Table 6.
DTI-based studying of the ASD at childhood.
| Work | Autistic group | Control group | sMRI | Brain regions | Data analysis | Findings | ||||
|---|---|---|---|---|---|---|---|---|---|---|
| Age, y | IQ | Size | Age, y | IQ | Size | |||||
| Sundaram et al., 2008 | 2–7 | – | 50 AD, ASP, NPDD | 3–11 | – | 16 | 3 T; 3 mm; 1000 b; 6 GD | Frontal lobe's long-/short-range association fibers | DT; ROI; MANCOVA; DTI studio | ASD: Increased MD in short/long-range fibers and decreased FA in short-range fibers; insignificant negative correlation between FA and GARS AQ (social isolation subscale) |
| Brito et al., 2009 | 8–12 | – | 8 AD | 8–12 | – | 8 | 1.5 T; 5 mm; 1000 b; 12 GD | Frontopontine and corticospinal tracts; frontal subcortical WM | ROI; ANOVA; EpiInfo | ASD: Decreased FA in the anterior CC, right CST, posterior limb of right and left ICs, left superior cerebellar peduncle, and right and left middle cerebellar peduncles |
| Cheung et al., 2009 | 7–12 | 78–122 | 13 AD | 7–13 | 92–132 | 14 | 1.5 T; 5 mm; 1200 b; 25 GD | Left orbitofrontal cortex; precentral gyrus; bilateral frontal pole | VBA; GLM; Linear regression; SPM2; SPSS15.0 | ASD: Decreased FA in prefrontal lobes, ventral and middle temporal lobe, and cerebellum; increased FA in superior longitudinal fasc and left occipital lobe; strongly correlated higher ADI-R scores and lower FA in prefrontal lobes and ventral temporal lobes; negatively correlated communication and social reciprocity impairments (ADI-A and ADI-B) throughout fronto-striato-temporal pathways and posterior CC |
| Ke et al. (2009) | 6–11 | 82–120 | 12 HFA | 7–12 | 82–118 | 10 | 1.5 T; 3 mm; 1000 b; 15 GD | Bilateral middle frontal gyrus; left inferior frontal gyrus; left superior temporal gyrus; right middle temporal gyrus; right frontal lobe | VBA; WM density; SPM5 | ASD: Decreased WM density in right frontal lobe, left parietal lobe and right anterior cingulate; increased WM density in right frontal lobe, left parietal lobe and left cingulate gyrus; decreased FA in the frontal lobe and left temporal lobe; positively correlated CARS score and FA in right frontal lobe; no significant correlation between ADI-R scores and mean FA |
| Sivaswamy et al., 2010 | 2–9 | – | 27 AD, ASP, NPDD | 2–9 | – | 16 | 3 T; 3 mm; 1000 b; 6 GD | Superior, middle, and inferior cerebellar peduncles tracts | ROI; ANCOVA; SPSS17.0 | ASD: Increased MD of bilateral superior cerebellar peduncles; increased FA of right middle cerebellar peduncle; reversed FA asymmetry pattern in middle and inferior cerebellar peduncles |
| Kumar et al., 2010 | 2–9 | – | 32 AD, ASP, NPDD | 2–9 | – | 16 | 3 T; 3 mm; 1000 b; 6 GD | Uncinate fasc.; inferior fronto-occipital fasc.; arcuate fasc.; CC; CST | TBSS; DT; ANOVA; DTI studio 2.40 | ASD and DevI: Decreased FA in right uncinate fasc., right cingulum, and CC; increased MD in right arcuate fasc.; DevI: Decreased FA in bilateral fronto-occipital fasc |
| Barnea-Goraly et al., 2010 | 9–14 | 69–103 | 13 AD | 8–12 | 107–123 | 11 | 1.5 T; 3 mm; 900 b; 6 GD | Medial prefrontal WM; frontal corona radiate; CC genu, anterior forceps, and body | VBA; FSL | ASD and siblings: Decreased FA and λ|| in multiple regions across frontal, temporal, and parietal lobes; no significant differences in WM structure; ASD: No significant correlation between ADOS and ADI-R subscale scores and FA or λ|| |
| Weinstein et al., 2011 | 2–4 | – | 22 AD | 2–5 | – | 32 | 1.5 T; 3 mm; 1000 b; 15 GD | CC genu and body; left superior longitudinal fasc.; right and left cingulum | TBSS; DT; MRI studio; FSL | ASD: Increased FA and decreased λ⊥ in CC genu and body, left superior longitudinal fasc., and bilateral cingulum |
