Diffusion-weighted magnetic resonance imaging of the brain of neurologically normal dogs

Megan J. MacLellan Department of Clinical Sciences, College of Veterinary Medicine, University of Minnesota, Saint Paul, MN 55108.

Search for other papers by Megan J. MacLellan in
Current site
Google Scholar
PubMed
Close
 DVM, MS
,
Christopher P. Ober Department of Clinical Sciences, College of Veterinary Medicine, University of Minnesota, Saint Paul, MN 55108.

Search for other papers by Christopher P. Ober in
Current site
Google Scholar
PubMed
Close
 DVM, PHD
,
Daniel A. Feeney Department of Clinical Sciences, College of Veterinary Medicine, University of Minnesota, Saint Paul, MN 55108.

Search for other papers by Daniel A. Feeney in
Current site
Google Scholar
PubMed
Close
 DVM
, and
Carl R. Jessen Department of Clinical Sciences, College of Veterinary Medicine, University of Minnesota, Saint Paul, MN 55108.

Search for other papers by Carl R. Jessen in
Current site
Google Scholar
PubMed
Close
 DVM, PHD

Abstract

OBJECTIVE To acquire MRI diffusion data (apparent diffusion coefficient [ADC] and fractional anisotropy [FA] values, including separate measures for gray and white matter) at 3.0 T for multiple locations of the brain of neurologically normal dogs.

ANIMALS: 13 neurologically normal dogs recruited from a group of patients undergoing tibial plateau leveling osteotomy.

PROCEDURES: MRI duration ranged from 20 to 30 minutes, including obtaining preliminary images to exclude pathological changes (T2-weighted fluid-attenuated inversion recovery transverse and dorsal images) and diffusion-weighted images.,

RESULTS: Globally, there were significant differences between mean values for gray and white matter in the cerebral lobes and cerebellum for ADC (range of means for gray matter, 0.8349 × 10−3 s/mm2 to 0.9273 × 10−3 s/mm2; range of means for white matter, 0.6897 × 10−3 s/mm2 to 0.7332 × 10−3 s/mm2) and FA (range of means for gray matter, 0.1978 to 0.2364; range of means for white matter, 0.5136 to 0.6144). These values also differed among cerebral lobes. In most areas, a positive correlation was detected between ADC values and patient age but not between FA values and patient age.

CONCLUSIONS AND CLINICAL RELEVANCE: Cerebral interlobar and cerebellar diffusion values differed significantly, especially in the gray matter. Information about diffusion values in neurologically normal dogs may be used to diagnose and monitor abnormalities and was the first step in determining the clinical use of diffusion imaging. This information provided an important starting point for the clinical application of diffusion imaging of the canine brain.

Abstract

OBJECTIVE To acquire MRI diffusion data (apparent diffusion coefficient [ADC] and fractional anisotropy [FA] values, including separate measures for gray and white matter) at 3.0 T for multiple locations of the brain of neurologically normal dogs.

ANIMALS: 13 neurologically normal dogs recruited from a group of patients undergoing tibial plateau leveling osteotomy.

PROCEDURES: MRI duration ranged from 20 to 30 minutes, including obtaining preliminary images to exclude pathological changes (T2-weighted fluid-attenuated inversion recovery transverse and dorsal images) and diffusion-weighted images.,

RESULTS: Globally, there were significant differences between mean values for gray and white matter in the cerebral lobes and cerebellum for ADC (range of means for gray matter, 0.8349 × 10−3 s/mm2 to 0.9273 × 10−3 s/mm2; range of means for white matter, 0.6897 × 10−3 s/mm2 to 0.7332 × 10−3 s/mm2) and FA (range of means for gray matter, 0.1978 to 0.2364; range of means for white matter, 0.5136 to 0.6144). These values also differed among cerebral lobes. In most areas, a positive correlation was detected between ADC values and patient age but not between FA values and patient age.

CONCLUSIONS AND CLINICAL RELEVANCE: Cerebral interlobar and cerebellar diffusion values differed significantly, especially in the gray matter. Information about diffusion values in neurologically normal dogs may be used to diagnose and monitor abnormalities and was the first step in determining the clinical use of diffusion imaging. This information provided an important starting point for the clinical application of diffusion imaging of the canine brain.

Magnetic resonance imaging is an essential component in the evaluation of intracranial disease in human and veterinary medicine for the purposes of making a diagnosis, planning and performing treatment, and monitoring response to treatment.1 However, specificity of conventional MRI can be limited because there is considerable overlap in the imaging characteristics of various disease processes.2–4 Advances in MRI, such as diffusion-weighted imaging, have expanded the morphology-based imaging of MRI to include functional, cellular, metabolic, cytoarchitectural, and hemodynamic information that can aid in the diagnosis, grading, therapeutic planning, and monitoring of brain tumors and other intracranial conditions in humans.1

