Data Analysis Of Developing Rat Brain Imaging And Its Application

Investigation of the brain development is important in preventing,diagnosing,intervening and therapy of human brain diseases.The non-invasive medical imaging techniques have been widely used in the researches on human brain development.Imaging experiment of animals could be used to investigate the influence of gene,environment and drug in brain development,which plays an irreplaceable role in the investigations of brain development.Because of its advantages of abundant disease models,short growth cycle and so on,rat is a frequently used animal model in study of brain development.The imaging data analysis methods of adult rats have been widely established until now.However,because of the considerable differences between infancy and adult rats in the shape,size and structure of brain,the existing data analysis methods of adult rat brain images are not suitable for the infancy rat brain images.Therefore,this study aimed to establish a software platform for the multimodal image analysis of developing rat,including voxel-wise data analysis platform and quantitative analysis platform based on region of interest(ROI).According to its features,the growth of the rat is divided into three stages,including neonatal stage(P7),infancy stage(P21)and adult stage(P60).Voxel-wise analysis is the most widely used data analysis method,with which we can perform voxel-by-voxel statistical analysis.The standard brain template and atlas are important tools for performing voxel-wise analysis.Because the size and shape of each rats are different,it is necessary to standardize the brain images of each rat into a unified space firstly in voxel-wise analysis.The standard brain template offers a unified reference space for normalization of subjects' images.The standard brain atlas which is in the same space of the standard brain template can provide precious location of analysis results.Therefore,we established the developing rat standard brain templates based on multimodal brain images of developing rats.The developing rat standard brain templates contain T2-weighted structural template of developing rat brain,DTI white matter template of developing rat brain and FDG-PET functional template of developing rat brain.In this study,we constructed the standard brain atlases of developing rats at each stage,which had detail brain regions and were registered with the developing rat brain template at the corresponding developing stage.Voxel-wise statistical analysis platform of developing rats was constructed through embedding the above templates and atlases into SPM,and modifying its source programs.To test the feasibility of this voxel-wise statistical analysis platform,we performed quantitative assessment on accuracy of spatial normalization,voxel-wise statistical analysis results and results location through simulation data.And the results show that the platform can detect simulated lesion well.ROI-based quantitative statistical analysis method is another widely used analysis methods of brain image.Accurately extracting the ROI is the key to doing the followup analysis.The traditional quantitative statistical analysis methods based on ROI are manually delineated by researchers with anatomical knowledge,which is timeconsuming,labor-intensive,subjective,and poorly reproducible.Based on the established standard brain templates and atlases of developing rats,a method of automatically extracting ROI was proposed.The results of this method were compared with the results of the expert's manual delineation which was used as criteria.The Dice coefficient and relative error on volume were selected as the paratemers to evaluate the comparation.Results show that the ROIs extracted by the automatical method is accurate,which is high consist with the ROIs extracted by expert.The developing rat brain imaging data analysis platform could be used to not only investigate the developing rules of rat brain structure and function,but also analyze the rat brain imaging data of a specific developing stage to study the rat brains' structure and function in the developing stage.With the platform,we analyzed the developing rat brain images,including T2 weighted magnetic resonance image,diffussion tensor image and FDG-PET image,to explore the developing rules of brain sturcure and function.In addition,we studied the abnormal brain functional acitivity of ischemic stroke in acute stage based on the middle cerebral artery occlusion(MCAO)rat model in adult stage.In summary,a multimodal brain imaging data analysis platform for developing rats has been established based on the developing rat model.With the platform,we investigated the development of rat brain structure and function,which could promote the understanding of brain development and provide a reference for prevention of brain diseases during development.