Clinical Values Of MR-Diffusion Tensor Imaging In Full-term Neonates With Hypoxic Ischemic Brain Injury

Objective1. To explore the feasibility and value of MR-DTI in full-term neonates with hypoxic ischemic brain injury.2. To investigate whether the fractional anisotropy (FA) values can quantitatively assess the severity of hypoxic ischemic brain injury and predict prognosis in full-term neonates.Material and Method1. SubjectsFifty term neonates were hospitalized in neonatal Wards of NariFang hospital between March to September in 2009 and took the conventional MRI, diffusion weighted imaging(DWI) and diffusion tensor imaging(DTI), all of them had different brain damage on the conventional MRI and DWI. Fifty cases of neonate(26 males and 24 females) of 39.7 mean weeks of gestation(37.4 to 42.0 weeks) and 3.32 kilograms mean weight(3.00 to 4.00 kg) were divided into three groups by the findings in the conventional MRI and DWI:mild brain injury(30 cases), moderate brain injury(11 cases), severe brain injury(9 cases). Forty cases of them underwent NBNA, the score range was from 25 to 38.All cases took cranial MRI examination between 2 to 18 days after birth, and the body conditions of them were stable. Inclusion criteria:①asphyxia in varying degrees (including prenatal factors, factors at birth and postnatal factors. Prenatal risk factors:secondary low blood pressure after maternal hemorrhage before birth, pregnancy hypertension and placental abnormalities; risk factors at birth:born with intrauterine embarrassed break and umbilical cord knotted around the neck; postnatal risk factors:as mainly due to some pathological conditions caused dysfunction of respiratory center and pulmonary ventilation obstacled, such as: hyaline membrane disease, apnea, seizures, intracranial hemorrhage and neonatal hypoxic-ischemic encephalopathy. Who meet one of the above criteria was included);②full-term children with 37 to 42 weeks of gestational age and 3.00 to 4.00 kg of weight;③MRI examination time was within 28 days after birth;④underwent the conventional MRI, DWI and DTI scans, the examiners sleeped deeply during scansing and the image quality was good. Exclusion criteria:①neonates with no clear history of asphyxia, hyperbilirubinemia and any congenital diseases;②premature infants and overdue births, less than 37 weeks or more than 42 weeks of gestational age, less than 3.00 kg or more than 4.00 kg of weight;③MRI examination time was more than 28 days after birth;④Infant did not take or match during MRI, DWI or DTI scaning caused image quality drop and unable to post-processing and analysis.2. Main instruments and equipments ①the Signa EXCITE 3.0T superconducting magnetic resonance scanner of U.S. General Electric (GE) Company;②The 8-channel phased array head coil and plastic foam to fix examiner’s head;③image post-processing was done through the Functool software of GE company’s AW 4.3 workstation.3. Perparations before examinationIntramuscular injection of sodium Rumi (2-4 mg/kg) and, if necessary, administration of oral chloral hydrate (30～100 mg/kg) were given in 30min before scanning. Oxygen was administered through nasal oxygen cannula, and the neonates were wrapped with blankets and fixed their head with plastic sponge and fitted the ear canal with cotton plugs to reduce noise. An experienced pediatrician was accomplished during scaning.4. MR examination methodsExaminer’s head will be placed on the 8-channel phased array head coil, and a plastic sponge will be placed in the occipital and temporal side of the infant in order to fix head position, located the positioning line in the eyebrow level. Conventional MRI sequences included:T1WI axial and sagittal plane, T2WI and T2Flair axial plane imaging. All of the axial scan sequences were located the scan baseline on the canthomeatal line, with the thickness 5mm, layer spacing of 0.5mm, FOV:20×20cm, acquisition times of 2, acquired 16 to 20 layers according to the actual situation; sagittal scan sequences with layer thickness 4mm, layer spacing of 0.1mm, FOV:20×20cm, acquisition times of 1, acquired 13 to 15 layers according to the actual situation of. Both DWI and DTI scans using single-shot echo-planar imaging(EPI) sequence.The parameters of DWI scan were:b values=0,1000s/mm2, TR 6200ms, TE 100.9ms, slice thickness 5mm, layer spacing 0mm, FOV= 20×20cm, acquisition times 2. The parameters of DTI scan were:b values 1000s/mm2, TR 7000ms, TE 100.9ms, slice thickness 5mm, spacing 0mm, FOV=20×20cm, acquisition times 1, acquired in 25 noncollinear diffusion gradient directions.5. Image post-processing and analysisAll of the DTI data were processed through the Functool software of GE company’s AW 4.3 workstation, the FA values were acquired by an experienced doctor in MR diagnosis. Choosed the plane of basal-ganglia to set the threshold value, the upper limit to maximum and the lower limit just to eliminate background noise without loss of useful information. Ignoring the existence of the focal or diffuse lesions, we choosed 6 ROIs during the processing of DTI data:posterior limb of internal capsule(PLIC), centrum semiovale(CSO), frontal white matter(FWM), parietal white matter(PWM), occipital white matter(OWM), temporal white matter(TWM). All measurements were carried out bilaterally symmetrical, the size of interest areas was 15～25mm2, the ROIs were round except the posterior limb region of interest for the long oval. Each ROI was measured three times to reduce error, averaged FA values for the corresponding region. Used the same window width and window level on the DTI image post-processing.Imaging method of posterior limb of internal capsule corticospinal tract:hand selected bilateral posterior limb of internal capsule as a "seed point"in the direction encoded color (DEC) map, terminate conditions of diffusion tensor tractography (DTT) track:FA<0.18, step<160, ADC> 0.01, observed the characteristics of three-dimensional conformation of the corticospinal. 