The Clinical Application Analysis Of SWI And DTI On Traumatic Brain Injury

Background and purposeIn recent years, with the rapid social and economic development, the number oftrauma caused by traffic accidents and facilities is increasing year by year. Amongthese, traumatic brain injury(TBI) had a proportion of about10%~15%, rankedsecond after limb trauma. But the mortality and morbidity of TBI was ranked first.According to statistics, trauma is the leading cause of death and disability in thegroup of young people at home and abroad, of which about50%~60%was due toTBI. Meanwhile, some TBI patients had dizziness, headache, seizures, behavioralcognitive dysfunction, limb movement disorder, and even the survival of persistentvegetative state and other complications after injury. Therefore, early diagnosis andaccurate assessment of the severity of TBI is conducive to treatment and effectiverehabilitation training in a timely manner, so as to reduce the mortality and morbidityand reduce sequelae.Imaging technology is the main method of TBI diagnosis, including computedtomography(CT), magnetic resonance imaging(MRI) and so on. Because of its rapidacquisition techniques (typically <30seconds), CT has been called the "Goldenstandard" of the acute TBI diagnosis. But CT had disadvantage on showingparenchymal damage, often lead to missed diagnosis or made too light judgment onthe severity of TBI. MRI made up the shortcomings of CT and could accurately display the injury site location. However, conventional CT and MRI could only showchanges of brain anatomy. Functional magnetic resonance imaging(fMRI), includingsusceptibility weighted imaging(SWI), diffusion tensor imaging(DTI), magneticresonance spectroscopic imaging(MRSI), blood oxygen level dependent functionalimaging(BOLD-fMRI), etc., could non-invasively provide pathophysiology andfunctional status information of TBI. Now fMRI had been used more frequently in thedetermination of TBI severity and prognosis prediction.SWI could clearly show the distribution of micro-hemorrhage, thus presume theextent of damage. DTI pioneered the in vivo detection of white matter changes. DTIcould detect the micro-structural changes in white matter, which is unparalleled inother technologies. TBI were often associated with axonal injury of brain tissueregardless of types. And the rest MR sequences were rarely able to detect thepresence of axonal injury, so the application of SWI and DTI on TBI was of greatimportance.The value of SWI and DTI had been confirmed by some reports, especially incases which had normal appearing on conventional CT and MRI, or GCS and clinicalsymptoms was not consistent with conventional imaging findings, as well as inconcussion, DAI and so on. SWI specialized in the detection of hemorrhagic braininjury. DTI mainly detect non-hemorrhagic axonal injury. Their combination couldfully show the extent of damage after TBI. But up to now, researches of SWI and DTIon TBI was not yet mature, there were still a lot of exploration space. Therefore, weexplore the diagnostic value of SWI and DTI on TBI in this thesis.Materials and methods80research subjects were divided into four groups, among which there were20health volunteer,20mild TBI patients,20moderate TBI patients and20severe TBIpatients. Glasgow Coma Scale (GCS) of60TBI patients were scored by twoexperienced neurosurgeons.1. The clinical application value of SWI in TBI: All patients underwentconventional MR and SWI scan. The involving the regions, numbers and areas of hemorrhagic lesions detected by conventional MRI and SWI were separately recordedby two experienced radiologists. The statistic differences between conventional MRIand SWI were compared. The Kruskal-Wallis test were underwent among the mildTBI group, moderate TBI group and severe TBI group, then Bonferroni method wereused to compare between each two groups. The correlation analysis between GCSscores and the number of involving regions, number of lesions, areas of lesionsdetected by conventional MRI and SWI were underwent by Spearman rankcorrelation analysis method. P<0.05was considered statistically significant.2. The clinical application value of DTI in TBI: All patients underwentconventional MRI and DTI scan. The past-processing were underwent by twoexperienced radiologists. Regions of interest(ROI) of37brain regions were outlinedand the