Diffusion imaging is a physiology-based type of MRI and uses differences in water movement through tissues to create contrast between tissues.1–3,5,6 In diffusion-weighted MRI, the locations of hydrogen atoms in water are phase encoded by 2 separate strong diffusion gradient pulses3,6–8; hydrogen atoms that move or diffuse between sequential pulses result in a signal void, whereas those with restricted diffusion result in a positive signal.1,3,6,8 Several quantitative values (eg, ADC and FA) can be acquired (mapped and calculated) from the obtained images.5,6,8,9 The ADC provides a numeric representation of unrestricted and random (referred to as isotropic) water diffusivity in each voxel; thus, it is a measure of the overall magnitude of water diffusion.2–8 The ADC requires at least 2 gradient signal intensities with separate b values. The b value is the sensitivity to diffusion, and a value of b = 0 s/mm2 is essentially a T2-weighted image. Typically, b values of 0 and 1,000 s/mm2 are used.3,6,7 The FA measures the anisotropic water diffusivity, which is the directionality of water diffusion attributable to the effects of surrounding tissues or cells.3,8,9 The main reason to calculate FA is to determine the integrity of brain parenchyma, specifically white matter tracts, in the vicinity of an intracranial lesion.1,2,8,9 The FA values have no units and range from 0 (complete isotropic diffusion) to 1 (complete anisotropic or highly directional diffusion).3,7,9

The main clinical applications for diffusion imaging in humans with intracranial disease include detecting ischemic infarcts, differentiating cytotoxic from vasogenic edema, assessing tumor grade and cellularity, examining postoperative injury, and evaluating integrity of white matter tracts.1,5–9 More recently, investigations have been conducted to determine whether diffusion imaging can be used with more specific diffuse diseases and those without a clear anatomic abnormality on other MRI sequences; these diseases include epilepsy, Alzheimer disease, multiple sclerosis, and Parkinson disease.2,10

Without a determination of values for clinically normal animals, the clinical diagnostic value of diffusion imaging is limited, and diffuse disease may be undiagnosed. Minimal information is available on values for clinically normal animals, and reports often do not include both ADC and FA measurements, nor do they separate values for gray and white matter.2,6,9 Diffusion values in clinically normal animals were measured in 2 recent studies.2,9 In 1 study,2 investigators evaluated ADC values (obtained at 1 T) of the brain of clinically normal dogs, but they did not separate gray matter from white matter or calculate FA values; evaluated only a few regions in the brain; and used only 1 Beagle. In the other study,9 investigators evaluated both ADC and FA values, although they also did not separate gray matter from white matter and evaluated only a few regions in the brain.

The purpose of the study reported here was to determine ADC and FA values at 3.0 T in multiple locations of the brain (including separate measures for gray matter and white matter) of neurologically normal dogs. We hypothesized that ADC and FA values for each location and tissue type would not differ on the basis of patient size or between right and left hemispheres; however, we also hypothesized that diffusion values would be affected by patient age because an increase in ADC has been detected with an increase in age of humans.11 Furthermore, we hypothesized that the white matter would have higher FA and lower ADC values, compared with values for the gray matter, and that there would be variation in diffusion values among brain lobes and regions. Finally, we hypothesized that there would be no significant interobserver differences in diffusion values.

Materials and Methods

Animals

Thirteen dogs were prospectively recruited for use in the study. For inclusion in the study, subjects had to be neurologically normal dogs undergoing tibial plateau leveling osteotomy. These dogs were selected on the basis that they were already undergoing an anesthetic procedure and that MRI would add < 1 hour to the procedure and would pose a minimal additional risk to each patient. Owner consent was obtained for use of each dog. Dogs were managed in accordance with a protocol approved by an institutional care and use committee.

Experimental procedures

Dogs were anesthetized for the tibial plateau leveling osteotomy. Because all dogs were clinical patients, anesthetic protocols were selected on a case-by-case basis by members of our anesthesiology service. Each dog was placed in sternal recumbency, and an MRI of the brain was obtained by use of a quadrature knee coil.a Then, T2-weighted FLAIR sequences were obtained for the transverse (TE, 120.6 milliseconds; TR, 8,002 milliseconds) and dorsal (TE, 126.5 milliseconds; TR, 8,202 milliseconds) planes. The T2-weighted FLAIR images were evaluated for morphological abnormalities; any dog with morphological abnormalities evident on T2-weighted images was excluded from the study.

Diffusion images, including diffusion-weighted (TE, 89.4 milliseconds; TR, 10,000 milliseconds) and diffusion tensor (TE, 86.5 milliseconds; TR, 10,000 milliseconds) sequences, were then obtained for the dorsal plane. Diffusion tensor sequences were obtained by use of 25 imaging directions with a matrix size of 256 × 256, slice thickness of 2.4 mm with no slice gap, echo train length of 1, and 2 excitations. Diffusion-weighted imaging sequences were obtained by use of 3 imaging directions with similar settings. Amplitudes evaluated included b = 0 s/mm2 and b = 1,000 s/mm2. Approximate scan time was 20 minutes. All images were obtained with a 3.0-T MRI scanner.b

Diffusion measurements, including ADC and FA, were obtained from MRI images by use of the diffusion tensor and T2-weighted sequences as an anatomic reference. All measurements were obtained with commercial software.c The ADC and FA measurements for each dog were calculated separately by 2 observers (MJM and CPO) using 3 circular ROIs. These values were used to calculate mean ADC and FA values for the gray and white matter. Areas evaluated for both the gray and white matter included the cerebral lobes (frontal, parietal, occipital, and temporal), cerebellum, thalamus, pons, medulla, and hippocampus. Measurements were obtained for the left and right hemispheres for each of these areas, except for the pons, medulla, and hippocampus. Three additional circular ROIs were placed at each site to generate overall (combined gray and white matter) mean ADC and FA values. Thus, there were 35 brain locations/dog. Combined gray and white matter values were included because not all lesions cannot be definitively classified as to tissue of origin. We believed that combined values would be useful as a comparison for patients with infiltrative lesions that cannot be localized to gray or white matter or that involve both areas.

