, Fig. 3) and may correspond to the profuse myelination that is known to occur during this age range (26).

Although reduced volume and delayed cortical thickness trajectories from age 5–18 years have been reported in ADHD (27,29), only two DTI studies have investigated age-related changes in WM microstructure. Silk et al (14) examined whole-brain mean FA from ages 8–18 years and found no significant group difference in that both the ADHD and control groups displayed a significant FA increase with age. The same group also investigated FA in the caudate, putamen, and thalamus within the same age range and reported median FA increased with age in the putamen and thalamus for both groups but only increased significantly in caudate for the ADHD group and not the controls (15). While these studies address the need to examine age-related microstructural changes, their examination of diffusion along all three major directions at the same time makes it difficult to compare their conflicting results with the literature or our findings. Furthermore, the appropriateness of FA for detecting microstructural changes in the primarily GM nuclei of the basal ganglia has not been established.

In our preliminary cross-sectional study we found that, unlike TDC, ADHD subjects failed to show a significant age-related kurtosis increase in either the PFC GM or WM for the mean (Fig. 2b) and the axial and radial directions (Table 2). Given the limited sample sizes and lack of knowledge regarding the effects of psychotropic medication, if any, on diffusion metrics, we can only speculate that this lack of a normal developmental trajectory may indicate an imperfection of neurodevelopmental mechanisms in ADHD (26). Longitudinal reexamination is required to determine what neuronal correlates may explain the higher kurtosis values in early adolescence within the ADHD group as well as determine the relationship between our observations and a report that attainment of peak cortical thickness in children with ADHD is delayed by 3 to 5 years (29).

In contrast to our kurtosis measures in TDC, diffusivity measures were sensitive only to WM age-related changes and failed to detect significant age-related changes in GM (Fig. 2c, Table 2). Although the significant WM decreases in diffusivity agrees with previous DTI studies of typical development (39), the lack of GM diffusivity changes suggests that diffusivity itself may not be the optimal metric to detect age-related microstructural changes in GM during adolescence (29). This highlights an advantage for kurtosis measures. Consistent with the normal developmental DTI literature (39), we identified WM decreases in diffusivity in both the axial and radial directions (Table 2) but with similar rates of change (D|| vs. D[perpendicular] white solid bars, Fig. 3) as opposed to the higher radial rate of change detected by kurtosis measures (K[perpendicular] vs. K|| white solid bars, Fig. 3). This decrease in diffusivity is expected with the formation of new structural barriers to diffusion in WM. Previous reports comparing degree of changes between D|| vs. D[perpendicular] have been inconsistent, ranging from a dominance of one metric to comparable rates (39,40). As opposed to TDC, the ADHD group lacked significant age-related changes in all of the WM diffusivity measures (Table 2), supporting the preliminary conclusion from both diffusion and kurtosis measures that the WM developmental trajectory is aberrant in ADHD. The negative results of the two-sample t-tests between group means for all metrics were not surprising considering the extent to which these measures dynamically change with age in TDC (Fig. 2).

Several DTI studies have shown that FA increases with age in typically developing children and adolescents (39,40). Similar to Makris et al’s (12) voxel-based results our FA mean analysis was unable to detect any significant age-related changes in WM of either group (Table 2). In contrast to other DTI studies (1,615), our group FA means also did not statistically differ. Many factors could explain these inconsistencies, including differences in methodology, cohort age range, and sample size. All previous studies calculated FA maps using fewer b-values and encoding directions than the present study, and all ROI DTI studies used much smaller, tract-based ROIs to compute group FA means. Furthermore, the measurement of FA is sensitive to partial volume effects of CSF that can make comparisons using different methodologies problematic (41).

In conclusion, this is the first study applying DKI to assess age effects on GM and WM microstructural changes in TDC and adolescents with ADHD. Our preliminary results of typical developmental changes in the PFC support the prevailing theory that ADHD is a disorder affecting WM of the frontostriatal circuit (14). Whereas TDC show a significant increase of WM tissue complexity from 12 to 18 years of age, dominated by a higher rate of change in the radial direction, WM microstructure in ADHD is stagnant. Directional diffusional and kurtosis analysis confirmed that this lack of typical age-related change occurs in both the axial and radial directions. Moreover, our study is the first to directly quantify an aberrant age-related trajectory in ADHD within GM microstructure; only kurtosis measures were able to detect this difference between TDC and adolescents with ADHD. These results suggest that the assessment of non-Gaussian directional diffusion using DKI may provide more sensitive metrics of changes in tissue microstructure than conventional DTI. Naturally, these results must be interpreted with caution because of the cross-sectional design, mixed ADHD clinical subtypes, possible confounding medications, and modest sample sizes. A longitudinal study with larger sample sizes will be needed to better characterize these potential developmental abnormalities in ADHD.