| Hong et al., 2011 | 7–11 | 84–126 | 18 HFA | 8–12 | 86–136 | 16 | 1.5 T; 2 mm; 1000 b; 15 GD | CC anterior third, anterior and posterior midbody, isthmus; and splenium | DT; ROI; WM density and volume; FSL; SPSS13.0 | ASD: Decreased WM density in CC anterior third; increased MD and decreased fiber number in anterior third transcallosal fiber tracts; no significant correlation between DTI indices and CARS |
| Ingalhalikar et al., 2011 | 8–13 | – | 45 HFA | 8–13 | – | 30 | 3 T; 2 mm; 1000 b; 30 GD | Middle occipital gyrus left; inferior occipital WM right; superior temporal WM right fornix | Classification; ROI; non-linear SVM | 84% specificity and 74% sensitivity by LOO CV based on FA in right occipital regions, left superior longitudinal fasc., EC; IC; MD in right occipital gyrus and right temporal WM; correlated SRS/SCQ autism scores and classification results |
| Cheon et al., 2011 | 9–13 | 100–124 | 17 ASP; NPDD | 8–12 | 103–125 | 17 | 1.5 T; 3 mm; 900 b; 30 GD | Anterior thalamic radiation; superior thalamic radiation; inferior longitudinal fasc | TBSS; ROI; ANOVA; DTIStudio; FSL; SPSS 11.5 | ASD: Decreased FA and increased MD in right anterior thalamic radiation, CC, and left uncinate fasc.; decreased FA in left anterior thalamic radiation, and right and left inferior longitudinal fasc.; increased λ⊥ in right and left anterior thalamic radiation, CC, left uncinate fasc., and left inferior longitudinal fasc.; decreased λ|| in left inferior longitudinal fasc.; negatively correlated SRS and FA in right anterior thalamic radiations and right uncinate fasc |
| Jeong et al., 2011 | 3–7 | – | 32 | 4–8 | – | 14 | 3 T; 3 mm; 1000 b; 6 GD | Bilateral arcuate fasc.; bilateral uncinate fasc.; CC genu. | DT; ROI; TBSS; FSL; SPM | ASD: Increased curvature and λ⊥ and decreased FA in parietotemporal junction for arcuate fasc., frontotemporal junction for uncinate fasc., and midline of CC genu; positively correlated curvature and λ⊥ in all ROIs; Controls and ASD: Negatively correlated curvature and FA in all ROIs; Controls: Positively correlated curvature and λ⊥ only in CC genu |
| Jou et al., 2011a | 7–15 | – | 15 AD | 9–14 | – | 8 | 3 T; 2.5 mm; 30 GD | Inferior fronto-occipital fasc.; superior longitudinal fasc.; uncinate fasc.; cingulum. | TBSS; FSL; SPSS17.0 | ASD: Decreased FA in numerous association, commissural, and projection tracts, especially, the forceps minor, fronto-occipital fasc., and superior longitudinal fasc.; no significant correlation between FA and SRS scores |
| Wan et al., 2012 | 6–8 | – | 5 LFA | 9–14 | – | 5 | 3 T; 1.5 mm; 1000 b; 25 GD | Arcuate fasc | PT; ROI; FSL | Nonverbal ASD: no usual leftward pattern of arcuate fasc. asymmetry (actually the reversed pattern for four out of the five nonverbal subjects) |
| Walker et al., 2012 | 3–7 | – | 39 AD | 3–7 | – | 39 AD | 1.5 T; 2.5 mm; 1100 b; 50 GD | Cerebellum; CC genu, body and splenium; CSTs; pons | TBSS; RESTORE | ASD: Decreased FA in various WM regions; ncreased MD in posterior brain regions Significant age group interaction, indicating differences in FA and MD developmental trends between the two groups |
| Nagae et al., 2012 | 7–18 ASD/-LI: 7–15 ASD/+LI | – | 18 ASD/-LI; 17 ASD/+LI | 7-18 | – | 25 | 3T; 2mm; 1000 b; 30 GD | Superior longitudinal fasc.; temporal lobe component of the superior longitudinal fasc.; CSTs | DT; GLM; HRL; DTIStudio; SPSS 19.0 | ASD/-LI: Increased MD in CSTs ASD/+LI: Increased MD in left hemisphere superior longitudinal fasc fiber tracts and temporal portion of superior longitudinal fasc ASD/±LI: Significant negative correlation between left hemisphere superior longitudinal fasc MD and clinical language ability assessment |
| Poustka et al., 2012 | 8–12 | 97–125 | 18 AD | 8–12 | 98–128 | 18 | 1.5 T; 2 mm; 1000 b; 6 GD | Fornix; superior longitudinal fasc.; uncinate fasc.; CC | VBA; DT; SPM5; NeuroQlab | ASD: Decreased FA in bilateral uncinate fasc and right superior longitudinal fasc.; negative correlation between FA of affected fiber tracts and autism communication and interaction scores (ADI-R and ADOS); no significant altered GM or WM concentration after correction for multiple comparisons |