Size and position of each ROI were adjusted to maximize the amount of tissue within the ROI while avoiding inclusion of other structures (such as CSF in adjacent sulci). Size of each ROI was highly variable among various sites within a dog and also among dogs and was determined on an individual basis by each observer (Figures 1 and 2). For the measurements of combined gray and white matter, ROI area was 30 mm3, whereas individual ROI areas were typically between 5 and 12 mm3. Standardization of slices for each ROI was not performed, and each area and slice were selected by each observer. All 4 sequences were used to determine the areas to be measured by each observer because this provided a more realistic clinical application.

Figure 1—
Figure 1—

Representative T2-FLAIR (A) and diffusion tensor (B) images and diffusion-weighted MRI images for evaluation of ADC (C) and FA (D) at the level of the lateral ventricles in a neurologically normal dog. In each panel, a circular ROI was placed in the hippocampus (triangle), white matter of the occipital lobe (diamond), and gray matter of the frontal lobe (square).

Citation: American Journal of Veterinary Research 78, 5; 10.2460/ajvr.78.5.601

Figure 2—
Figure 2—

Representative T2-FLAIR (A) and diffusion tensor (B) images and diffusion-weighted MRI images for evaluation of the ADC (C) and FA (D) at the level of the thalamus and cerebellum in a neurologically normal dog. In each panel, a circular ROI was placed in the thalamus (triangle) and the gray matter of the cerebellum (square).

Citation: American Journal of Veterinary Research 78, 5; 10.2460/ajvr.78.5.601

Statistical analysis

All statistical analyses were performed by use of commercially available statistical software.d The same statistical method was used for evaluation of ADC and FA values. Data points were evaluated by use of the Kolmogorov-Smirnov goodness-of-fit test for each of 35 brain locations to determine whether the data were normally distributed. The significant probability reported for this testing involved use of the Lilliefors significance correction.

Whenever 2 groups were compared, all continuous variables were evaluated separately by use of independent t tests to compare mean values within tissue types, among tissue types, or between right and left hemispheres. The ADC and FA values were also evaluated for each observer by use of independent t tests. Correlation of subject age with all tissue variables was performed by use of the Pearson correlation coefficient and Spearman rank correlation coefficient. A 1-way ANOVA was used for the evaluation of means within a lobe (gray matter, white matter, and combined gray and white matter for both the right and left hemispheres [6 variables]) and for comparison between the gray and white matter among 5 sites (4 cerebral lobes and the cerebellum). We made an assumption of unequal variance and used the Tamhane T2 post hoc test to determine whether there were significant relationships for ADC or FA values between tissue types or locations. All testing for equal or unequal variances was conducted by use of the Levene test.

All significant differences among means were under the assumption of unequal variance between groups. All testing involved use of a 2-tailed statistic for reporting P values. Values were considered significant at P < 0.05.

Results

Animals

Thirteen neurologically normal dogs undergoing anesthesia for tibial plateau leveling osteotomy were enrolled in the study. There were 5 males (4 neutered males and 1 sexually intact male) and 8 spayed females. Mean age of dogs was 6.0 years (range, 3.6 to 9.0 years), and mean body weight was 37.8 kg (range, 23.5 to 55.3 kg). Dogs comprised 3 Labrador Retrievers, and 1 each of Bernese Mountain Dog, Golden Retriever, Gordon Setter, Flat-Coated Retriever, Akita, Llewellyn Setter, Newfoundland, German Shepherd Dog, Redbone Coonhound, and Chesapeake Bay Retriever. No abnormalities were identified during evaluation of the general sequences (T2-weighted FLAIR images) obtained before acquisition of the diffusion-weighted images.

Diffusion-weighted MRI

Data for gray matter and for white matter were normally distributed. Data for combined gray and white matter in all locations, except for the left frontal lobe (P = 0.047), were normally distributed. Given the normality of the data and the likelihood that the 1 location with significant results represented a type 1 error, parametric testing was performed, and mean and SD values were reported. Mean ADC and FA values for the 35 measurements were summarized (Table 1). Comparisons of ADC and FA were performed for left and right hemispheres, observers, age of dogs, and lobe (both within and among lobes) and for gray matter, white matter, and combined gray and white matter.

Table 1—

Mean ± SD values for ADC and FA on diffusion-weighted MRI images obtained from 13 neurologically normal dogs.