Contract grant sponsor: Litwin Foundation (to J.A.H.); Contract grant sponsor: National Institutes of Health (NIH); Contract grant numbers: 1R01AG027852 and 1R01EB007656 (to J.A.H.).


1. Ashtari M, Kumra S, Bhaskar SL, et al. Attention-deficit/hyperactivity disorder: a preliminary diffusion tensor imaging study. Biol Psychiatry. 2005;57:448–455. [PubMed]
2. Krain AL, Castellanos FX. Brain development and ADHD. Clin Psychol Rev. 2006;26:433–444. [PubMed]
3. Paloyelis Y, Mehta MA, Kuntsi J, Asherson P. Functional MRI in ADHD: a systematic literature review. Expert Rev Neurother. 2007;7:1337–1356. [PMC free article] [PubMed]
4. Valera EM, Faraone SV, Murray KE, Seidman LJ. Meta-analysis of structural imaging findings in attention-deficit/hyperactivity disorder. Biol Psychiatry. 2007;61:1361–1369. [PubMed]
5. Zang YF, Jin Z, Weng XC, et al. Functional MRI in attention-deficit hyperactivity disorder: evidence for hypofrontality. Brain Dev. 2005;27:544–550. [PubMed]
6. Bechtel N, Kobel M, Penner IK, et al. Decreased fractional anisotropy in the middle cerebellar peduncle in children with epilepsy and/or attention deficit/hyperactivity disorder: a preliminary study. Epilepsy Behav. 2009;15:294–298. [PubMed]
7. Cao Q, Sun L, Gong G, et al. The macrostructural and microstructural abnormalities of corpus callosum in children with attention deficit/hyperactivity disorder: a combined morphometric and diffusion tensor MRI study. Brain Res. 2010;1310:172–180. [PubMed]
8. Casey BJ, Epstein JN, Buhle J, et al. Frontostriatal connectivity and its role in cognitive control in parent-child dyads with ADHD. Am J Psychiatry. 2007;164:1729–1736. [PubMed]
9. Chao TC, Chou MC, Yang P, Chung HW, Wu MT. Effects of interpolation methods in spatial normalization of diffusion tensor imaging data on group comparison of fractional isotropy. Magn Reson Imaging. 2008;27:681–690. [PubMed]
10. Davenport ND, Karatekin C, White T, Lim KO. Differential fractional anisotropy abnormalities in adolescents with ADHD or schizophrenia. Psychiatry Res. 2010;181:193–198. [PMC free article] [PubMed]
11. Hamilton LS, Levitt JG, O’Neill J, et al. Reduced WM integrity in attention-deficit hyperactivity disorder. Neuroreport. 2008;19:1705–1708. [PMC free article] [PubMed]
12. Makris N, Buka SL, Biederman J, et al. Attention and executive systems abnormalities in adults with childhood ADHD: a DT-MRI study of connections. Cereb Cortex. 2008;18:1210–1220. [PubMed]
13. Pavuluri MN, Yang S, Kamineni K, et al. Diffusion tensor imaging study of white matter fiber tracts in pediatric bipolar disorder and attention-deficit/hyperactivity disorder. Biol Psychiatry. 2008;65:586–593. [PMC free article] [PubMed]
14. Silk TJ, Vance A, Rinehart N, Bradshaw JL, Cunnington R. White-matter abnormalities in attention deficit hyperactivity disorder: a diffusion tensor imaging study. Hum Brain Mapp. 2009;30:2757–2765. [PubMed]
15. Silk TJ, Vance A, Rinehart N, Bradshaw JL, Cunnington R. Structural development of the basal ganglia in attention deficit hyper-activity disorder: a diffusion tensor imaging study. Psychiatry Res. 2009;172:220–225. [PubMed]
16. Basser PJ, Mattiello J, Le Bihan D. Estimation of the effective self-diffusion tensor from the NMR spin echo. J Magn Reson. 1994;103:247–254. [PubMed]
17. Basser PJ, Mattiello J, Le Bihan D. MR diffusion tensor spectroscopy and imaging. Biophys J. 1994;66:259–267. [PubMed]
18. Basser PJ. Inferring microstructural features and the physiological state of tissues from diffusion-weighted images. NMR Biomed. 1995;8:333–344. [PubMed]
19. Jensen JH, Helpern JA. In: Jensen JH, Helpern JA, editors. Quantifying non-Gaussian water diffusion by means of pulsed-field-gradient MRI; Proc 11th Annual Meeting ISMRM; Toronto. 2003. pp. 19pp. 2154–2161.
