Influence of background and absorption correction on nuclear quantification of left ventricular end-diastolic volume.

Left ventricular (LV) end-diastolic volume (EDV) was determined in millilitres by equilibrium radionuclide ventriculography (MUGA) in 38 patients and in 15 normal individuals by applying corrections for background (BC), LV blood self-absorption and for absorption by the thoracic and LV walls. Volume calibration was performed by measuring a syringe containing the patient's venous blood with the gamma camera. Single plane cineventriculography (CVG) served as the reference method. Without absorption correction, LV volume values were underestimated by a factor of 3.6 on the average. Several background models (uniform, parabolic, no background) were investigated. Without BC, EDV was overestimated, and with uniform BC, EDV was underestimated. EDV calculated using a parabolic BC with a correction for absorption yielded the best correlation (r = 0.96) with volumes (VOL) obtained by CVG (VOLMUGA [ml] = -6.83 + 1.060 X VOLCVG).