LocationADC (X 10−3 s/mm2)FA
Cerebellum
 Left hemisphere
  Gray0.8349 ± 0.05520.2320 ± 0.0381
  White0.7038 ± 0.05730.5542 ± 0.0759
  Gray and white0.7495 ± 0.05270.3619 ± 0.0540
 Right hemisphere
  Gray0.8369 ± 0.06540.2296 ± 0.0319
  White0.7196 ± 0.05730.5437 ± 0.0633
  Gray and white0.7460 ± 0.04080.3582 ± 0.0598
Frontal lobe
 Left hemisphere
  Gray0.9225 ± 0.07520.2003 ± 0.0334
  White0.7092 ± 0.04460.6102 ± 0.0742
  Gray and white0.8620 ± 0.05960.3765 ± 0.0508
 Right hemisphere
  Gray0.9107 ± 0.05640.2194 ± 0.0396
  White0.7151 ± 0.02890.6144 ± 0.0666
  Gray and white0.8487 ± 0.06150.3620 ± 0.0491
Parietal lobe
 Left hemisphere
  Gray0.9200 ± 0.05490.2135 ± 0.0460
  White0.6979 ± 0.06500.5994 ± 0.0721
  Gray and white0.8539 ± 0.05610.3696 ± 0.0630
 Right hemisphere
  Gray0.9273 ± 0.06390.2232 ± 0.0423
  White0.7332 ± 0.04000.6128 ± 0.0754
  Gray and white0.8358 ± 0.05240.3864 ± 0.0576
Occipital lobe
 Left hemisphere
  Gray0.8632 ± 0.05800.2267 ± 0.0379
  White0.6897 ± 0.02720.5837 ± 0.0423
  Gray and white0.7885 ± 0.03910.3744 ± 0.0448
 Right hemisphere
  Gray0.8866 ± 0.03760.2364 ± 0.0365
  White0.6936 ± 0.04130.5882 ± 0.0788
  Gray and white0.7982 ± 0.03630.3754 ± 0.0578
Temporal lobe
 Left hemisphere
  Gray0.8829 ± 0.05490.1978 ± 0.0335
  White0.7127 ± 0.05970.5136 ± 0.0842
  Gray and white0.8638 ± 0.05080.2943 ± 0.0534
 Right hemisphere
  Gray0.8767 ± 0.04820.2058 ± 0.0336
  White0.7071 ± 0.05690.5236 ± 0.0783
  Gray and white0.8696 ± 0.04370.3006 ± 0.0568
Hippocampus0.8776 ± 0.07370.2212 ± 0.0467
 Thalamus
 Left hemisphere0.7804 ± 0.04260.2807 ± 0.0487
 Right hemisphere0.7766 ± 0.03680.2900 ± 0.0466
Pons0.7724 ± 0.04520.5347 ± 0.1082
Medulla0.7005 ± 0.05270.3502 ± 0.0605

Significant differences were identified between gray and white matter, independent of the location. White matter had a significantly (P values ranged from < 0.001 to 0.005) higher FA value and a significantly (P values ranged from < 0.001 to 0.020) lower ADC value, compared with results for gray matter in all locations. The FA and ADC values for combined tissue measurements in all locations were intermediate relative to the values for the gray matter or white matter (Figure 3). Mean FA values for combined tissues differed significantly (P < 0.001) from the values for gray matter or white matter. Mean ADC values for the combined tissues differed significantly from the values for gray matter or white matter, except for a few areas in the cerebellum, frontal lobe, parietal lobe, and temporal lobe. In the cerebellum, combined tissue measurements for the left and right hemispheres did not differ significantly from the white matter measurements for the left and right hemispheres (white matter and combined tissues of the left hemisphere [P = 0.497], white matter of the left hemisphere and combined tissues of the right hemisphere [P = 0.475], white matter of the right hemisphere and combined tissues of the right hemisphere [P = 0.958], and white matter of the right hemisphere and combined tissues of the left hemisphere [P = 0.947]). In the frontal lobe, values did not differ significantly for gray matter of the left hemisphere and combined tissues of the left hemisphere (P = 0.392), gray matter of the left hemisphere and combined tissues of the right hemisphere (P = 0.161), gray matter of the right hemisphere and combined tissues of the left hemisphere (P = 0.483), and gray matter of the right hemisphere and combined tissues of the right hemisphere (P = 0.181). Only 2 areas in the parietal lobe had a combined tissue measurement that did not differ significantly from the measurement for gray matter or white matter (gray matter of the left hemisphere and combined tissues of the left hemisphere [P = 0.082] and gray matter of the right hemisphere and combined tissues of the left hemisphere [P = 0.070]). In the temporal lobe, combined tissue measurements for the left and right hemispheres did not differ significantly from the gray matter measurements for the left and right hemispheres (gray matter and combined tissues of the left hemisphere [P = 0.999], gray matter of the left hemisphere and combined tissues of the right hemisphere [P = 1.000], gray matter of the right hemisphere and combined tissues of the right hemisphere [P = 1.000], and gray matter of the right hemisphere and combined tissues of the left hemisphere [P = 1.000]). Finally, ADCs for gray matter of the left and right hemispheres of the thalamus were not significantly (P = 1.000) different from the values for the corresponding combined tissues of the left and right hemispheres.

