THE INFLUENCE OF MR FIELD-STRENGTH ON THE DETECTION OF FOCAL LIVER-LESIONS WITH SUPERPARAMAGNETIC IRON-OXIDE

The purpose of this study was to compare the value of low- vs high-fie ld MR systems in the detection of focal liver lesions after IV adminis tration of iron oxide particles. A prospective study was undertaken wh ich included 20 patients with focal liver lesions on CT or US, or stro ng clinical suspicion of focal liver disease. Iron oxide particles wer e administered in an IV drip infusion over 30 min. Magnetic resonance imaging was subsequently performed on a 0.2 and a 1.5-T system. Both e xaminations were performed in one session. Turbo spin-echo T2-weighted sequences were used for further analysis (at 0.2 T: TR 4050 ms, TE 96 ms, 1.5 T: TR 3000 ms, TE 103 ms). After randomisation, images were a nalysed by two blinded readers. The evaluation included lesion counts, determination of lesion conspicuity and overall image quality (both g raded on a scale 1-5). Quantitative analysis was performed on 29 lesio ns. Lesion-to-liver signal intensity and contrast-to-noise ratios (CNR s) were calculated. The total lesion count (cumulative counts for two observers) was 59 on the high-field system and 63 on the low-field sys tem. Statistical analysis showed no significant difference. On both sy stems median value for lesion conspicuity was 3. No statistically sign ificant difference was found. Global image quality was rated higher (p = 0.0017). Quantitative analysis showed no significant difference for lesion-to-liver signal intensity ratios or CNRs. Although subjective image quality is significantly better on the high-field system, this d oes not result in better lesion detection or better lesion conspicuity . No significant difference in objective quantitative parameters was f ound in our series.