Texture Feature Ratios from Relative CBV Maps of Perfusion MRI Are Associated with Patient Survival in Glioblastoma

Document Type


Publication Date


Publication Title

AJNR. American journal of neuroradiology


BACKGROUND AND PURPOSE: Texture analysis has been applied to medical images to assist in tumor tissue classification and characterization. In this study, we obtained textural features from parametric (relative CBV) maps of dynamic susceptibility contrast-enhanced MR images in glioblastoma and assessed their relationship with patient survival.

MATERIALS AND METHODS: MR perfusion data of 24 patients with glioblastoma from The Cancer Genome Atlas were analyzed in this study. One- and 2D texture feature ratios and kinetic textural features based on relative CBV values in the contrast-enhancing and nonenhancing lesions of the tumor were obtained. Receiver operating characteristic, Kaplan-Meier, and multivariate Cox proportional hazards regression analyses were used to assess the relationship between texture feature ratios and overall survival.

RESULTS: Several feature ratios are capable of stratifying survival in a statistically significant manner. These feature ratios correspond to homogeneity (P = .008, based on the log-rank test), angular second moment (P = .003), inverse difference moment (P = .013), and entropy (P = .008). Multivariate Cox proportional hazards regression analysis showed that homogeneity, angular second moment, inverse difference moment, and entropy from the contrast-enhancing lesion were significantly associated with overall survival. For the nonenhancing lesion, skewness and variance ratios of relative CBV texture were associated with overall survival in a statistically significant manner. For the kinetic texture analysis, the Haralick correlation feature showed a P value close to .05.

CONCLUSIONS: Our study revealed that texture feature ratios from contrast-enhancing and nonenhancing lesions and kinetic texture analysis obtained from perfusion parametric maps provide useful information for predicting survival in patients with glioblastoma.

Medical Subject Headings

Adult; Aged; Brain Neoplasms; Female; Glioblastoma; Humans; Image Enhancement; Image Interpretation, Computer-Assisted; Kaplan-Meier Estimate; Magnetic Resonance Imaging; Male; Middle Aged; Multivariate Analysis; Proportional Hazards Models; ROC Curve

PubMed ID






First Page


Last Page