fractional anisotropy(FA) values were recorded. The37brain regions were asfollows:1knee of corpus callosum,2body of corpus callosum,3splenium of corpuscallosum,4right cingulate bundle,5left cingulate bundle,6right centrum semiovale,7left centrum semiovale,8right frontal lobe,9left frontal lobe,10right parietal lobe,11left parietal lobe,12right occipital lobe,13left occipital lobe,14right temporallobe,15left temporal lobe,16right anterior limb of internal capsule,17left anteriorlimb of internal capsule,18right knee of internal capsule,19left knee of internalcapsule,20right posterior limb of internal capsule,21left posterior limb of internalcapsule,22right external capsule,23left external capsule,24right caudate nucleus,25left caudate nucleus,26right basal ganglia,27left basal ganglia,28right thalamus,29left thalamus,30right cerebral peduncle,31left cerebral peduncle,32right part ofthe pons,33left part of pons,34right part of midbrain,35left part of midbrain,36right cerebellar hemisphere,37left cerebellar hemisphere. For each brain region’s FAvalues, statistic comparisons among the mild TBI group, moderate TBI group andsevere TBI group were made, and corresponding test methods were used to makepairwise comparison. The correlation analysis between GCS scores and FA values ofeach region were performed by Spearman rank correlation analysis method. P<0.05was considered statistically significant. Results1. For the mild TBI group, moderate TBI group, severe TBI group, and total TBIgroup, the differences of involving regions’ number, lesions’ number detected byconventional MRI and SWI are statistically significant (P<0.05). the number detectedby SWI is higher. The differences between each two groups were statisticallysignificant (P<0.05). Severe TBI group got the maximum number of involvingregions, lesions, and the largest areas. Followed by moderate group and mild group.The GCS scores are highly negatively correlated with the number of involvingregions, number of lesions, areas of lesions detected by conventional MRI and SWI.In descending order of relevance: GCS with areas detected by SWI(r=-0.982,P=0.000), the number of lesions detected by SWI(r=-0.941, P=0.000), the number ofinvolving regions detected by SWI(r=-0.900, P=0.000), the number of involvingregions detected by conventional MRI(r=-0.792, P=0.000), the number of lesionsdetected by conventional MRI(r=-0.775, P=0.000).2. For all the four groups, the data of FA values in different regions are subject tonormal distribution. In the control group, the value of the splenium of corpuscallosum get the highest FA values(879.58±45.08). The FA values of the four groupsare compared, results showed that: the corpus callosum, bilateral cingulate bundle,bilateral centrum semiovale, bilateral occipital temporal lobe, bilateral internalcapsule, bilateral external capsule, bilateral thalamus, bilateral cerebral peduncle,pons, midbrain, bilateral cerebellar hemispheres, the right basal ganglia and rightcaudate nucleus, totally30regions, differ in FA values between the fourgroups(P<0.05). Among these30regions,18regions’ FA values, for example corpuscallosum region, gradually reduce as the severity of TBI aggravate. Further, the resultof pairwise comparisons among the four groups showed that: for control group andmild TBI group, control and moderate, control and severe, mild and moderate, mildand severe, moderate and severe, the number of regions that differences of FA valuesare statistically significant are15,24,31,10,30,21. For most regions, FA values’order is severe <moderate <mild <control.The result of the correlation analysis between GCS scores and FA values of eachregion showed that: For totally30regions, for exmaple, the corpus callosum, bilateralcingulate bundle, the FA values get positive correlation with the GCS. Wherein, the correlation coefficient of the right cingulum is maximum (r=0.872), followed by theright anterior limb of internal capsule(r=0.801), left cingulum (r=0.787), the spleniumof corpus callosum(r=0.775), the body of corpus callosum (r=0.765).Conclusions1. The clinical applications of SWI in detecting hemorrhagic lesions after TBIhave great value.2. The clinical applications of DTI in detecting the severity of white matterdamage after TBI have great value.3. The clinical applications of SWI and DTI is valuable in deciding the TBIseverity.