Figure 3—
Figure 3—

Mean ADC values for the white matter, gray matter, and combined tissues of the left hemisphere (black bars) and right hemisphere (gray bars) of the frontal lobes of 13 neurologically normal dogs.

Citation: American Journal of Veterinary Research 78, 5; 10.2460/ajvr.78.5.601

No significant differences were identified for ADC and FA between the left and right hemispheres (all P > 0.80 for ADC and all P > 0.95 for FA), which allowed further statistical analysis to determine the mean results for the 2 hemispheres. For ADC measurements, no significant differences were identified between observers (gray matter and white matter) or among lobes (white matter and combined tissues). For FA measurements, no significant correlations or differences were identified for age (gray matter, white matter, and combined tissues) or among lobes (gray matter, white matter, and combined tissues).

Comparisons among lobes revealed no significant differences for ADC in white matter or for ADC or FA in combined tissues. Significant interlobar differences in FA were identified for the white matter in several locations and for ADC and FA in gray matter. For ADC measurements of gray matter, values for the cerebellum differed significantly (P ≤ 0.001) from those for the frontal lobes and occipital lobes, and values for the occipital lobes differed significantly (P = 0.021) from those for the temporal lobes. For FA measurements of gray matter, values for the cerebellum differed significantly (P = 0.032) from those for the parietal lobes, and values for the parietal lobes differed significantly (P = 0.035) from those for the temporal lobes. For FA measurements of white matter, values for the cerebellum differed significantly from those for the frontal lobes (P = 0.017) and values for the parietal lobes differed significantly from those for the frontal lobes (P ≤ 0.001), occipital lobes (P ≤ 0.001), and temporal lobes (P = 0.014).

Significant differences in ADC between observers were detected for the right hemisphere of the thalamus (mean difference, 0.031 × 10−3 s/mm2 [P = 0.033]), left hemisphere of the thalamus (mean difference, 0.041 × 10−3 s/mm2 [P = 0.013]), and right hemisphere of the frontal lobe (mean difference, 0.057 × 10−3 s/mm2 [P = 0.042]), compared with values for combined tissues. Several additional locations with significant differences in FA between observers were identified, including in the cerebellum (white matter of the right hemisphere [mean difference, 0.063; P = 0.035]), frontal lobes (white matter of the left hemisphere [mean difference, 0.091; P ≤ 0.001], gray matter of the left hemisphere [mean difference, 0.031; P = 0.029], white matter of the right hemisphere [mean difference, 0.053; P = 0.031], and gray matter of the right hemisphere [mean difference, 0.051; P = 0.049]), parietal lobes (white matter of the right hemisphere [mean difference, 0.072; P = 0.004]), temporal lobes (white matter of the left hemisphere [mean difference, 0.068; P = 0.044] and white matter of the right hemisphere [mean difference, 0.057; P = 0.044]), and combined tissues of the right hemisphere (mean difference, 0.052; P = 0.011) and medulla (mean difference, 0.52; P = 0.026).

A significant positive correlation was identified between ADC and patient age for multiple areas in all regions (gray matter, white matter, and combined tissues; Table 2). However, there was not a significant correlation between FA and patient age.

Table 2—

Significant* values for the correlation (Pearson correlation) between age and ADC values in diffusion-weighted MRI images of brains of 13 neurologically normal dogs.

LocationrP value
Cerebellum left hemisphere gray and white matter combined0.4180.034
Frontal lobe
 Left hemisphere white matter0.4460.023
 Left hemisphere gray and white matter combined0.4550.020
Parietal lobe
 Left hemisphere white matter0.429.0029
 Left hemisphere gray and white matter combined0.423.0031
 Right hemisphere white matter0.5080.008
 Right hemisphere gray and white matter combined0.4880.012
Occipital lobe left hemisphere gray matter0.4450.023
Temporal lobe left hemisphere gray and white matter combined0.5470.004
Hippocampus0.673< 0.001
Thalamus
 Left hemisphere gray and white matter combined0.4850.012
 Right hemisphere gray and white matter combined0.4590.018
Pons0.4920.011

Values were considered significant at P < 0.05.

Discussion

In the study reported here, ADC and FA values were significantly different between gray matter and white matter, as hypothesized. Differences in tissue structure of gray and white matter can explain these differences for both ADC and FA. White matter is predominantly composed of axons; thus, the diffusion pattern is highly directional along the axonal pathways, which results in lower ADC and higher FA values than for gray matter.12 Gray matter is predominately composed of cell bodies, which results in a less directionally oriented pattern and higher diffusion freedom.13 It is not surprising that the combined tissue values (values encompassing components of both gray and white matter) in both lobes would be between the individual values for gray matter and white matter.

The present study provided important information with respect to diffusion values within the brain of neurologically normal dogs, such as the lack of difference between the right and left cerebral hemispheres and the manner in which architecture of the brain (eg, differences between lobes and structural differences between gray and white matter) can significantly alter diffusion patterns and measurements. These ADC and FA values can be extrapolated and used when performing evaluations to detect abnormalities.

