Morphological Analysis Of Brain Structure MRI Of Healthy People Based On Deep Learning

Objective:To establish the reference range of basic data of brain volume of whole brain and subcortical structure of normal adults in northern Shaanxi,and to conduct early screening and diagnosis of neuropsychiatric diseases related to pathological brain atrophy in northern Shaanxi from the direction of medical imaging.To analyze the changes of brain volume with age in normal adults during aging,and to explore the interaction of subcortical structure on sex and age,so as to further explore the law and mechanism of rational brain atrophy in healthy life between 20 and 80 years old.Materials and Methods:From November 2019 to February 2021,275 cases of brain MRI(magnetic resonance imaging)scanning were collected from the affiliated Hospital of Yan’an University,34 cases were excluded,and a total of 241 cases were included,including117 males and 124 females.According to their age,they were divided into three groups:20-39 years old group,40-59 years old group and 60-79 years old group.The corresponding structural boundaries were automatically drawn layer by layer by using the deep learning quantitative software U-AI,and the volumes of the whole brain and subcortical structures,including absolute volume and relative volume,were automatically measured,so as to determine the reference range of brain structures of different ages and genders.SPSS20.0 was used to analyze the changes of whole brain and subcortical volume among different sex and age groups by t-test and analysis of variance statistics.Correlation and regression analysis were used to determine the relationship between the relative volume of whole brain and subcortical structure,subcortical laterality index and the proportion of subcortical structure in whole brain with age.Results:1.There were no significant differences in sex constituent ratio,age and education level among the three groups.(P > 0.05).2.The absolute volume of whole brain tissue,whole brain gray matter,whole brain white matter and whole cerebrospinal fluid was larger in males,which was closely related to intracranial volume,and the relative volume was the largest in gray matter,the second in white matter,and the smallest in cerebrospinal fluid,and with the increase of age,gender differences gradually appeared in whole brain tissue,whole brain gray matter and cerebrospinal fluid,and the proportion of females whole brain tissue and whole brain gray matter was larger,while that of cerebrospinal fluid was larger in males.3.Two-way ANOVA was used to compare the whole brain and subcortical volume differences in 20-39 years old group,40-59 years old group and 60-79 years old group.It was found that most of the subcortical structures except caudate nucleus,there was a significant age main effect between the two groups(P < 0.05).In addition,the age × sex interaction of hippocampus and other basal ganglia was found.4.There was a significant negative correlation between gray matter volume and age,cerebrospinal fluid,white matter,hippocampal volume and GM/WM ratio,and a significant quadratic correlation model with age.In addition,in the subcortical structure(R2≥1.6,P<0.05).In addition,in the subcortical structure,except caudate nucleus,the volume of basal nuclei such as putamen and thalamus was significantly negatively correlated with age(P< 0.05),especially in thalamus and putamen,which were higher than the average rate of global brain atrophy.At the same time,it was also found that the hippocampus had significant right laterality(P<0.05).Conclusion:1.The whole brain and subcortical volumes of healthy adults of different genders and different age groups were measured and analyzed by using the neural quantitative software of deep learning algorithm,and the reference ranges of whole brain and subcortical structural volumes of different genders at different ages were determined.it is helpful to distinguish normal aging from abnormal brain atrophy.2.After the brain volume was normalized and the relative volume was used to eliminate the influence of intracranial volume,the linear model of gray matter and the quadratic model of white matter,cerebrospinal fluid and GM/WM ratio were found in the age track of whole brain volume,and different age trajectory patterns were found in subcortical structures,such as the quadratic model of hippocampus and the linear model of thalamus and putamen.Thus,an important regression model is used to explain the changes of brain aging in imaging.3.It is found that hippocampal laterality and sex × age interaction.Understanding these changes may be of certain reference significance to the course,diagnosis,treatment and prognosis of diseases related to hippocampal volume atrophy(PD and AD,etc.).