Diffusion tensor imaging in children and adolescents: reproducibility, hemispheric, and age-related differences

Abstract

We evaluated intra-rater, inter-rater, and between-scan reproducibility, hemispheric differences, and the effect of age on apparent diffusion coefficient (ADC) and fractional anisotropy (FA) in healthy children (age range 5.5-19.1 years) examined with a clinical diffusion tensor imaging (DTI) protocol at 1.5 T, using a region of interest (ROI) methodology. Measures of reliability and precision were assessed in six ROIs using two different ROI shapes (polygonal and ellipsoidal).

Results: Highly reproducible values of ADC and FA were obtained with the polygonal method on intra-rater (coefficients of variation<or=2.7%) and inter-rater (coefficients of variation<or=4.8%) reproducibility. For between-scan reproducibility, the coefficients of variation were <or=5.0%. Mean asymmetry indices were in the range from -4% to 9% for FA and from -6% to 3% for ADC. ADC showed significant negative correlation with age in 13 of 15 examined fiber tracts and FA increased significantly in three fiber tracts. Our results show that the evaluated DTI protocol is suitable for clinical application in pediatric population.

Figure 1. ROIs for study of age-related changes

Fifteen regions of interest used to study age-related changes in DTI parameters overlaid on color map slices. CP (1): cerebral peduncle, TWM (2): temporal white matter, FWM (3): frontal white matter, ALIC (4): anterior limb of the internal capsule, PLIC (5): posterior limb of the internal capsule, GCC (6): genu, SCC (7): splenium, BCC (8): body of the corpus callosum, AWM (9): anterior white matter, TOWM (10): temporo-occipital white matter, SLF (11): superior longitudinal fasciculus, SCR (12): superior corona radiata, SFOF (13): superior fronto-occipital fasciculus, CING (14): cingulum, CSO (15): centrum semiovale.

Figure 2. ROIs for reproducibility analysis

Upper panel: Regions of interest (ROIs) are shown overlaid on the color map; polygonal ROIs are displayed on the left side, ellipsoid ROIs on the right side. Lower panel: The location of the color map is shown on the corresponding T2 weighted image. The following ROIs were evaluated: a) from top to bottom: genu of the corpus callosum (6), anterior limb of the internal capsule (4), posterior limb of the internal capsule (5), b) superior corona radiata (12), c) cerebral peduncle (1), d) cingulum (14). Numbers in brackets relate these ROIs to the 15 ROIs in Figure 1.

Figure 3. Scan – rescan - comparison

Color-coded FA maps of initial scan (left panel) and repeated scan (right panel, 3 days later) from a 17.9 year old male. Of the 22 acquired slices, the four slices used in the ROI analysis are shown; slices depict a) cingulum b) superior corona radiata c) anterior and posterior limb of the internal capsule, genu of the corpus callosum, d) cerebral peduncle.

Figure 4. Intra-, inter-rater and between-scan reproducibility of measurements of ADC and FA

Coefficients of variation (CV₁, see text) are depicted as boxplots, N=20, pooled data for right and left measurements. Data are shown for each tract (abscissa); ROIs are identified in the caption of Figure 1. Separate diagrams are shown for the ellipsoid ROI method (upper half) and the polygonal method (lower half). ADC measurements are shown in the left half of the diagram, FA measurements in the right half of the diagram. Intra-rater and inter-rater reproducibility are shown for the ellipsoid ROI approach, intra-rater, inter-rater and between-scan CV₁s are shown for the polygonal ROI approach. The polygonal ROI method has significant better reproducibility (lower CV₁ values) than the ellipsoid approach, both for ADC and FA and for intra- and inter-rater measurements. The intra-rater measurements have lower CV₁ values than those between different operators. The between-scan reproducibility analysis yields significantly higher CV₁ for ADC values when compared to intra- (p<.001) and inter-rater (p=.043) results. FA reproducibility was significantly better in the intra-rater analysis (p<.001), but not in the inter-rater analysis (p=.084), when compared to the between-scan analysis. ADC generally shows better reproducibility than FA.

Figure 5. Asymmetry indices

Mean values and standard deviations of asymmetry indices calculated for each region. Regions are represented by numbers as given in the caption of Figure 1. Asterisk (*) indicates P<0.05 after Bonferroni-correction.

Figure 6. Age-related differences

Scatter plots for all 15 examined fiber tracts of age (abscissa) and DTI parameter (ordinate). FA (scaled by a factor of 1000); ADC (μm²/s). ROIs are identified as given in the caption of Figure 1. Asterisk (*) indicates significant correlations.