Comparison of mean values for dogs of the present study with values for a study9 on ADC values in gray and white matter of healthy humans revealed that values were within similar ranges. Unfortunately, investigators of that study9 of humans did not evaluate FA values for comparison. Differences in interlobar values for FA (gray matter and white matter) and ADC (gray matter) indicated that global values may not be meaningful when comparing normal with abnormal structures, and specific reference values may be required for each lobe. The increasing ADC values detected in white matter of multiple locations with increasing age presumably were attributable to physiologic alterations in tissue structure and composition, as opposed to pathological changes, and recognition of this change would be important when assessing brains for abnormalities. The exact cause for such changes in dogs may be similar to that reported for humans, whereby there is demyelination with loss of axonal integrity with age, which results in an increase in extracellular volume.11 However, it has also been reported9 that ADC in humans is not affected by age; thus, effects of age on ADC are still debatable.

Differences between the 2 observers for FA values, which ranged from 7% to 16%, and for ADC values in 3 locations, all of which were < 10%, may have been related to observer inexperience because these represented the first 13 patients for which such measurements have been obtained or possibly to inherent error with the available resolution. Although there was some interobserver variation, all comparable ADC and FA values for combined gray and white matter tissues were within the ranges reported for another veterinary study,8 which may indicate variation even within lobes. Although the FA values for white matter in the study reported here are almost 2 times as high as those reported in the aforementioned study,8 it should be mentioned that direct comparison of results may not be ideal, given the differences in separations between the tissues for measurement purposes as well as machine measurement versus human measurement. Investigators in that other study8 separated measurements into fiber tracts instead of white and gray matter, as was done for the present study, which also makes it difficult to directly compare results.

Comparison of equivalent values for the present study with those in another study2 in which investigators evaluated ADC values in certain regions of the brain of clinically normal dogs revealed that only a few were similar. Comparable locations included the hippocampus, thalamus, cortex cerebri, and piriform lobe, with the right and left hemispheres when possible. The cerebral cortex was measured in the region of the frontal lobe in the aforementioned study,2 which therefore enabled us to make comparisons with results for the study reported here. The ADC values for the thalamus of the right hemisphere, cerebral cortex of the right and left hemispheres, and piriform lobe of the right hemisphere for dogs of the present study were within the range of values and within 1 SD of the mean value reported for that other study.2 The ADC value for the hippocampus in the present study was lower than that in that other study, but it was within 1 SD of the mean value for that other study.2 In contrast, ADC values for both the thalamus and piriform lobe of the left hemisphere of the present study were not within the range or within 1 SD of the mean value for that other study.2 Some differences between results for these 2 studies may be attributable to the type of imaging system, strength of the magnet, and technique used for obtaining images. In that other study,2 a 1.0-T magnete was used, whereas in the study reported here, a 3.0-T magnetb was used. In addition, the b value was only 800 s/mm2 in that other study,2 whereas the b value was 1,000 s/mm2 in the present study. Each of these factors may have contributed to unequal values between the 2 studies. Furthermore, ADC values can be altered by several factors, including strength of the magnet (4% to 9%), vendor (7%), and coil (8%).14 Another possible factor was the type of patient. In the present study, multiple breeds were included, although all dogs weighed > 23 kg. In the aforementioned study,2 investigators used purpose-bred Beagles (mean ± SD body weight, 9.5 ± 1.6 kg). Body weight did not affect ADC measurements in the present study and was not definitively evaluated in that other study2 because the dogs were essentially all the same size. It is unclear whether differences in results between these 2 studies were affected by technical differences or patient differences. Given that there were differences in only a few regions, size of a patient may not be a factor for ADC values.

Another difference between the present study and that other study2 was the significant differences between left and right hemispheres in that other study, which were not evident in the study reported here. It has been found in humans15 and chimpanzees16 that right-handed patients typically have higher FA values for the left hemisphere than for the right hemisphere. This may also be clinically relevant in canine patients because dogs demonstrate a paw preference.17,18 In 1 study,17 the distribution in canine patients was skewed toward the right paw. The reason for a lack of this finding in the present study is unclear.

Differences in ADC values for the thalamus between observers, which represented 2 of 3 values with significant differences, were thought to be a combination of the anatomic structures and inexperience of the 2 observers. The thalamus is closely associated with the internal capsule and other structures; thus, it was thought that this association likely resulted in partial volume averaging and inclusion of these structures in the diffusion measurements, which made the measurements highly dependent on the slice selected and the surrounding structures involved. Additionally, given the number of comparisons performed as part of this study, it is possible that some of the significant differences actually represented a type 1 error.19

A possible limitation of the present study included the fact that these measurements typically have not been obtained, meaning some degree of error may have been expected with the initiation of a new measurement technique. However, because results for this study are similar to results for other studies, error in measurement technique (if any) may not have been an important factor. In addition, a limitation of the present study was the fact that most patients undergoing tibial plateau leveling osteotomy are medium- to large-breed dogs. Although no significant difference was identified for measured values on the basis of body weight, it is possible that there would be a difference for smaller breeds and between small breeds relative to medium to large breeds. Additionally, there is likely lower maximal spatial resolution for small patients, as each pixel represents a larger proportion of the patient's anatomy, which may result in greater measurement error or variability.

