Automated multidetector computed tomography evaluation of subacutely infarcted myocardium

Abstract

Background: Delayed enhanced (DE) multidetector computed tomography (MDCT) can identify acute and chronic myocardial infarct. To our knowledge, automated techniques for infarct quantification on DE-MDCT have not been used.

Objective: We evaluated an automated signal intensity (SI) threshold method for quantification of subacute myocardial infarct and identified and quantified microvascular obstruction (MO) in subacute infarct.

Methods: DE-MDCT imaging was performed on 5 pigs 6-7 days after mid left anterior descending artery occlusion-reperfusion. DE-MDCT images were compared with triphenyl tetrazolium chloride (TTC) staining for infarct quantification and with hematoxylin and eosin (H&E) staining for MO quantification. Pixels with SI more than the mean SI of a remote normal myocardial region (SI(remote)) plus 2 times the standard deviation (SI(remote) + 2 SD) value were considered infarct pixels. The ratio of infarct to total area of a given slice, the percentage of infarct area per slice (PIS), was calculated. MO as a percentage of total infarct area was also calculated.

Results: The average density values on DE-MDCT (5 minutes after contrast injection) were remote normal myocardium of 93 +/- 19 Hounsfield units (HU), infarct myocardium of 159 +/- 40 HU, blood of 140 +/- 26 HU, and MO of 85 +/- 30 HU. PIS(MDCT) showed substantial agreement with PIS(TTC) (y = 1.003x + 4.12; R = 0.90, P < 0.05). A relation was also shown between MO determined by MDCT compared with H&E staining (y = 0.74x + 3.4).

Conclusions: We show the feasibility of using a semiautomated SI threshold technique for quantification of subacute myocardial infarct. We also show the persistent MO in subacute myocardial infarct on DE-MDCT images.