Diagnostic accuracy of perfusion weighted MRI in differentiating neoplastic and non-neoplastic brain lesions using histopathology as Gold standard.
Objectives: Perfusion-weighted imaging may be performed as a complementary examination to conventional MRI techniques. The clinical applications of PWI in the evaluation of focal brain lesions include the differentiation between neoplastic and non-neoplastic brain lesions, primary tumors and solitary metastasis and, in the post-treatment follow up, the differentiation of tumoral recurrence and radio necrosis. To determine the diagnostic accuracy of perfusion weighted MRI in differentiating. Neoplastic and non-neoplastic brain lesions keeping histopathology as gold standard. Study Design: Cross sectional (validation) study. Setting: The study was conducted in Radiology department of Allied Hospital Faisalabad. Period: Six months from 25-March-2016 to 24-Sep-2016. Material and Methods: Permission for this study was taken from the hospital ethical review committee. A total of 125 patients were included and MRI examination was performed with the patients in supine position using a 1.5-T Philips MRI unit and a body phased-array coil. A preload of paramagnetic contrast agent (gadolinium) was administered 30 seconds before acquisition of dynamic images, followed by a standard dose 10 seconds after starting imaging acquisitions. Results: Patients ranged between 15-65 years of age. Mean age of the patients was 47.6±10.6 years. There were 65 males (52%) and 60 females (48%). PW MRI showed sensitivity 81.67%, specificity 81.54%, positive predictive value 80.33%, negative predictive value 82.81% and diagnostic accuracy 81.60%. ROC and likelihood ratio was measured for age and gender. For age (15-40 years) likelihood ratio was 7.187 and for age 41-65 likelihood ratio was 48.665. For males likelihood ratio was 31.759 while for females 16.188. Conclusion: In conclusion, our results suggest a promising role for perfusion MR imaging in the distinction between neoplastic and non-neoplastic lesions.