Shun Yu1, Zhongshuai Zhang2, Qiang Bao1, Jiawei Su1, Mengxiao Liu2, Qinglei Shi2, Wenchao Cai1. 1. 1 Department of Radiology, Provincial Clinical College of Fujian Medical University, Fujian Provincial Hospital , Fuzhou, Fujian , China. 2. 2 Diagnosis Imaging, Siemens Healthcare Ltd. , Shanghai , China.
Abstract
OBJECTIVES: To study the value of diffusion kurtosis imaging (DKI) in diagnosis of parotid gland disease (PGD) with different pathological patterns and parotid adenolymphoma (PAL). METHODS: 57 patients with different kinds of PGD were enrolled and underwent DKI and conventional diffusion-weighted imaging (DWI). All patients were categorized into different groups according to their pathological patterns. The result of calculating the value of DKI-derived parameters (Kmean, Krad, Kax, Dmean, Drad, Dax, and FA) and apparent diffusion coefficient (ADC) of DWI were used to study their diagnostic accuracy in PGD with different pathological patterns. The binary logistic regression method was used to confirm clinical valuable diffusion parameters (obtained with DKI and DWI models) for diagnosing PAL. Using MedCalc 13.0, receiver operating characteristic (ROC) analysis was conducted to evaluate the diagnostic value of confirmed parameters based on the logistic regression equation. RESULTS: Both DKI parameters and conventional ADC showed statistical significance in diagnosing PGD with different pathological patterns (p < .01). By using the DKI model, kurtosis coefficients showed higher diagnostic capability than diffusion coefficients ([Kmean+Krad + Kax] vs [Dmean +Drad + Dax]: 22 vs 15, p < .01) did in the differential diagnosis among different PGD groups. In the diagnosis of PAL among all PGD patterns, the ROC analysis demonstrated that the area under curve (AUC) FA +Kax [0.881 ± 0.057 (0.824 to 0.938)] is higher than that when using FA [0.629 ± 0.095 (0.534 to 0.724)] and Kax [0.800 ± 0.070 (0.730 to 0.870)] alone (p < .05), with sensitivity, specificity, accuracy, and both positive and negative predictive values of 71.43, 95.78, 91.77, 76.92, and 94.44%, respectively. CONCLUSIONS: DKI showed higher diagnostic capacity in the differential diagnosis of PGD with different pathological patterns. Combined parameters of DKI can differentiate PAL from other PGD pathological patterns with a high degree of accuracy. This technique shows great potential for DKI in the differential diagnosis of PGD within a certain pathological category.
OBJECTIVES: To study the value of diffusion kurtosis imaging (DKI) in diagnosis of parotid gland disease (PGD) with different pathological patterns and parotid adenolymphoma (PAL). METHODS: 57 patients with different kinds of PGD were enrolled and underwent DKI and conventional diffusion-weighted imaging (DWI). All patients were categorized into different groups according to their pathological patterns. The result of calculating the value of DKI-derived parameters (Kmean, Krad, Kax, Dmean, Drad, Dax, and FA) and apparent diffusion coefficient (ADC) of DWI were used to study their diagnostic accuracy in PGD with different pathological patterns. The binary logistic regression method was used to confirm clinical valuable diffusion parameters (obtained with DKI and DWI models) for diagnosing PAL. Using MedCalc 13.0, receiver operating characteristic (ROC) analysis was conducted to evaluate the diagnostic value of confirmed parameters based on the logistic regression equation. RESULTS: Both DKI parameters and conventional ADC showed statistical significance in diagnosing PGD with different pathological patterns (p < .01). By using the DKI model, kurtosis coefficients showed higher diagnostic capability than diffusion coefficients ([Kmean+Krad + Kax] vs [Dmean +Drad + Dax]: 22 vs 15, p < .01) did in the differential diagnosis among different PGD groups. In the diagnosis of PAL among all PGD patterns, the ROC analysis demonstrated that the area under curve (AUC) FA +Kax [0.881 ± 0.057 (0.824 to 0.938)] is higher than that when using FA [0.629 ± 0.095 (0.534 to 0.724)] and Kax [0.800 ± 0.070 (0.730 to 0.870)] alone (p < .05), with sensitivity, specificity, accuracy, and both positive and negative predictive values of 71.43, 95.78, 91.77, 76.92, and 94.44%, respectively. CONCLUSIONS: DKI showed higher diagnostic capacity in the differential diagnosis of PGD with different pathological patterns. Combined parameters of DKI can differentiate PAL from other PGD pathological patterns with a high degree of accuracy. This technique shows great potential for DKI in the differential diagnosis of PGD within a certain pathological category.
Authors: Andrew B Rosenkrantz; Anwar R Padhani; Thomas L Chenevert; Dow-Mu Koh; Frederik De Keyzer; Bachir Taouli; Denis Le Bihan Journal: J Magn Reson Imaging Date: 2015-06-26 Impact factor: 4.813
Authors: Nathaniel D Kelm; Kathryn L West; Robert P Carson; Daniel F Gochberg; Kevin C Ess; Mark D Does Journal: Neuroimage Date: 2015-09-21 Impact factor: 6.556