LAUSR

OBJECTIVE Meningioma grading is a relevant subject with respect to therapy decisions in complete or partial resection, observation and radiotherapy as higher grades are associated with tumour growth and recurrence. The differentiation of low and intermediate grades is particularly challenging. This study attempts to apply radiomics-based shape and texture analysis on routine multiparametric MRI from different scanners and institutions for grading. METHODS We utilized MR imaging data (T1-/T2-weighted, T1-weighted-contrast-enhanced [T1CE], FLAIR, DWI, ADC) of grade I (n=46) and grade II (n=25) non-treated meningiomas with histologic work-up. Two experienced radiologists performed manual tumour segmentations on FLAIR, T1CE and ADC images in consensus. The MR imaging data were preprocessed through T1CE&T1-subtraction, co-registration, resampling, and normalization. PyRadiomics-package was used to generate 990 shape/texture features. Step-wise dimension reduction and robust radiomics feature selection were performed. Biopsy results were used as standard of reference. RESULTS Four statistically independent radiomics features were identified as showing the strongest predictive values for higher tumour grades: roundness-of-FLAIR-shape (area-under-curve [AUC]: 0.80), cluster-shades-of-FLAIR/T1CE-grey-level (AUC: 0.80), DWI/ADC-grey-level-variability (AUC: 0.72), and FLAIR/T1CE-grey-level-energy (AUC: 0.76). In a multivariate-logistic-regression-model, the combination of the features led to an AUC of 0.91 for the differentiation of grade I and grade II meningiomas. CONCLUSIONS Our results indicate, that radiomics-based feature analysis applied on routine MR is viable for meningioma grading, and a multivariate-logistic-regression-model yielded strong classification performances. More advanced tumour stages are identifiable through certain shape parameters of the lesion, textural patterns in morphological MR sequences, and DWI/ADC-variability.