MR diffusion kurtosis and intravoxel incoherent motion related perfusion imagings in assessment and grading of brain tumors

Background: Primary brain tumors refer to a diversified group of benign as well as malignant neoplasms that arise from the parenchymal tissue of the brain or the nearby components. Neoplasms are a crucial cause of morbidity & mortality in adults as well as children, usually result in marked disabilities and causing high burden in the patient’s family along with the health care system. This work aimed to determine the importance of MR Diffusion kurtosis and Intravoxel incoherent motion related perfusion imaging in assessment and grading of brain tumors.Methods: This prospective study was carried out on 50 cases with histopathologically proven, clinically and or MRI manifested with brain tumors and with no comorbidities. Routine MRI, diffusion kurtosis, PWI and IVIM were performed to all patients.Results: Regarding assessment of the brain tumors, the estimated conventional diffusion weighted image (DWI) values with 95.22% sensitivity, 80.67% specificity, 96.5% PPV and 40% NPV, the estimated conventional apparent diffusion coefficient (ADC) values with 87.8% sensitivity, 59.56% specificity, 90% PPV and 50% NPV, conventional T1 with contrast enhancement pattern of the brain tumors with 94.24% sensitivity, 70.44% specificity, 90.1% PPV and 50% NPV, the estimated relative cerebral blood volume (RCBV) values with 96.24% sensitivity, 92.89% specificity, 97.4% PPV and 66.7% NPV, the estimated diffusion kurtosis imaging (DKI) with 97.30% sensitivity, 94.1% specificity, 90.23% PPV and 50% NPV, the estimated advanced diffusion weighted image (IVIM parameter) values with 98.22% sensitivity, 99.23% specificity, 97.5% PPV and 50% NPV, the estimated advanced apparent diffusion coefficient (IVIM parameter) values with 97.83% sensitivity, 91.24% specificity, 90% PPV and 50% NPV, the estimated Diffusion coefficient (D*) (IVIM parameter) values with 89.17% sensitivity, 55.56% specificity, 91.1% PPV and 55% NPV, the estimated perfusion fraction (F%) (IVIM parameter) values with 99.12% sensitivity, 100% specificity, 98% PPV and 79.8% NPV. The estimated DWI values couldn’t predict high grade tumors in our study (P=0.357and AUC=0.650) at cut-off > 0.83 with 90.24% sensitivity, 66.67% specificity, 92.5% PPV and 60% NPV. ADC couldn't predict high grade tumors (P=0.836 and AUC=0.535). (D*) couldn't predict high grade tumors (P=0.752 and AUC=0.556). (F%) predicted high grade brain tumors at cut-off >7.3 with 95.12% sensitivity,100% specificity,100% PPV and 81.8% NPV. RCBV significantly predicted high grade tumors at cut-off >6.2 with 90.24% sensitivity, 88.89% specificity, 97.4% PPV and 66.7% NPV. DKI significantly predicted high grade tumors at cut-off >286 with 90.24% sensitivity, 44.44% specificity, 88.1% PPV and 50% NPV.Conclusions: DKI at cut-off >286, Perfusion fraction at cutoff >7.3, RCBV at cutoff > 6.2 and diffusion enhancement can significantly predict high grades of BTs (p< 0.05). However, DWI, ADC, and diffusion coefficient D* can’t predict high grade of BTs (p< 0.05).