Use of Multimodal Imaging, 3D Segmented Atlas and Computational Analysis to
Evaluate CNS Damage Following Traumatic Brain Injury in Rats.
P. KULKARNI1, S. P. FINKLESTEIN2, *J. REN2, R. WU2, M. DAVENPORT2, M. NEDELMAN3, C. F. FERRIS1;
1Northeastern
ABSTRACT
Program #552.25
Poster #M11
Univ., Boston, MA; 2Biotrofix, Waltham, MA; 3Ekam Imaging, Boston, MA
MAIN RESULTS
Traumatic brain injury was delivered to the right dorsolateral frontoparietal cortex of adult
male rats (n=8) using the fluid percussion method. Five days later animals were imaged
using a 7.0 Tesla MR scanner for T1 and T2 relaxivity values, diffusion tensor imaging for
tractography and quantitative anisotropy and functional BOLD imaging in response to
hypercapnia (CO2 challenge) under awake conditions. The data for each imaging modality
was registered into a 3D segmented and annotated rat atlas delineating 152 brains areas. As
expected, the injury caused significant disruption in fiber tracts at the site of impact and
apparent shearing of fiber tracts that extended from the corpus callosum ventrally down the
external capsule and optic tract at the level of the lateral geniculate. When compared to
contralateral side of the brain there were significant differences in T2 relaxivity and several
measures of diffusivity as determined with quantitative anisotropy. Interestingly, areas distant
to the cortical insult were affected, particularly hypothalamus, amygdala and hippocampus.
These changes in subcortical brain areas may reflect pathological alterations in
microarchitecture affecting emotion and cognition.
CO2 Challenge
contralateral
ipsilateral
T2 Relaxivity Maps
Materials and Methods
Custom Radiofrequency Technology
Acclimating Rats for the Imaging
All animals were lightly anesthetized and placed into a copy of
the restraining system used during awake imaging. When fully
conscious, the animals were placed into a dark mock scanner
tube with a recording of a standard MRI pulse sequence
playing in the background. The reduction in autonomic and
somatic response measures of arousal and stress improve
the signal resolution and MR image quality. (King et al. 2005 J
Neurosci Meth.)
Imaging Technology
head coil
Quad transmit/receive
volume coil built into
the head holder
Optimal signal-to-noise
and field homogeneity
restrainer
Study Protocol
Adult, male Sprague Dawley® Rat were used in this study. Experiments were conducted using a Bruker Biospec 7.0T/20-cm
USR horizontal magnet (Bruker, Billerica, Massachusetts) and a 20-G/cm magnetic field gradient insert (ID = 12 cm) capable
of a 120-µs rise time (Bruker). Radiofrequency signals were sent and received with the quad-coil electronics built into the
animal restrainer. On test days, animals (n=8) were placed in the restrainer, and positioned in the magnet. Following Scans
were collected for each subject.
Anatomical scans : At the beginning of each imaging session, a high-resolution anatomical data set was collected using the
RARE pulse sequence (20 slice; 1 mm; field of vision [FOV] 3.0 cm; 256 × 256; repetition time [TR] 2.5 sec; echo time [TE]
12.4 msec; NEX 6; 6.5-minute acquisition time).
Functional Scan: were acquired using a multi-slice half Fourier acquisition, single shot, turbo spin echo sequence. A single
scanning session acquired 20 slices, 1mm thick, every 6.0 seconds (FOV 3.0 cm, matrix size 96 x 96, ETL 36, NEX 1)
repeated 100 times for a total time of 10 minutes. CO2 was administered through nose cone at 5 min into the imaging session.
T1 Measuremnt: Variable TR images were acquired using RARE pulse sequence (TE=12.5 and TR: 450, 800, 1400, 2200,
6000, msec.) Images were acquired with a field of view [FOV] 3 cm2, data matrix = 128×128×20 slices, thickness = 1 mm.
T1 values were calculated using Paravision™ software.
T2 Measurement: Images acquired using a multi-slice multi-echo (MSME) pulse sequence. The echo time (TE), was 11 ms,
and 16 echoes were acquired during imaging with a recovery time (TR) of 2500 ms. Images were acquired with a field of view
[FOV] 3 cm2, data matrix = 256×256×20 slices, thickness = 1 mm. T2 relaxation values were obtained from all the slices
ParaVison 5.1 software. The T2 values were computed using the equation; y=A+C*exp(-t/T2) (S.D. weighted). Where, A =
absolute bias, C = signal intensity, t = echo time and T2 = spin-spin relaxation time.
Diffusion Tensor Imaging: DTI images were acquired using 3D EPI pulse sequence with (TE = 19 ms, 8 segments, TR =500
msec). The data was collected in 10 direction with one B0 image and B value=1000.The data was processed using MedINRIA
( V1.9.4).
Method of TBI Induction
In the fluid percussion model, a small patch of dura is exposed, and a seal is made between the dura and an externallyapplied fluid-filled connector tube. Using a gravity-operated device specifically designed for this purpose, a preciselycalibrated fluid percussion wave is then transmitted from the connector tube to the dura [1-4]. This results in focal brain
contusion and cell death under the point of the connector tube on the dura and a percussion wave through the brain that
causes focal white matter damage and neuronal death in the distant hippocampus. The brain injury results in focal motor
deficits in the opposite limbs and memory disturbances
Data Analysis
Each subject was registered to a 3D segmented and annotated rat brain atlas (Ekam Imaging Inc). The alignment process was
facilitated by an interactive graphic user interface. The affine registration involved translation, rotation, and scaling in all 3
dimensions independently. The matrices that transformed the subject’s anatomy to the atlas space were used to embed each
slice within the atlas. All transformed pixel locations of the anatomy images were tagged with the segmented atlas regions,
creating a fully segmented representation of each subject. Parameter values for each ROI was computed based on each
segmented map.
Shown are the average (n=8) significant increase in BOLD signal intensity
(yellow/red) following 5% CO2 challenge over the rostral/caudal boundaries of the TBI
lesion. There were no significant differences in vascular responsivity between
lesioned and control sides. The T2 relaxivity maps show a change in transverse
relaxation in the rostral/caudal boundaries of the TBI lesion (arrows) in a
representative rat. This change reflects an increase in water due to edema
associated with the trauma. There were no significant changes in T1 relaxivity.
SUMMARY AND CONCLUSIONS
Magnetic resonance imaging was used to characterize the extent of brain injury in
response to a unilateral insult to the frontoparietal cortex. As expected, high
resolution anatomical images clearly show the rostral/caudal limits of tissue
disruption in the cortical mantel on the side of the trauma. While there were no
significant changes in T1 relaxivity, T2 measures clearly showed the extent of the
trauma as reflected by the increase of water with edema. When challenged with
CO2 there was no difference between the ipsilateral and contralateral halves of the
brain. This indicates no change in vascular responsivity and would predict normal
coupling between metabolically active brain regions and cerebral blood flow.
DTI with tractography showed disruption in myelinated fiber tracts on the side of the
brain trauma (see white circles in DTI panel). Quantitative anisotropy when
combined with co-registration into a 3D, segmented, annotated MRI rat atlas
detected many areas that showed alterations in water diffusivity. Interestingly, many
of these area were subcortical, particularly the hypothalamus, amygdala and
hippocampus. These data indicate trauma directed to the cortex can alter the
microarchitecture of deeper brain areas. This finding my help to explain the
changes in emotion and cognition associated with TBI.