Dynamic contrast enhanced (DCE) MRI derived perfusion parameters are widely used for glioma grading. Several studies have shown that relative blood volume (rCBV) derived from DCE-MRI correlates excellently with grade of glioma ^1-3. Traditionally rCBV is calculated by dividing CBV value of tumor region with the CBV value from the corresponding contra-lateral region by identifying and placing region of interest (ROI) ^1-3. This technique is tedious when contra-lateral side is also involved by the tumor or vasculatures contaminate the ROI placement and needs user expertise. The main aim of this study was to develop an automatic method to normalize CBV so that the user-induced biasness in glioma grading due to ROI placement can be reduced.A total of 20 histologically conformed glioma patient data (10 high grade, 10 low-grade) were included retrospectively in this study after the approval of institutional ethics committee. DCE-MRI images were acquired for all patients on a 3T MRI scanner (Philips, The Netherlands). Imaging protocol: DCE-MRI (TR/TE=4.4/2.1ms, 10o flip angle, 240 × 240 mm² FOV, 128×128 matrix, 12 slice with 6mm thickness, 32 dynamic with 3.9s temporality, contrast dose 0.1 mmol/kg body weight, 3.5 ml/sec injection rate, contrast used Gd-BOPTA), T1 weighted inversion recovery (TR/TE/IT=2064/20/800ms, 5mm slice, 400x340 matrix, 230x230 mm² FOV, 90 flip angle). CBV map were quantified from the DCE-MRI data. T1 weighed image was registered with the first stack of the dynamic data and then segmented into grey matter, white matter and CSF using SPM8. Histogram of CBV values from segmented normal grey and white mater region were extracted, mean and standard deviation of CBV were estimated by fitting normal distributions to the measured histograms. The CBV map of whole stack was then divided by estimated mean CBV value from normal grey and white region to generate normalized CBV maps. Color maps of normalized CBV were generated for better visualization.Sensitivity and specificity of normalized CBV in glioma grading is shown in Table 1. CBV normalized with normal white matter gives better results as compared to grey matter. CBV normalized by white matter and grey matter of a high grade glioma patient is shown in Figure 1.Normalized CBV provides a better contrast between tumor and normal region. The technique reduces the user dependency and biasness in glioma grading. Our study suggest the automatic normalization technique may improve glioma grading decision making for clinicians; however a rigorous study for the technique with increased sample size is required.