Diffusion kurtosis imaging features of renal cell carcinoma: a preliminary study.

Abstract

To explore the feasibility of diffusion kurtosis imaging (DKI) in differentiating different types of renal cell carcinoma (RCC). 36 patients with clear cell RCC (CCRCC, low-grade,n = 20 and high-grade, n = 16), 19 with papillary RCC, 11 with chromophobe RCC, and 9 with collecting duct carcinoma (CDC) were examined with DKI technique. b values of 0, 500 and 1000 s/mm2 were adopted. The DKI parameters, i.e., mean diffusivity (MD), mean kurtosis (MK), kurtosis anisotropy (KA), radial kurtosis (RK) and signa-to-noise ration (SNR) of DKI images at different b values were used. The mean SNRs of DKI images at b = 0, 500 and 1000 s/mm2 were 32.8, 14.2 and 9.18, respectively. For MD parameter, a significant higher value was shown in CCRCC than those of papillary RCC, chromophobe RCC and CDC (p < 0.05). In addition, both chromophobe RCC and CDC have larger MD values than papillary RCC (p < 0.05), however, there was no significant differences between chromophobe RCC and CDC (p > 0.05). For MK, KA and RK parameters, a significant higher value was shown in papillary RCC than those of CCRCC, chromophobe RCC and CDC (p < 0.05). Moreover, both chromophobe RCC and CDC have significantly larger values of MK, KA and RK than CCRCC (p < 0.05). Our preliminary study demonstrated significant differences in the DKI parameters between the subtypes of RCCs, given an adequate SNR of DKI images. 1.The MD value is the best parameter to distinguish CCRCC from other RCCs.2.The MK, KA and RK values are the best parameters to distinguish papillary RCC from other RCCs.3.DKI is able to provide images with sufficient SNRs in kidney disease.