Multiparametric MRI for the differentiation of brain glial cell hyperplasia from low-grade glioma

Abstract Background For individualized clinical treatment of patients,The differentiation between brain glial cell hyperplasia and low-grade glioma is of significant importance. Objective Based on Multiparametric MRI images,combining with clinical risk factors,we construct a radiomics-clinical model and nomogram for the differentiation of brain glial cell hyperplasia from low-grade glioma. Methods We retrospectively included patients with brain glial cell hyperplasia and low-grade glioma who underwent surgery at the First Affiliated Hospital of Soochow University from March 2016 to March 2022.A total of 41 patients of brain glial cell hyperplasia and 87 patients of low-grade glioma were included in this study,then divide them into training group and validation group randomly with a ratio of 7: 3.We extracted radiomics features from T1-weighted imaging(T1WI),T2-weighted imaging(T2WI), diffusion-weighted imaging(DWI), contrast-enhanced T1-weighted imaging (T1-enhanced),then built LASSO, SVM and RF model,and we selected a model with higher efficiency to calculate the Rad-score (radiomics score) of every patient. To obtain the independent risk factors,we screened the Rad-score and clinical risk factors by univariate and multivariate logistic regression analysis in turn, then we constructed radiomics-clinical model, and evaluated their performance. Results Of the included 128 cases ,brain glial cell hyperplasia and low-grade gliomas were randomly divided into 10 groups, and 7 of them were used as training group and 3 as validation group. The radiomics-clinical model were constructed with two independent risk factors——mass effect and Rad-score,which AUCs of the training group and validation group were 0.847 and 0.858. The diagnostic accuracy, sensitivity, and specificity of the validation group were 0.821,0.750,0.852. Conclusion Combining with radiomics constructed by multiparametric MRI images and clinical features,the radiomics-clinical model and nomogram which were constructed to differentiate between brain glial cell hyperplasia and low-grade glioma had a good performance.