Contribution of mono-exponential, bi-exponential and stretched exponential model-based diffusion-weighted MR imaging in the diagnosis and differentiation of uterine cervical carcinoma.

Abstract

To investigate the potential of various metrics derived from mono-exponential model (MEM), bi-exponential model (BEM) and stretched exponential model (SEM)-based diffusion-weighted imaging (DWI) in diagnosing and differentiating the pathological subtypes and grades of uterine cervical carcinoma. 71 newly diagnosed patients with cervical carcinoma (50 cases of squamous cell carcinoma [SCC] and 21 cases of adenocarcinoma [AC]) and 32 healthy volunteers received DWI with multiple b values. The apparent diffusion coefficient (ADC), pure molecular diffusion (D), pseudo-diffusion coefficient (D*), perfusion fraction (f), water molecular diffusion heterogeneity index (alpha), and distributed diffusion coefficient (DDC) were calculated and compared between tumour and normal cervix, among different pathological subtypes and grades. All of the parameters were significantly lower in cervical carcinoma than normal cervical stroma except alpha. SCC showed lower ADC, D, f and DDC values and higher D* value than AC; D and DDC values of SCC and ADC and D values of AC were lower in the poorly differentiated group than those in the well-moderately differentiated group. Compared with MEM, diffusion parameters from BEM and SEM may offer additional information in cervical carcinoma diagnosis, predicting pathological tumour subtypes and grades, while f and D showed promising significance. • DWI-derived parameters by different models are related but provide diversified information. • Commonly used ADC by MEM of DWI overestimates the tissue water diffusivity. • DWI processed by BEM could separate blood perfusion from true diffusion effects. • The derived diffusion-related and perfusion-related parameters by BEM are superior to ADC.