20. Jensen JH, Helpern JA, Ramani A, Lu H, Kaczynski K. Diffusional kurtosis imaging: the quantification of non-gaussian water diffusion by means of magnetic resonance imaging. Magn Reson Med. 2005;53:1432–1440. [PubMed]
21. Lu H, Jensen JH, Ramani A, Helpern JA. Three-dimensional characterization of non-gaussian water diffusion in humans using diffusion kurtosis imaging. NMR Biomed. 2006;19:236–247. [PubMed]
22. Jensen JH, Helpern JA. MRI quantification of non-Gaussian water diffusion by kurtosis analysis. NMR Biomed. 2010 doi: 10.1002/nbm.1518. [Epub ahead of print] [PMC free article] [PubMed] [Cross Ref]
23. Cheung MM, Hui ES, Chan KC, Helpern JA, Qi L, Wu EX. Does diffusion kurtosis imaging lead to better neural tissue characterization? A rodent brain maturation study. Neuroimage. 2009;45:386–392. [PubMed]
24. Hui ES, Cheung MM, Qi L, Wu EX. Toward better MR characterization of neural tissues using directional diffusion kurtosis analysis. Neuroimage. 2008;42:122–134. [PubMed]
25. Song SK, Sun SW, Ramsbottom MJ, Chang C, Russell J, Cross AH. Dysmyelination revealed through MRI as increased radial (but unchanged axial) diffusion of water. Neuroimage. 2002;17:1429–1436. [PubMed]
26. Huttenlocher PR, Dabholkar AS. Regional differences in synapto-genesis in human cerebral cortex. J Comp Neurol. 1997;387:167–178. [PubMed]
27. Castellanos FX, Lee PP, Sharp W, et al. Developmental trajectories of brain volume abnormalities in children and adolescents with attention-deficit/hyperactivity disorder. JAMA. 2002;288:1740–1748. [PubMed]
28. Mostofsky SH, Cooper KL, Kates WR, Denckla MB, Kaufmann WE. Smaller prefrontal and premotor volumes in boys with attention-deficit/hyperactivity disorder. Biol Psychiatry. 2002;52:785–794. [PubMed]
29. Shaw P, Eckstrand K, Sharp W, et al. Attention-deficit/hyperactivity disorder is characterized by a delay in cortical maturation. Proc Natl Acad Sci U S A. 2007;104:19649–19654. [PubMed]
30. Poot DH, den Dekker AJ, Achten E, Verhoye M, Sijbers J. Optimal experimental design for diffusion kurtosis imaging. IEEE Trans Med Imaging. 2010;29:819–829. [PubMed]
31. Wechsler D. Wechsler abbreviated scale of intelligence (WASI) San Antonio, TX: Psychological Corporation; 1999.
32. American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4. Washington, DC: American Psychiatric Association; 2000. text revision.
33. Kaufman J, Birmaher B, Brent D, Rao U, Ryan N. Diagnostic interview. Kiddie-sads-present and lifetime version (K-SADS-PL) — screen interview. 1996
34. Conners CK. Conners’ rating scales-revised technical/user’s manual. North Tonawanda, NY: Multi-Health Systems; 1997a/b.
35. Hollingshead AB. Four factor index of social status. New Haven, CT: Yale University, Dept. of Sociology; 1975.
36. Tabesh A, Jensen JH, Ardekani BA, Helpern JA. Estimation of tensors and tensor-derived measures in diffusional kurtosis imaging. Magn Reson Med. 2011 in press. [PMC free article] [PubMed]
37. Collignon A, Maes F, Delaere D, Vandermeulen D, Suetens P, Marchal G. Automated multi-modality image registration based on information theory. In: Bizais Y, Barillot C, Di Paola R, editors. Proc Information Processing in Medical Imaging. Dordrecht, The Netherlands: Kluwer Academic Publishers; 1995. pp. 263–274.
38. Ashburner J, Friston KJ. Unified segmentation. Neuroimage. 2005;26:839–851. [PubMed]
39. Lebel C, Walker L, Leemans A, Phillips L, Beaulieu C. Microstructural maturation of the human brain from childhood to adulthood. Neuroimage. 2008;40:1044–1055. [PubMed]
40. Giorgio A, Watkins KE, Chadwick M, et al. Longitudinal changes in grey and white matter during adolescence. Neuroimage. 2010;49:94–103. [PubMed]
41. Ma X, Kadah YM, LaConte SM, Hu X. Enhancing measured diffusion anisotropy in gray matter by eliminating CSF contamination with FLAIR. Magn Reson Med. 2004;51:423–427. [PubMed]