Objective To compare contrast-enhanced magnetic resonance imaging （ceMRI） with nuclear metabolic imaging for the assessment of myocardial viability in patients with chronic ischemic heart disease. Methods Twenty patients with suspected chronic ischemic heart disease underwent ceMRI and technetium-99m sestamibi single-photon emission computed tomography （SPECT）. Patients with positive SPECT results also underwent

18 F-fluorodeoxyglucose （FDG） SPECT. In a 17-segment model

the segmental extent of hyperenhancement （SEH） by ceMRI was compared with segmental FDG and sestamibi uptake by SPECT. Correlation between the extent of hyperenhancement by ceMRI and left ventricular function was analyzed. Results Seven patients got negative results both in ceMRI and technetium-99m sestamibi SPECT. The rest 13 patients with positive results then underwent

18 F-FDG SPECT. In 221 segments of 13 patients

SEH was （2.1±8.2）%

（25.0±13.7）%

and （57.7±23.6）% in segments with normal metabolism/perfusion

metabolism/perfusion mismatch

and matched defects

respectively

and there were significant differences between either two of them （all P

<

0.05）. By receiver operating characteristic curve analysis

the area under the curve was 0.95 for the differentiation between viable and non-viable segments. At the cutoff value of 34%

SEH optimally differentiated viable from non-viable segments defined by SPECT. Using this threshold

the sensitivity and specificity of ceMRI to detect non-viable myocardium as defined by SPECT were 92% and 93%

respectively. Hyperenhancement size by ceMRI was correlated negatively with the left ventricular ejection fraction （r=-0.90

P

<

0.01） and positively with left ventricular volumes （r=0.62 for end-diastolic volume

r=0.75 for end-systolic volume

both P

<

0.05）. Conclusion CeMRI allows assessment of myocardial viability with a high accuracy in patients with chronic ischemic heart disease.