Early diagnosis and treatment of non-alcoholic fatty pancreatic disease (NAFPD) can effectively intervene in the development of type 2 diabetes. This study aimed to evaluate the utility of an improved ultrasound method for diagnosing NAFPD.

All patients underwent abdominal ultrasonography (US) and magnetic resonance imaging (MRI) mDixon-Quant technique. Patients with a pancreatic fat fraction (PFF) > 6.2% were in the NAFPD group (NA) and the rest were in the normal group (NC). MRI mDixon-Quant technique was used to evaluate the diagnostic efficiency of NAFPD with improved ultrasound diagnosis.

This study included 46 participants. The MRI mDixon-Quant had good repeatability and reproducibility in measuring PFF. The kappa value of the improved version ultrasound (IVUS) method and MRI diagnosis was 0.760 (95% confidence interval [CI] = 0.662-0.858, P < 0.001). That of traditional version ultrasound (TVUS) and MRI diagnosis was 0.497 (P < 0.001). Statistical analyses revealed that pancreatic grading based on both ultrasound evaluation methods correlated with PFF, with IVUS (R² = 0.812) superior to TVUS (R² = 0.496). Body weight, body mass index, triglyceride level, abdominal circumference, abdominal visceral fat, total abdominal fat, abdominal visceral fat area, and liver fat content were significantly higher in the NA versus NC group (P < 0.05). Fat fractions of the pancreatic head, body, and tail in the NA group were significantly higher than those in the NC group (P < 0.05).

IVUS more consistently predicts NAFPD and correlates better with MRI than TVUS.