Diagnosis of brain tumor with multi-modality imaging of 11C-MET PET combined with functional MRI

Abstract

Objective To evaluate the diagnostic efficiency of the parameters obtained from PET/MR in brain tumors. Methods In this prospective study, 28 patients (21 males, 7 females, age range: 6-82 years) with clinical suspicion of brain tumor from November 2012 to September 2015 underwent PET/MR multi-modality imaging. The examination of PET/MR included 11C-MET PET and multiple MR sequences. The qualities of images were estimated firstly. The ROC curve and the accuracy of SUVmax,ADCmean,rCBF and NAA/(Cho+ Cre) ratio were calculated. The pathology or final clinical diagnosis was taken as the standard. The diagnostic efficiency of the multi-modality imaging was determined based on the cutoff values of the four parameters. Two-sample t test was used to analyze the differences of parameters between glioma group and inflammatory group. Results SUVmax,ADCmean,rCBF and NAA/(Cho+ Cre) ratio were validated to be effective parameters in diagnosing brain tumors with the diagnostic accuracy of 89.3%(25/28), 82.1%(23/28), 78.6%(22/28) and 75.0%(21/28), respectively. The SUVmax exhibited the highest diagnostic accuracy, while the diagnostic efficiency of the combination of four parameters was superior to the separate parameter. The values of SUVmax,rCBF and NAA/(Cho+ Cre) ratio were significantly different between glioma group(n=10) and inflammatory group(n=11; t values: -2.31, -3.11, -2.77, all P<0.05). Conclusions PET/MR can provide a one-stop examination with multi-modality imaging of brain. The obtained parameters SUVmax,ADCmean,rCBF and NAA/(Cho+ Cre) ratio, especially their combination, have effective diagnostic values on brain tumor. Key words: Brain neoplasms; Positron-emission tomography; Magnetic resonance imaging; Methionine