| Purpose To explore the changes of imaging characteristics(number,length and curvature etc.)of lenticulostriate arteries(LSAs)of normal volunteers and high-risk volunteers(hypertension,hyperlipidemia,diabetes,heart disease,history of large artery infarction,smoking and drinking,etc.)with the increase of age;to explore the interactive relationship between the number,volume of lacunar infarctions(LIs)and the imaging characteristics of LSAs(the number and length etc.)in patients with LIs in basal ganglia infarction;to explore the relationship between the plaque characteristics of internal carotid artery(ICA)and / or middle cerebral artery M1segment(MCA-M1)and the LIs in basal ganglia and the imaging characteristics of LSAs.Materials and Methods Part one: Retrospective analysis of 200 volunteers,and 108 normal volunteers and 63 volunteers with high-risk of stoke were included.DWI,T1 WI,T2WI,T2-FLAIR,SWI and HR-VWI were performed;The number and length of lenticular arteries were measured repeatedly on the VWI image of minimum intensity project(min IP)by two radiologist for consistency appraisal tested by ICC analysis;SWI was used to evaluate thalamostriate vein good or poor visualization;The one way ANOVA analysis of variance was used to test the difference of BMI,total number and total length of LSAs among the three groups of normal volunteers(< 30 years old,30-50 years old and > 50 years old),and student Newman Keuls was used to test the multiple differences;Non parameter Kruskal Wallis H test was used to test the difference of age and curvature among the three groups;Chi-square was used to test the difference of gender,medial group,LSAs dilation and thalamostriate vein poor visualization;Mann Whitney U was used to test the BMI,total number and total length of LSAs among high-risk volunteers < 52 years old and ≥ 52 years old;Chi-square test was used to detect gender,medial group,LSAs dilation,thalamostriate vein poor visualization,and clinical baseline high-risk factors(hypertension,hyperlipidemia,diabetes,etc.);Pearson correlation analysis was usedto detect normal and high-risk volunteers Correlation analysis was used to detect the correlation between LSAs image characteristics(total number and length of LSAs etc.)of high-risk volunteers and age,and partial correlation analysis was used to correct the correlation.Part two: 237 cases of cerebral ischemic stroke or transient ischemic attack were analyzed retrospectively,among which 56 cases were in the LIs group and 44 cases were in the non LIs group of basal ganglia.The scanning sequence of subjects included DWI,T1 WI,T2WI,T2-FLAIR,VWI plain scan and enhancement.Using neuspin(Neusoft spin,Shenyang,China)software to convert VWI raw image to coronal and carry out min IP to get post-processing image.One radiologist and one neurosurgeon were used to evaluate the LSAs and LIs and ICA and/or MCA-M1plaques;ICC coefficient was used to test the consistency of the measurement of LSAs of two physicians;Chi-square test was used to compare the differences of baseline clinical characteristics between patients with and without LIs in basal ganglia,include gender,hypertension,hyperlipidemia,diabetes,coronary heart disease,large artery infarction,smoking and drinking history;Non parameter Mann Whitney U-test or Fisher’s exact test were used to compare the age,total number,total length,average length and visual evaluation of LSAs between the LIs group and the non LIs group;Binary logistic regression analysis was used to find out the independent predictors of LIs group;Paired t-test was used to detect the differences of total number,total length and average of LSAs in bilateral basal ganglias,as well as ipsilateral ICA and MCA-M1 plaque assessment;Mann Whitney U was used after further grouping,and use multifactor logistic regression analysis to find out the risk factors of basal ganglia LIs;After further grouping,pearson correlation analysis was used to detect the correlation of two continuous variables,otherwise spearman correlation analysis was used;Stepwise linear regression analysis to find out the independent prediction factors of LSAs.Results Part one: 108 normal volunteers and 63 high-risk of stroke volunteers were included in this study.The consistency of LSAS quantity and length(0.738-0.977)between the two measurements was excellent.The normal volunteers were divided into threegroups of < 30 years old(totally 57 volunteers,14 men and 43 women,mean age25.5±2.6 years old),30-50 years old(totally 25 volunteers,12 men and 13 women,mean age 39.4±7.1 years old)and > 50 years old(totally 26 volunteers,10 men and16 women,mean age 59.5±6.4 years old).The total number of LSAs(P = 0.018)and total length(P = 0.020)were statistically different,and gradually decreased before 52 years old,then increased slightly.The total number(r =-0.289,P = 0.002)and total length(r =-0.303,P = 0.001)of LSAs of normal volunteers were negatively correlated with age.After adjustment of partial correlation analysis,only the linear length of LSAs(r =-0.385,P = 0.025)was negatively correlated with age.High-risk volunteers were divided into <52(totally 16 volunteers,11 men and 5 women,mean age 37.3±6.2 years old)and ≥52 years old groups(totally 47 volunteers,25 men and22 women,mean age 60.7±6.6 years old),the total number of LSAs(P = 0.037),history of diabetes mellitus(P = 0.025),history of LIs in basal ganglia(P = 0.032),and thalamostriate vein poor visualization(P = 0.050)were negatively correlated with age in high-risk volunteers,and the total number and length of LSAS increased slightly before 41 years old,and then decreased gradually.There was a negative correlation between the total number of LSAs and age(r =-0.267,P = 0.035)in high-risk volunteers.After adjustment of partial correlation analysis,age had no direct relationship with the total number(r =-0.033,P = 0.871)and total length(r =-0.007,P = 0.972)of LSAs,but there was a significant positive correlation between the LIs history of basal ganglia(r = 0.405,P = 0.040)and the wall thickening of MCA-M1(r = 0.408,P = 0.039).Part two: 237 patients with TIA or stroke were analysed retrospectively,and 100 patients were enrolled finally,namely 56 patients with LIs and 44 without.The results showed that there was a good agreement among the observers(0.585-0.994);The incidence of male(P = 0.046),baseline NIHSS(P = 0.010),hypertension(P = 0.049)and the history of large artery infarction(P = 0.009)was higher in patients with basal ganglia LIs,and the total number(P = 0.007),total length(P = 0.001),average length(P = 0.011)and visual evaluation(P = 0.001)was lower;After binary logistic regression analysis,the shortening of the mean length of LSAs may increase the incidence of LIs in basal ganglia(or 8.98,95% CI 1.37-59.11,P = 0.022).100 patients were sub-divided into 84 basal ganglia and 116 basal ganglia.By binary logistic regression analysis,less total LSAs(OR of 1.54,P = 0.002),more ipsilateral ICA plaques(OR of 0.50,P = 0.009),more ipsilateral MCA-M1 upper wall plaques(OR of 0.33,P = 0.016)and lower wall plaques(OR of 0.31,P = 0.003)increased the incidence of LIs in basal ganglia;After multiple regression analysis,it was found that the visual assessment of LSAs,the volume of LIs and the presence of ipsilateral MCA-M1 plaques were independent predictors of LSAs(P < 0.05).Conclusion The decrease of the number and length of LSAs in high-risk stroke volunteers with the increase of age was presented earlier than that in normal stroke volunteers;The increase of age in high-risk stroke volunteers did not directly affect the total number and length of LSAs,but will cause more LIs in basal ganglia and severer thickening of the MCA-M1,and then affect the LSAs.Less total number of LSAs and severer ipsilateral ICA / MCA-M1 plaques promoted the LIs in basal ganglia;The number and volume of LIs and ipsilateral MCA-M1 plaques affected the total number and length of LSAs.Larger volume of LIs and more plaques on the upper and lower walls of MCA-M1 could predict the decrease of LSAs. |
PDF Full Text Request