Related Experiment Video
Updated: Jul 7, 2025
Glioblastoma Relapse Post-Resection Model for Therapeutic Hydrogel Investigations
Published on: February 24, 2023
The development of a combined clinico-radiomics model for predicting post-operative recurrence in atypical
Leihao Ren1,2,3, Jiawei Chen1,2,3, Jiaojiao Deng1,2,3
1Department of Neurosurgery, Huashan Hospital, Shanghai Medical College, Fudan University, Shanghai, China.
Purpose:
Atypical meningiomas could manifest early recurrence after surgery and even adjuvant radiotherapy. We aimed to construct a clinico-radiomics model to predict post-operative recurrence of atypical meningiomas based on clinicopathological and radiomics features.
Materials And Methods:
The study cohort was comprised of 224 patients from two neurosurgical centers. 164 patients from center I were divided to the training cohort for model development and the testing cohort for internal validation. 60 patients from center II were used for external validation. Clinicopathological characteristics, radiological semantic, and radiomics features were collected. A radiomic signature was comprised of four radiomics features. A clinico-radiomics model combining the radiomics signature and clinical characteristics was constructed to predict the recurrence of atypical meningiomas.
Results:
1920 radiomics features were extracted from the T1 Contrast and T2-FLAIR sequences of patients in center I. The radiomics signature was able to differentiate post-operative patients into low-risk and high-risk groups based on tumor recurrence (P < 0.001). A clinic-radiomics model was established by combining age, extent of resection, Ki-67 index, surgical history and the radiomics signature for recurrence prediction in atypical meningiomas. The model achieved a good prediction performance with the integrated AUC of 0.858 (0.802-0.915), 0.781 (0.649-0.912) and 0.840 (0.747-0.933) in the training, internal validation and external validation cohort, respectively.
Conclusions:
The present study established a radiomics signature and a clinico-radiomics model with a favorable performance in predicting tumor recurrence for atypical meningiomas.
