[Echocontrast ventriculography. Determination of left ventricular functional parameters with special reference to...

In 57 consecutive patients we examined whether digital subtraction echocardiography (DSE) is able to determine left ventricular function parameters, in particular the ejection fraction (LV-EF). The most important step in DSE consists of the subtraction of digitized grey level information of the original echocardiogram from a phase-identical contrast image. In addition LV-EF, end-diastolic (EDV) and end-systolic volumes (ESV) were calculated by manual definition of endocardium in the original (Na) and contrast-echocardiograms (Ko). The results were then compared with the invasive cardiac catheterisation (Hk) data. We used 5.5% oxypolygelatine (Gelifundol S) as echo-contrast material and injected 2-4 ml as a bolus directly into the left ventricle. LV-EF calculated on the basis of automatic contour definition by DSE demonstrated a better correlation with cineangiographic data (r = 0.89, SEE 5.8% in 44 patients) than the manual input of endocardial borders in original echocardiograms (r = 0.71, SEE 8.1% in 47 patients). Good correlations were also found between EF-Ko and EF-Hk with r = 0.92 and an SEE of 4.3%; because there is a good agreement between EF-Ko and EF-DSE (r = 0.91, SEE 5.3%) these calculations closely approach the results of the invasive data. The EF-Na underestimates the EF-Hk markedly as a consequence of a systematic error: as compared to the contrast image the volumes of the original echocardiograms, in particular the ESV are overestimated and the ejection fraction as a derived parameter is calculated as too small. ESV-Hk versus ESV-Na: y = 0.82x + 24.4, ESV-Hk versus ESV-DSE: y = 0.88x + 37.0.(ABSTRACT TRUNCATED AT 250 WORDS)