Abstract:
Objective To analyze the correlation and consistency of MRI IDEAL-IQ and quantitative CT (QCT) in the quantitation of pancreatic fat content in patients with type 2 diabetes.
Methods Sixty-seven volunteers with type 2 diabetes (33 men and 34 women, mean age years) underwent both MRI IDEAL-IQ and QCT examinations. The regions of interest (ROI) were placed at the head, body, and tail of the pancreas to obtain MRI pancreas fat fraction (MR-PFF) and CT-PFF. The correlation between MR-PFF and CT-PFF was analyzed, and the consistency of them was analyzed by Bland-Altman. The correlations between MR-PFF and liver fat content, abdominal circumference, and body mass index (BMI) were also analyzed.
Results In sixty-seven volunteers, the mean abdominal circumference was (83.34±10.1) cm, BMI was (24.02±2.96) kg/m2, and liver fat content was (5.28±2.76)%. The abdominal circumference and liver fat content were correlated with MR-PFF (P < 0.05). MR-PFF and CT-PFF were (12.21±10.71)% and (19.9±10.61)%, respectively. CT-PFF showed a significant linear correlation with MR-PFF (r=0.842, P < 0.001). MR-PFF=-4.169+0.818×CT-PFF. The Bland-Altman analysis results showed that the 95% CI of consistency was -4.15%-19.75% (P=0.044), and there were 3 values outside of the interval.
Conclusions The pancreatic fat contents measured by QCT and MRI have good correlation and consistency, and QCT can be used to measure pancreatic fat content.