Effect of power Doppler and digital subtraction techniques on the comparison of myocardial contrast echocardiography with SPECT

Abstract

Objective: To compare the accuracy and feasibility of harmonic power Doppler and digitally subtracted colour coded grey scale imaging for the assessment of perfusion defect severity by single photon emission computed tomography (SPECT) in an unselected group of patients.

Design: Cohort study.

Setting: Regional cardiothoracic unit.

Patients: 49 patients (mean (SD) age 61 (11) years; 27 women, 22 men) with known or suspected coronary artery disease were studied with simultaneous myocardial contrast echo (MCE) and SPECT after standard dipyridamole stress.

Main outcome measures: Regional myocardial perfusion by SPECT, performed with (99m)Tc tetrafosmin, scored qualitatively and also quantitated as per cent maximum activity.

Results: Normal perfusion was identified by SPECT in 225 of 270 segments (83%). Contrast echo images were interpretable in 92% of patients. The proportion of normal MCE by grey scale, subtracted, and power Doppler techniques were respectively 76%, 74%, and 88% (p < 0.05) at > 80% of maximum counts, compared with 65%, 69%, and 61% at < 60% of maximum counts. For each technique, specificity was lowest in the lateral wall, although power Doppler was the least affected. Grey scale and subtraction techniques were least accurate in the septal wall, but power Doppler showed particular problems in the apex. On a per patient analysis, the sensitivity was 67%, 75%, and 83% for detection of coronary artery disease using grey scale, colour coded, and power Doppler, respectively, with a significant difference between power Doppler and grey scale only (p < 0.05). Specificity was also the highest for power Doppler, at 55%, but not significantly different from subtracted colour coded images.

Conclusions: Myocardial contrast echo using harmonic power Doppler has greater accuracy than with grey scale imaging and digital subtraction. However, power Doppler appears to be less sensitive for mild perfusion defects.

Figure 1

Concordant diagnosis with multiple image processing strategies. The lateral perfusion defect by single photon emission computed tomography (SPECT) is matched by lateral defects on grey scale, colour coded, and power Doppler images (arrows).

Figure 2

False positive contrast perfusion defect caused by a "septal stripe" (arrow). This appearance is caused by inhomogeneity of the resting image and reflects the disadvantage of simple subtraction.

Figure 3

False positive perfusion defect because of attenuation. An apical perfusion defect is detected using each technique. However, power Doppler shows another apparent defect in the basal septum (arrow), not matched on SPECT or colour coded imaging. This false positive is attributed to attenuation.

Figure 4

Apical defect shown by both single photon emission computed tomography (SPECT) and contrast echo (arrow), with unmatched lateral wall defect by contrast echocardiography (large arrow). This artefact is caused by overlying lung.

Figure 5

Apparently normal image with digital subtraction and colour coding in a patient with a mild perfusion defect in the basal septum on the SPECT and power Doppler images (arrows). This false negative contrast pattern with digital subtraction and colour coding reflects difficulties in alignment and superimposition of the images.

Figure 6

Discordance between single photon emission computed tomography (SPECT) with all echo techniques. The large septal perfusion defect by SPECT (arrow) is probably the result of left bundle branch block in a patient without a history of coronary artery disease.

Figure 7

Proportion of abnormal segments by myocardial contrast echo (defined by any defect), compared with quantitative SPECT activity (as per cent of maximum).

Figure 8

Sensitivity, specificity, and accuracy of grey scale, colour coded, and power Doppler techniques for contrast echocardiography.