| Lai et al., 2012 | 7–15 | – | 16 LFA | 7–16 | – | 18 | 1.5 T; 5 mm; 1000 b; 25 GD | Arcuate and inferior fronto-occipital fasc | PT; t-test FSL4.1 | ASD: Decreased FA in left part of left dorsal pathway; significantly lower tensor norms for ventral tract |
| Billeci et al., 2012 | 4–8 | 47–93 | 22 AD, NPDD | 3–8 | 90–108 | 10 | 1.5 T; 3 mm; 1000 b; 25 GD | CC; cingulum; arcuate fasc | TBSS; DT; ANCOVA; FSL; BET; FNIRT; eDTI; SPSS | ASD: Increased FA in major WM pathways, especially in CC, cingulum, arcuate fasc and IC; increased fiber length and FA in cingulum and CC, and increased MD in indirect segments of right arcuate and the cingulum; significant correlation between MD of arcuate fasc., CC, and cingulum and expressive language abilities |
| Mills et al., 2013 | 8–11 | VIQ: 77–117 | 10 HFA | 7–11 | 109–133 | 17 | 1.5 T; 2.5 mm; 1000 b; 51 GD | Superior longitudinal fasc.; temporal and partial subsections of superior longitudinal fasc.; inferior fronto-occipital fasc | DT; ROI; FreeSurfer; DTIStudio | ASD: Increased MD and λ⊥ in right inferior longitudinal fasc Positive correlation between higher FA and lower MD and λ⊥ in inferior longitudinal fasc produced more morphologically accurate language |
| Duerden et al., 2014 | 8–13 | 86–122 | 30 AD | 8–13 | 100–128 | 30 | 1.5 T; 3 mm; 1000 b; 35 GD | Cingulum bundle; posterior limb of IC; corona radiata | ROI; TBSS; MANCOVA; Camino; FLIRT | ASD and self-injury subjects: Decreased FA and Increased MD in left posterior limb of ICs Positive correlation between self-injury and increased λ⊥ in bilateral posterior limbs of IC and corona radiata. |
| Joseph et al., 2014 | 4–6 | VIQ: 73–119 | 20 AD | 8–11 | VIQ: 102–132 | 20 | 3 T; 2 mm; 1000 b; 15 GD | Arcuate fasc | PT; ANOVA; FSL4.1.2 | Decreased leftward/increased rightward asymmetry of pars opercularis correlated with higher language ability and bilaterally increased FA and decreased λ⊥ of the arcuate fasc |
| Peterson et al., 2015 | 9–12 | 86–118 | 36 HFA | 9–12 | 100–118 | 37 | 3 T; 2.2 mm; 800 b; 32 GD | Left and right hemispheric WM regions | ROI; ANCOVA; CATNAP; RESTORE; LDDMM | ASD: Increased MD throughout left hemisphere, particularly in outer-zone cortical WM Controls and ASD: Negative correlation between MD and age in left-hemisphere WM regions |
Additional abbreviations: b, b-value (s/mm2); f, females; fasc., fasciculus; m, males; Rf, reference;
T, MRI magnetic field strength (tesla); y, year;
ADI-R, autism diagnostic interview - revised; ADOS, autism diagnistic observation schedule;
ANCOVA, analysis of covariances; ANOVA, analysis of variances;
AQ, autistic quotient; CARS, childhood autism rating scale; CC, corpus callossum;
CV, cross-validation; CST, cortico-spinal tract; DevI, developmentally impaired;
DT, deterministic tractography; EC, external capsule; FWE, family-wise error;
GARS, Gilliam autism rating scale; GD, gradient directions; GLM, general linear model;
HRL, hierarchic regression model; IC, internal capsule;
LDDMM, large deformation diffeomorphic metric mapping; LOO, leave-one-out; ±LI, with/without LI;
MANCOVA, multivariate analysis of covariances; MANOVA, multivariate analysis of variances;
PT, probabilistic tractography;
RCLGCM, random coefficient linear growth curve model;
SCQ, social communication questionnaire; SLI, specific LI; SNR, signal-to-noise ratio;
SPM, statistical parametric mapping; SRS, social responsiveness scale; SVM, support vector machine;
TBSS, tract-based spatial statistics;
3D Slicer, BET, CATNAP, DTIprep, DTI studio;
eDTI, FiberViewer, FLIRT, FNIRT, FSL, MRI studio, NeuroQlab, RESTORE;
SAS, SPM, SPSS, data processing packages.