The Value Of Mri EDWI Combined With Image Texture Features In The Evaluation Of Vascular Architecture In Peripheral Lung Cancer

Objective:to explore the value of mri eDWI combined with image texture features in the evaluation of vascular architecture of peripheral lung cancer,so as to provide valuable imaging basis for the diagnosis,treatment and prognosis of lung cancer.Materials and methods:a total of 12 patients with peripheral lung cancer who visited our hospital from August 2017 to August 2018 and obtained postoperative pathology were collected.All the patients were found with chest nodules or masses by CT scanning in our hospital or other hospitals,and mri scanning was performed in our hospital.The machine model was GE Discovery750 3.0t.Conventional T1WI,T2WI axial position and fat suppression technology,T2WI coronal position,eDWI sequence(b value=0,30,50,100,150,200,300,500,800,1000,1500s/mm~2).The scanned image was transferred to GE ADW 4.6 workstation.MADC module in Function Tools was used to calculate multi-b value image parameters,and Standard ADC(Standard diffusion coefficient),Slow ADC(Slow diffusion coefficient),Fast ADC(Fast diffusion coefficient)and F value(Fast diffusion fraction)were obtained.Will get medical hui shadow eWDI images uploaded to remit group radiology cloud platform(http://mics.radcloud.cn/),get second order character(GLCM gray level co-occurrence matrix,gray-level run-length matrix GLRLM step)and higher order features(NGTDM neighborhood grayscale difference matrix and gray area the size of the matrix GLSZM),calculate the average under different b values texture characteristic value,and calculate the average texture characteristic value of each coefficient of variation CV,the texture feature of variation coefficient is greater than 1,select corresponding texture feature;CD31,VEGF and SMA staining were performed on 12 lung cancer samples by immunohistochemical method,and CD31 expression,VEGF expression and SMA expression were measured by Image J,and MPI was calculated.Statistical methods were used to analyze the correlation between eDWI parameters,image texture characteristics and CD31 expression,VEGF expression/score,SMA expression and MPI.Results:1.EDWI parameters of 12 patients with lung cancer were not significantly correlated with CD31 expression,SMA expression and MPI(P>0.05).2.There was no significant correlation between Standard ADC,Slow ADC and Fast ADC and VEGF expression/score in eDWI parameters of 12 patients with lung cancer(P>0.05).F value was positively correlated with VEGF score,r=0.592,P=0.042<0.05.3.VEGF expression and CD31 expression,CD31 expression and SMA expression were positively correlated in angiogenic parameters,with r values of0.718 and 0.849,and P values of 0.009 and 0.000,respectively(P<0.05).VEGF score and MPI values were negatively correlated,with r values of-0.615 and P values of 0.033<0.05.4.eDWI has 7 texture feature in images(Joint Entropy,Inverse Difference Moment Normalized,Inverse Difference Normalized,Correlation,Sum Entropy,Run Entropy,Zone Entropy)and negatively correlated with the MPI,the rest of the image texture features are no obvious correlation with vascular construct parameters.Conclusion:1.There is a correlation between the F value in the eDWI parameter value of peripheral lung cancer and VEGF score,and the F value can indirectly reflect the vascular architecture information of peripheral lung cancer to some extent.2.Multiple texture features of eDWI images are correlated with MPI,and image texture features can indirectly reflect vascular architecture information of peripheral lung cancer to some extent.3.Magnetic resonance eDWI technology combined with image texture feature analysis can indirectly reflect more vascular architecture information of peripheral lung cancer.