Objective To investigate whether timing of image acquisition influenced infarct size estimation using delayed CeMRI,and the association of left ventricular ejection fraction between magnetic resol3anee imaging and left ventrieulography Was also studied.Method From Junary 2005 to April 2006,27 first,onset AMI patients [23 male,mean age(54.3±10.5)years]were enrolledinthistudr.Allpatients receivedleft ventrictdographyas well as coronary angiography.The average checking time was(13.2±5.2)clays after the onset of AMI.MR imaging was performed with a 1.5-T magnet(SIMENS).After breath-hold eine images were acquired,patients re.ceived afI intravenous bolus of 0.05 mmol/kg Gd-DTPA at a rate of 5 ml/8.A first-pass perfusion scan was ac.qllired.Then a second bolus of 0.15 mmoVkg Gd-DTPA was give.at a rate of 2 mE/Is.After the hyperenhancement localized,the typical short axis slice with hyperenhancement WaS chosen to repeat imaging for IlleasuriIin.farct size every5minutesfrom5minutes after secondinjection ofcontrast until 20minutes.Results Twexty-seren patients showed hyperenhancement at the delayed CeMRI and hypoenhancement at the first pass enhancement(FPE).The average infarct size estimated by CeMRI WaS(17.9士9.8)%of LV nlass.Myocardial enhancement at a repesentative short-axis slice WIllS(7.2±6.2)%of LV Imss at 5 minutes,(8.5±7.4)%at 10 minutes,(7.3±6.3)%at 15 minutes and(6.9-t-6.4)%at 20 minutes respectively.There WltlS significant difference be-tween lmfninmes and 20-minutes enhancement size(P<0.05).Correlations of EF obtained by cineventriculo-grapIIy and MR irr,lg were significant(r=0.867,P<0.01).There were also correlations between infarction size and pe.k CK(r:O.819,P<O.01),a8 well ills peaI[cTNT(r=0.517,P<0.05)levels.Tuning of image acquisition iufluenced infarct size quantification using delayed CeMRI when TI Was kept constant.