| BackgroundIn today’s interdisciplinary prevalence,the exploration of neurology combined with imaging methods in the study of animal and human brain diseases has gradually developed into a hotspot.The study of brain gray matter and white matter structure by magnetic resonance technology not only provides an objective in vivo non-invasive research method,but is also of great significance for providing different perspectives on various neurological diseases.Using these technologies,a wealth of imaging data can be obtained,combined with knowledge of mathematics and statistics,to obtain more complete information about brain structure and diseases.Therefore,making full use of and mining the information generated by neuroimaging is of great help to researchers in understanding the pathological changes of neurological diseases,diagnosis and classification,and prognosis prediction.This paper uses the commonly used small animal structure magnetic resonance imaging sequence and human brain scattering tensor imaging sequence to study brain structure under intracranial pathologic changes:Objective:To use VBM technology to explore the changes in brain structure in the pathological progression of a-syn abnormal aggregation rats,and whether these changes in brain structure are related to behaviorMethod:1.Thirty-two healthy adult male SD rats were randomly divided into a 3-month group and a 6-month group,and the two groups were again divided into:PBS injection group and a-syn human full-length fiber injection group.2.Behavioral experiment compares the difference between exercise ability and non-exercise ability of rats.3.All the T2-weighted images of the rat’s head to obtain scanned images that can be used for analysis.4.Use VBM technology to perform gray matter and white matter segmentation and other processing on the T2 images of the rat’s head,and then perform comparative analysis,and visually map the different brain structures.5.After extracting the voxel volume of the different brain structure,analyze the correlation with behavior.6.Observe the expression differences of endogenous p-a-syn(Ser129),DAT and TH in rats by immunohistochemical technique.7.Observe the expression of MAP2 in the substantia nigra,cortex,and hippocampus CA1 regions of rats by immunofluorescence technology.Results:1.The step-by-step experiment,coat hanger experiment,and sucrose experiment were observed to produce differences in the experiment.There was no difference in the Morris water maze test and the apomorphine test without rotation.2.Determine the key parameters such as the layer distance and scanning range(whether to cover the rat cerebellum)of the T2 weighted image of the rat brain.3.Comparing the experimental group with the control group at 6 months after operation,the gray and white matter areas with differences on the T2-weighted image are:neocortex,part of the primary sensory cortex,and white matter in the thalamus.There was no difference between the groups after 3 months.4.The voxel volume changes of the above-mentioned different brain structures are correlated with the behavioral changes of the experimental rats in the 6-month group.5.The results of immunohistochemistry suggested that the expression of endogenous pa-syn(Ser129)increased in the experimental group;the expression of DAT in the striatum decreased in the experimental group of the 3-month group and the 6-month group;TH was in the substantia nigra expression decreased in the experimental group of the 6-month group.6.The immunofluorescence results showed that in the cortex area,the expression of MAP2 decreased in the experimental groups of the two major groups;the experimental group decreased in the substantia nigra area only for 6 months;there was no decrease in the hippocampal CA1 area.Conclusion:1.The study confirmed the feasibility of VBM technology to study the pathological progress of abnormal aggregation of a-syn in vivo.2.The study found that the different brain structures in the 6-month postoperative group were:neocortex,part of the primary sensory cortex,and white matter in the thalamus,and the voxel volume changes of these different brain structures were correlated with the behavioral changes of rats in the experimental group..3.The different brain structure may be the early involvement area on imaging of this pathological progress.Objective:Use HARDI technology to reconstruct,track,and extract feature values of 6 types of 12 white matter fiber structures in the head,and use machine learning methods to study the feature values for complication epilepsy prediction researchMethod:1.Collect DTI data and demographic information of 574 cases of children with intraventricular hemorrhage or/and periventricular leukomalacia and normal group.2.Using HARDI analysis technology to reconstruct,track,and extract feature values of the CC,AF,TC,MCP,SCP and ICP fibers.The phenomenon of neuron migration around the ventricle was observed.3.Use machine learning methods to construct mathematical models(logistic regression,naive Bayes,support vector machines,random forests)for the eigenvalues of the white matter fibers,and explore the discriminative performance of the above models to predict complication epilepsy.4.The logistic regression and other models are further observed for calibration to judge the actual performance of the constructed model.5.Construct the model and analyze the calibration curve according to the different classifications of the data set.Results:1.The CC,AF,TC,MCP,SCP and ICP fibers extracted by HARDI technology have obvious changes in morphology compared with the normal group with the aggravation of the disease grade.2.The constructed logistic regression model performs well on the PVL and IVH combined PVL data sets.3.In PVL disease,corpus callosum(length decrease),right arcuate(length decreas、volume increase),right thalamicortical fiber(volume increase),MCP(volume decrease、right fiber length increase)can be incorporated into the logistic regression prediction model to achieve a better discrimination effect.The accuracy rate is 85%.4.In IVH combined with PVL disease,the corpus callosum(length increase、volume decrease),the right arcuate(length decrease),the left MCP(length increase、FA value decrease),the left ICP(length decreased)and the right SCP(volume decreased)can be incorporated into the logistic regression prediction model to achieve a better discrimination effect.The accuracy rate is 89%.Conclusion:1.The characteristic values of different white matter fibers extracted by HARDI technology can be used to predict complication epilepsy.2.The corpus callosum(length decrease),right arcuate(length decrease、volume increase),right thalamicortical fiber(volume increase),MCP(volume decrease、right fiber length increase)of changes can be used to predict the occurrence of epilepsy in the PVL data set of this paper.3.The corpus callosum(length increase、volume decrease),the right arcuate fasciculus(length decrease),the left MCP(length increase、FA value decrease),the left ICP(length decreased)and the right SCP(volume decreased)of changes can be used to predict the occurrence of epilepsy in the IVH combined PVL data set of this paper. |
PDF Full Text Request