It could not be determined whether size of the patients in another study2 and the study reported here was the cause of unequal values. Potential differences between small-breed dogs and larger dogs could stem from actual tissue differences or simply may be an effect of the spatial resolution of the imaging system. One area for possible additional studies would be evaluation of values in neurologically normal small-breed dogs. Finally, another limitation was the study size (13 dogs). Ideally, a larger patient population would have been recruited.

In the study reported here, it was determined that diffusion values were symmetric between the 2 cerebral hemispheres and that there were marked differences between gray and white matter, which may be useful in further localizing intracranial disease with respect to gray and white matter. In addition, differences in diffusion values among lobes may support that there are substantial anatomic differences among lobes, depending on function of the specific lobes. Obtaining values for clinically normal animals is the first step in determining the use of diffusion imaging for characterization of abnormalities in clinical settings. The authors believe the data reported here were an important starting point toward clinical application of diffusion-weighted imaging and diffusion tensor imaging of the canine brain.

Acknowledgments

Supported in part by a Small Companion Animal Grant from the University of Minnesota and an American College of Veterinary Radiology Research Grant.

Presented as a poster at the American College of Veterinary Radiology Annual Conference, Minneapolis, October 2015.

ABBREVIATIONS

ADC

Apparent diffusion coefficient

FA

Fractional anisotropy

FLAIR

Fluid-attenuated inversion recovery

ROI

Region of interest

TE

Echo time

TR

Repetition time

Footnotes

a.

HD T/R quad extremity, Invivio, Pewaukee, Wis.

b.

GE Signa HDx 3.0-T MRI scanner, GE Medical Systems, Milwaukee, Wis.

c.

FuncTool, GE Medical Systems, Milwaukee, Wis.

d.

SPSS, version 20, SPSS Inc, Chicago, Ill.

e.

Philips Integra Gyroscan, Philips Healthcare, Hamburg, Germany.

References

  • 1. Cha S. Update on brain tumor imaging: from anatomy to physiology. Am J Neuroradiol 2006; 27:475487.

  • 2. Hartmann A, Soffler C, Failing K, et al. Diffusion-weighted magnetic resonance imaging of the normal canine brain. Vet Radiol Ultrasound 2014; 55:592598.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 3. Neil JJ. Diffusion imaging concepts for clinicians. J Magn Reson Imaging 2008; 27:17.

  • 4. Cherubini GB, Mantis P, Martinez TA, et al. Utility of magnetic resonance imaging for distinguishing neoplastic from non-neoplastic brain lesions in dogs and cats. Vet Radiol Ultrasound 2005; 46:384387.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5. Sener RN. Diffusion MRI: apparent diffusion coefficient (ADC) values in the normal brain and a classification of brain disorders based on ADC values. Comput Med Imaging Graph 2001; 25:299326.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6. Sutherland-Smith J, King R, Faissler D, et al. Magnetic resonance imaging apparent diffusion coefficients for histologically confirmed intracranial lesions in dogs. Vet Radiol Ultrasound 2011; 52:142148.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 7. Price SJ, Tozer DJ, Gillard JH. Methodology of diffusion-weighted, diffusion tensor and magnetization transfer imaging. Br J Radiol 2011; 84:121126.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8. Hecht S, Adams WH. MRI of brain disease in veterinary patients. Part 1: basic principles and congenital brain disorders. Vet Clin North Am Small Anim Pract 2010; 40:2138.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9. Anaya Garcia MS, Hernández-Anaya JS, Meléndez OM, et al. In vivo study of cerebral white matter in the dog using diffusion tensor tractography. Vet Radiol Ultrasound 2015; 56:188195.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10. Helenius J, Soinne L, Perkio J, et al. Diffusion-weighted MR imaging in normal human brains in various age groups. Am J Neuroradiol 2002; 23:194199.

    • Search Google Scholar
    • Export Citation
  • 11. Watanabe M, Sakai O, Ozonoff A, et al. Age-related apparent diffusion coefficient changes in the normal brain. Radiology 2013; 266:575582.

  • 12. Citation [white matter]. In: Encyclopedia and dictionary of medicine, nursing, and allied health online. Available at: www.medical-dictionary.thefreedictionary.com. Accessed Dec 13, 2015.

    • Search Google Scholar
    • Export Citation
  • 13. Citation [grey matter]. In: Encyclopedia and dictionary of medicine, nursing, and allied health online. Available at: www.medical-dictionary.thefreedictionary.com. Accessed Dec 13, 2015.

    • Search Google Scholar
    • Export Citation
  • 14. Sasaki M, Yamada K, Watanabe Y, et al. Variability in absolute apparent diffusion coefficient values across different platforms may be substantial: a multivendor, multi-institutional comparison study. Radiology 2008; 249:624630.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15. Büchel C, Raedler T, Sommer M, et al. White matter asymmetry in the human brain; a diffusion tensor MRI study. Cereb Cortex 2004; 14:945951.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16. Li L, Preuss TM, Rilling JK, et al. Chimpanzee (Pan troglodytes) precentral corticospinal system asymmetry and handedness: a diffusion magnetic resonance imaging study. PLoS One 2010; 5:e12886.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17. Tan U. Paw preferences in dogs. Int J Neurosci 1987; 32:825829.

