The Clinical implication of trimodality PET/CT-MR in head and neck tumor

Objectives To compare the diagnostic efficacies of PET/CT, MRI and trimodality PET/CT-MR.

Methods Fourteen consecutive patients (2 benign and 12 malignancies) with head and neck tumor were assessed with trimodality PET/CT-MR .PET/CT scan first , and then MRI conventional sequence and diffusion weighted imaging (DWI) with multi b were performed. Reviewed all data of tumors,including standardized uptake value (SUVmax and SUVmean) and apparent diffusion coefficient (ADCmean,ADCmax and ADCmin) used multiple index model (small b 0-200, medium b 300-1500 and high b1700-4500 s/mm2) . Two blinded and independent readers assessed all images randomly to decide the malignant or benign. Using SPSS16.0 statistics software with Pearson’s correlation test to detect the relationships between SUV and ADC.

Results The mean SUVmax ±SD in patients with a benign and malignancy is 5.10±1.80and 11.02±5.62, there is the significant between benign and malignancy group (T=-2.85, P=0.03). The mean ADCmean ±SD in patients with a benign and malignancy is 0.30±0.10 and 0.15±0.04, there is the significant statistical differences between benign and malignancy group (T=-3.82, P=0.02).The sensitivity and specificity and accuracy of PET/CT ,MRI and PET/CT-MRI were 100.00%, 91.67%,100.00%,100.00%, 92.86% ,92.86%, 100.00%,100.00% and 100.00% respectively．The linear correlation analysis result of between SUV and ADC of fourteen patients with head and neck tumor is in the below Table1. There are not significant correlations between SUV and ADC(r<0.5, P>0.5).

Conclusions Trimodality PET/CT-MR can improve diagnosis confidence of head and neck tumor. In the differential diagnosis of benign and malignant both SUV and ADC are very important.

Research Support no