Title: Radiomic Features Predict Ki-67 Expression Level and Survival in Lower Grade Gliomas
|
| Authors: Li, YM; Qian, ZH; Xu, KB; Wang, K; Fan, X; Li, SW; Liu, X; Wang, YY; Jiang, T
|
| Author Full Names: Li, Yiming; Qian, Zenghui; Xu, Kaibin; Wang, Kai; Fan, Xing; Li, Shaowu; Liu, Xing; Wang, Yinyan; Jiang, Tao
|
| Source: JOURNAL OF NEURO-ONCOLOGY, 135 (2):317-324; 10.1007/s11060-017-2576-8 NOV 2017
|
| Language: English
|
| Abstract: To investigate the radiomic features associated with Ki-67 expression in lower grade gliomas and assess the prognostic values of these features. Patients with lower grade gliomas (n = 117) were randomly assigned into the training (n = 78) and validation (n = 39) sets. A total of 431 radiological features were extracted from each patient. Differential radiological features between the low and high Ki-67 expression groups were screened by significance analysis of microarrays. Then, generalized linear analysis was performed to select features that could predict the Ki-67 expression level. Predictive efficiencies were further evaluated in the validation set. Cox regression analysis was performed to investigate the prognostic values of Ki-67 expression level and Ki-67-related radiological features. A group of nine radiological features were screened for prediction of Ki-67 expression status; these achieved accuracies of 83.3% and 88.6% (areas under the curves, 0.91 and 0.93) in the training and validation sets, respectively. Of these features, only spherical disproportion (SD) was found to be a prognostic factor. Patients in the high SD group exhibited worse outcomes in the whole cohort (overall survival, p < 0.0001; progression-free survival, p < 0.0001). Ki-67 expression level and SD were independent prognostic factors in the multivariate Cox regression analysis. This study identified a radiomic signature for prediction of Ki-67 expression level as well as a prognostic radiological feature in patients with lower grade gliomas.
|
| ISSN: 0167-594X
|
| eISSN: 1573-7373
|
| IDS Number: FL4PH
|
| Unique ID: WOS:000414212300011
|
| PubMed ID: 28900812
*Click Here to View Full Record
