Abstract
BackgroundMeningioma invasion can be preoperatively recognized by radiomics features, which significantly contributes to treatment decision-making. Here, we aimed to evaluate the comparative performance of radiomics signatures derived from varying regions of interests (ROIs) in predicting BI and ascertaining the optimal width of the peritumoral regions needed for accurate analysis.MethodsFive hundred and five patients from Wuhan Union Hospital (internal cohort) and 214 cases from Taihe Hospital (external validation cohort) pathologically diagnosed as meningioma were included in our study. Feature selection was performed from 1,015 radiomics features respectively obtained from nine different ROIs (brain-tumor interface (BTI)2–5mm; whole tumor; the amalgamation of the two regions) on contrast-enhanced T1-weighted imaging using least-absolute shrinkage and selection operator and random forest. Principal component analysis with varimax rotation was employed for feature reduction. Receiver operator curve was utilized for assessing discrimination of the classifier. Furthermore, clinical index was used to detect the predictive power.ResultsModel obtained from BTI4mm ROI has the maximum AUC in the training set (0.891 (0.85, 0.932)), internal validation set (0.851 (0.743, 0.96)), and external validation set (0.881 (0.833, 0.928)) and displayed statistically significant results between nine radiomics models. The most predictive radiomics features are almost entirely generated from GLCM and GLDM statistics. The addition of PEV to radiomics features (BTI4mm) enhanced model discrimination of invasive meningiomas.ConclusionsThe combined model (radiomics classifier with BTI4mm ROI + PEV) had greater diagnostic performance than other models and its clinical application may positively contribute to the management of meningioma patients.
Highlights
Brain invasion (BI), described as abnormal tumor projections into the basal parenchyma, which lacks an overlapping layer of the leptomeninges, has previously been considered to possess therapeutic predictive and prognostic benefits in the management of meningiomas [1,2,3,4,5,6,7]
Patients were included in the study if the following criteria were met: (a) recently diagnosed intracranial meningiomas awaiting resection; (b) undergone relevant MRI and blood testing in the same hospital, 4 and 2 weeks prior to surgery; (c) MR images were accessible in DICOM format, enveloping T1WI and T2 fluid-attenuated inversion recovery (FLAIR) and contrast-enhanced T1-weighted (CE-T1WI) sequences; (d) MR images were free of visible artefacts or spatial distortions; (e) full blood test results were accessible in the electronic medical database system
We discovered that BTI4mm regions of interests (ROIs) had a higher Receiver operator curve (ROC)-area under curve values (AUC) and Precision-recall curve (PRC)-AUC, compared with other models both in the training set, internal validation set, and external validation set and acquired the highest AUC value of 0.891, 0.851, and 0.881, respectively
Summary
Brain invasion (BI), described as abnormal tumor projections into the basal parenchyma, which lacks an overlapping layer of the leptomeninges, has previously been considered to possess therapeutic predictive and prognostic benefits in the management of meningiomas [1,2,3,4,5,6,7]. Though preoperative BI detection is critically important, it continues to present challenges in clinical practice In this respect, quantitative image analysis, or radiomics, has displayed potential in BI detection. The situation is comparable in invasive meningiomas, wherein diffuse (single tumor cells extending into the encircling parenchyma) and cluster-like (clustered islands or nests of tumor cells) invasive patterns are sometimes too cryptic to be recognized during medical imaging [10]. When using this analytical technique, it remains important to include the peritumoral regions in the radiomics analysis to effectively identify meningioma invasiveness [11]. We aimed to evaluate the comparative performance of radiomics signatures derived from varying regions of interests (ROIs) in predicting BI and ascertaining the optimal width of the peritumoral regions needed for accurate analysis
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
