Dixon sequences play a major role in measurement of fat fraction (FF, PDFF) in MRI. Methods for quantifying FF are based on Dixon, which are widely available on modern tomographs regardless of the manufacturer. Existing software solutions for automatic calculation of fat fraction, which have proven their high accuracy in quantifying FF (for example, “IDEAL IQ” for GE, “Liver Lab” for Siemens, etc.). Software techniques are not available on all tomographs, as they are an additional paid option. Nevertheless, in the arsenal of a radiologist there are several approaches to calculating the fat fraction using standard Dickson sequences available on any MRI scanner, namely: according to formulas based on signaling characteristics in phase (In-phase) and antiphase (Out-phase): FF1= (In-Out)/2*In*100%; or by images weighted by water (Water) and fat (Fat): FF2= Fat/(Fat+water)*100% [1,2].

The aim of the work was to compare the quantitative assessment of the fat fraction (FF) in CT and MRI using generally accepted formulas in comparison with the specified phantom indicators.