Effects of regions of interest methods on apparent coefficient measurement of the parotid gland in early Sjogren's syndrome at 3T MRI.

BACKGROUND: The apparent diffusion coefficient (ADC) has been used to assess parotid gland abnormalities in Sjogren's syndrome (SS) patients; however, few data exist on the influence of region of interest (ROI) methods on ADC measurements. PURPOSE: To assess the influence of ROI methods on ADC measurement, and their diagnostic ability in detecting parotid gland abnormalities in early SS patients. MATERIAL AND METHODS: Thirteen early SS patients underwent parotid gland diffusion-weighted imaging scans at a 3.0 T MR unit. Two readers independently measured the parotid gland ADC value using three different ROIs (whole-gland [WG], single-slice [SS], and reader-based circular [RBC]). The ADC value based on three different ROIs (ADC-ROI(WG), ADC-ROI(SS), ADC-ROI(RBC)) were compared between the SS group and a matched healthy control (HC) group (n = 19). Receiver operating characteristic (ROC) curves and intra-class correlation coefficients (ICC) were used to determine the diagnostic ability and reproducibility of the parameters. RESULTS: The ADC-ROI(WG), ADC-ROI(SS), and ADC-ROI(RBC) in the SS group were all significantly higher than those in HC group (all P < 0.05). The ADC-ROI(WG) showed better diagnostic ability than did ADC-ROI(RBC) (P = 0.0200), while no significant difference was found between ADC-ROI(WG) and ADC-ROI(SS) (P = 0.4636). The ROI(WG) method showed the best inter- and intra-reader agreement (ICC, 0.902 and 0.928, respectively), followed by ROI(SS) and ROI(RBC). CONCLUSION: The ROI methods can influence the parotid gland ADC measurements and their diagnostic ability. Considering our results, we suggest using in clinical practice single-slice ROIs to measure the ADC of the parotid gland.