Author: Hyo-Min Cho, Cheolpyo Hong, Changwoo Lee, Sunyoung Lee 👨🔬
Affiliation: Daegu Catholic University, Korea Research Institute of Standards and Science (KRISS), Korea Research Institute of Standards and Science 🌍
Purpose: With the increasing prevalence of metabolic syndrome, the importance of diagnosing obesity and fatty liver diseases has grown significantly. As a result, there is a rising demand for non-invasive methods to assess body fat using medical imaging technology. This study aims to develop reference materials to improve fat quantification accuracy using various MRI systems and enable harmonization of measurement values across different equipment.
Methods: High-intensity focused ultrasonic emulsification was used to prepare surfactant-free micro-emulsions of three fat fractions, (9.12 ± 0.02) %, (18.32 ± 0.04) %, and (27.86 ± 0.05) %. The stability and homogeneity of the RMs were determined using the Karl-Fisher titration method. Each fat content-specific reference material was encapsulated in vials and assembled into a phantom for MRI imaging. PDFF data were acquired using three different MRI systems in three different site. During data acquisition, the unique CSE MRI PDFF protocols of each manufacturer were applied. PDFF data for identical reference materials were collected from each system, and the differences between the measured and reference values were analyzed. A linear regression model was used to derive the correlation between the fat fraction of the reference materials and the measurements from each device.
Results: The PDFF values of the reference materials measured by MRI demonstrated a high correlation in the linear regression model, with Siemens Healthineers systems showing the highest accuracy. Differences in PDFF measurements among manufacturers were identified, and the feasibility of harmonizing measurement values across devices using identical reference materials was demonstrated.
Conclusion:
This study established a foundation for advancing PDFF measurement techniques using MRI by enhancing the accuracy of fat quantification and demonstrating the potential for harmonization across devices. The developed reference materials enable comparison and evaluation of data across multiple institutions and MRI systems, contributing to improved reliability in imaging-based fat quantification diagnostics.