ObjectiveTo investigate the utility of stretched exponential model diffusion-weighted imaging (DWI) for diagnosing of advanced liver fibrosis.MethodsThe patients with chronic liver disease complicated with vary degrees of fibrosis confirmed by pathological examination underwent DWI using different b-values (0, 50, 600 s/mm2) at the First Affiliated Hospital of Chengdu Medical College from June 2015 to February 2020 were collected. In addition, patients who underwent upper abdominal MRI examination in the same hospital at the same time and had no liver disease or disease affecting liver function were collected as a control group. The apparent diffusion coefficient (ADC) was calculated by using a mono-exponential model. The distributed diffusion coefficient (DDC) and water molecular diffusion heterogeneity index (α) were calculated by using a stretched exponential model. The fibrosis stage was evaluated by using the Metavir scoring system. The ADC, DDC, and α among different fibrosis groups were compared. The receiver operating characteristic (ROC) curve was used to analyze the diagnostic efficacy of these three quantitative parameters for advanced liver fibrosis.ResultsA total of 42 patients with chronic liver disease were collected in this study, including mild liver fibrosis (S1–S2, n=16) and advanced liver fibrosi (≥S3, n=24); 15 patients in the control group. The values of ADC, DDC, and α of the patients with mild liver fibrosis and advanced liver fibrosis were significantly lower than those of the control patients (P<0.05). The area under the ROC curve of ADC, DCC, and α in diagnosing liver fibrosis (≥S1) was 0.915, 0.974, and 0.835, respectively, which in diagnosing advanced liver fibrosis (≥S3) was 0.744, 0.869, and 0.758, respectively. However, further the area under ROC curve among these three metrics had no statistical differences (P>0.05).ConclusionDDC based on stretched exponential model is valuable for diagnosis of advanced liver fibrosis.