  • 18. Quaranta A, Siniscalchi M, Frate A, et al. Paw preferences in dogs: relations between lateralized behaviours and immunity. Behav Brain Res 2004; 153:521525.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 19. Lane DM. The problem of too many statistical tests: subgroup analyses in a study comparing the effectiveness of online and live lectures. Numeracy 2013; 6:13.

    • Search Google Scholar
    • Export Citation
  • Figure 1—

    Representative T2-FLAIR (A) and diffusion tensor (B) images and diffusion-weighted MRI images for evaluation of ADC (C) and FA (D) at the level of the lateral ventricles in a neurologically normal dog. In each panel, a circular ROI was placed in the hippocampus (triangle), white matter of the occipital lobe (diamond), and gray matter of the frontal lobe (square).

  • Figure 2—

    Representative T2-FLAIR (A) and diffusion tensor (B) images and diffusion-weighted MRI images for evaluation of the ADC (C) and FA (D) at the level of the thalamus and cerebellum in a neurologically normal dog. In each panel, a circular ROI was placed in the thalamus (triangle) and the gray matter of the cerebellum (square).

  • Figure 3—

    Mean ADC values for the white matter, gray matter, and combined tissues of the left hemisphere (black bars) and right hemisphere (gray bars) of the frontal lobes of 13 neurologically normal dogs.

  • 1. Cha S. Update on brain tumor imaging: from anatomy to physiology. Am J Neuroradiol 2006; 27:475487.

  • 2. Hartmann A, Soffler C, Failing K, et al. Diffusion-weighted magnetic resonance imaging of the normal canine brain. Vet Radiol Ultrasound 2014; 55:592598.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 3. Neil JJ. Diffusion imaging concepts for clinicians. J Magn Reson Imaging 2008; 27:17.

  • 4. Cherubini GB, Mantis P, Martinez TA, et al. Utility of magnetic resonance imaging for distinguishing neoplastic from non-neoplastic brain lesions in dogs and cats. Vet Radiol Ultrasound 2005; 46:384387.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5. Sener RN. Diffusion MRI: apparent diffusion coefficient (ADC) values in the normal brain and a classification of brain disorders based on ADC values. Comput Med Imaging Graph 2001; 25:299326.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6. Sutherland-Smith J, King R, Faissler D, et al. Magnetic resonance imaging apparent diffusion coefficients for histologically confirmed intracranial lesions in dogs. Vet Radiol Ultrasound 2011; 52:142148.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 7. Price SJ, Tozer DJ, Gillard JH. Methodology of diffusion-weighted, diffusion tensor and magnetization transfer imaging. Br J Radiol 2011; 84:121126.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8. Hecht S, Adams WH. MRI of brain disease in veterinary patients. Part 1: basic principles and congenital brain disorders. Vet Clin North Am Small Anim Pract 2010; 40:2138.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9. Anaya Garcia MS, Hernández-Anaya JS, Meléndez OM, et al. In vivo study of cerebral white matter in the dog using diffusion tensor tractography. Vet Radiol Ultrasound 2015; 56:188195.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10. Helenius J, Soinne L, Perkio J, et al. Diffusion-weighted MR imaging in normal human brains in various age groups. Am J Neuroradiol 2002; 23:194199.

    • Search Google Scholar
    • Export Citation
  • 11. Watanabe M, Sakai O, Ozonoff A, et al. Age-related apparent diffusion coefficient changes in the normal brain. Radiology 2013; 266:575582.

  • 12. Citation [white matter]. In: Encyclopedia and dictionary of medicine, nursing, and allied health online. Available at: www.medical-dictionary.thefreedictionary.com. Accessed Dec 13, 2015.

    • Search Google Scholar
    • Export Citation
  • 13. Citation [grey matter]. In: Encyclopedia and dictionary of medicine, nursing, and allied health online. Available at: www.medical-dictionary.thefreedictionary.com. Accessed Dec 13, 2015.

    • Search Google Scholar
    • Export Citation
  • 14. Sasaki M, Yamada K, Watanabe Y, et al. Variability in absolute apparent diffusion coefficient values across different platforms may be substantial: a multivendor, multi-institutional comparison study. Radiology 2008; 249:624630.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15. Büchel C, Raedler T, Sommer M, et al. White matter asymmetry in the human brain; a diffusion tensor MRI study. Cereb Cortex 2004; 14:945951.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16. Li L, Preuss TM, Rilling JK, et al. Chimpanzee (Pan troglodytes) precentral corticospinal system asymmetry and handedness: a diffusion magnetic resonance imaging study. PLoS One 2010; 5:e12886.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17. Tan U. Paw preferences in dogs. Int J Neurosci 1987; 32:825829.

  • 18. Quaranta A, Siniscalchi M, Frate A, et al. Paw preferences in dogs: relations between lateralized behaviours and immunity. Behav Brain Res 2004; 153:521525.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 19. Lane DM. The problem of too many statistical tests: subgroup analyses in a study comparing the effectiveness of online and live lectures. Numeracy 2013; 6:13.

    • Search Google Scholar
    • Export Citation

Advertisement