6. Statistical analysisData analysis was performed by using SPSS 13.0 sofe ware in this study. All data were subjected to 1-Sample Kolmogorov-Smirnov test for normality. Used Paired-Samples t test to compare bilateral FA values of one ROI. Merged the bilateral FA values if there was no difference and compared with the degree of hypoxic-ischemic brain injury and NBNA scores respectively if difference was exit between them. Used One-Way ANOVA and Multiple Comparisons to compared three groups of brain injury. Used Bivarite to analysis the correlation between FA values and NBNA scores. The statistical differences were considered as statistically significant when P values less than 0.05.ResultFifty cases of neonate(26 males and 24 females) had 39.7 mean weeks of gestation(37.4 to 42.0 weeks) and 3.32 kilograms mean weight(3.00 to 4.00 kg). Used Apgar Score to evaluate all neonates prenatal asphyxia, two of them did not have assessment after birth. Four cases had less than 3 Apgar Score in one minute, seventeen cases had more than 4 and less than 7, twenty-seven cases had normal scores. One had normal score in one minute but had score descent in followed twenty minutes. One severity infant had less than 7 score within 10 minute after birth.50 cases of full-term newborns, including 30 cases of mild brain injury, moderate brain injury in 11 cases,9 cases of severe brain injury. Mild brain injury showed stippled and strip high signals along cortical and subcortical white matter on T1WI (8/50) or supratentorial or subtentorial subarachnoid hemorrhage (5/50). Moderate brain damage including deep white matter lesions (20/50), characterized by punctate slightly high signal located symmetric in bilateral frontal deep white matter and banded high signal changes along the ventricular edge on T1WI, followed with focal cerebral edema (10/50). Besides the above imaging findings, the performance of neonates with severe brain injury contain any of the following:①basal ganglia and thalamus abnormal (3/50), the most common site of involvement is the putamen nucleus, followed by the globus pallidus and ventrolateral thalamus, showed symmetrical patchy heterogeneous high signal on T1WI, changes on T2WI was not obviously;②the myelination in posterior limb of internal capsule was obstructed, high signal disappeared (4/50) on T1WI;③The space in ventricle or outdoor, intraventricular hemorrhage (3/50) showed high signal on T1WI and low signal on T2WI, fluid level was often seen, with or without ventricular dilatation, large subarachnoid hemorrhage (3/50) showed high signal on T1WI and low signal on T2WI in the brain surface sulci, cisterns, cerebral fissure, straight sinus, tentorial sinus;④cystic necrosis in subcortical (1/50) and extensive brain edema (5/50).The gestational age and weight were not significantly different in three groups of neonates with brain damage. Bilateral FA values in PLIC, CSO, PWM, OWM was no significant difference while the bilateral FA values in FWM, TWM were significantly different. The FA values are in line with normal distribution, only TWML was not homogeneity of variance. FA values were significantly different in three groups of neonates with brain damage, multiple comparisons found that the FA value in TWML was not significant between moderate and severe brain injury group.colored fractional anisotropy (FA) map, directionally encoded color (DEC) map and the DTT three dimensional conformation of corticospinal tract within PLIC were used in this study. they can display changes of FA values in different regions of the brain and complicated white matter fasciculus. The transition from red to blue color representative of the gradual decrease of FA values in colored FA map, red represents the direction of moving left to right, green represents the direction of moving front to back, blue represents the direction of moving up to down in DEC map. The PLIC and CSO in severe brain injury group showed pale green color in FA maps, and yellow or red reduction, moderate and mild brain injury group increased gradually yellow or red. The white matter in severe brain injury group showed blue, moderate and mild brain injury group showed light green. The color of the ROI in severe brain injury group showed more mixed than moderate and mild brain injury groups in DEC maps, indicating that its orientation is not strong, the fiber bundles was significantly damaged. The DTT three-dimensional conformation of corticospinal tract within PLIC of severe brain injury group shows sparse, interrupted, shorter fibers, moderate and mild brain injury group had more fibers and less fibers interruptions.40 of all cases underwent NBNA in 5 to 7 days after birth and the NBNA score values ranging from 25 to 38. The FA value of each ROI in 40 cases of neonates are in line with normal distribution. There was a significant positive correlation between FA values and NBNA scores. The Lower FA values it was, the lower NBNA score, suggesting that poorer prognosis in the development of nervous system in neonates.Conclusion1. DTI can be used to study full-term neonates with hypoxic ischemic brain injury, FA value reflect the degree of molecular diffusion accurately.2. FA value is a sensitive indicator of hypoxic ischemic brain injury in full-term neonates, it can quantitatively assess the severity of brain damage.3. FA value would be a good indicator of prediction prognosis of nervous system development in full-term neonates with hypoxic ischemic brain injury, but it will take a long time to prove by longitudinal follow-up observation.