Radiomic signatures of meningiomas using the Ki-67 proliferation index as a prognostic marker of clinical outcomes

OBJECTIVE The clinical behavior of meningiomas is not entirely captured by its designated WHO grade, therefore other factors must be elucidated that portend increased tumor aggressiveness and associated risk recurrence. In this study, the authors identify multiparametric MRI radiomic signatures using Ki-67 as a prognostic marker outcomes independent grade. METHODS A retrospective analysis was conducted all resected between 2012 2018. Preoperative MR images were used for high-throughput feature extraction subsequently to develop machine learning algorithm stratify based on indices < 5% ≥ 5%, Progression-free survival (PFS) assessed prediction strata compared with histopathological Ki-67. RESULTS Three hundred forty-three included: 291 grade I, 43 II, 9 III. overall rate recurrence 19.8% (15.1% in 44.2% 77.8% III) over median follow-up 28.5 months. Grade II III tumors had higher than I tumors, albeit peritumoral edema volumes considerable variation meningioma Forty-six high-performing features (1 morphological, 7 intensity-based, 38 textural) identified build support vector model cutoff resultant areas under curve 0.83 (95% CI 0.78–0.89) 0.84 0.75–0.94) achieved discovery (n = 257) validation 86) data sets, respectively. Comparison versus learning–predicted showed excellent performance (overall accuracy > 80%), classification exhibiting greatest accuracy. Prediction classifier revealed shorter PFS (p 0.0001, log-rank test), which corroborates divergent patient observed CONCLUSIONS proliferation index may serve surrogate Machine preoperative Ki-67, provides enhanced analytical insights help improve diagnostic guide patient-specific treatment strategies.