Erythrocytes were labeled with 30 mCi of technetium-99m (““Tc) by using the in vivo“ or modified in vivo procedure. Acquisition was gated to the R-wave of the ECG and collected at 16 frames per cardiac cycle. Anterior, left lateral, and left anterior oblique (LAO) radionuclide angiograms were performed at rest; the LAO view which allowed maximum separation of left and right ventricles was used. Besting and exercise LAO views were performed with a parallel hole collimator with a caudal tilt of 15° and repeat LAO views were obtained during the last two minutes of each three-minute exercise stage. Here
Radionuclide data were processed by using a commercially available computer system and software (Medical Data Systems) and previously reported techniques. A second derivative technique was employed to identify the left ventricular region of interest in each frame. A background was chosen 5 pixels lateral to the left ventricular region of interest. Ejection fraction (EF) was calculated at rest and at maximal exercise from a background-corrected left ventricle counts vs time curve using a variable region of interest. Wall motion was accessed at rest in the anterior, left lateral and LAO positions. Wall motion at rest and maximal exercise was accessed subjectively in the LAO projection as the consensus of two experienced observers (one staff cardiologist and one staff nuclear medicine physician). An exercise regional wall motion abnormality was considered present if wall motion in any segment worsened significantly from rest to exercise.
End-diastolic and end-systolic volumes of the left ventricle were determined using a count-based method. A blood sample (10 ml) was obtained immediately after exercise for cardiac volume determination. The count activity in the blood pool was accessed directly on the gamma camera for two minutes. In order to determine left ventricular volume, left ventricular count activities were corrected for blood pool activity and time. A regression equation has been developed comparing the volume units determined in this fashion with volumes determined by contrast ventriculography. Correlation coefficients for end-diastolic and end-systolic volume determined in this fashion compared with those determined by contrast ventriculography were previously reported to be 0.85 and 0.94 respectively, with standard errors of 28 and 18 ml, respectively. End-diastolic volume index (EDVI) and end-systolic volume index (ESVI) were determined by dividing the end-diastolic volume and end-systolic volume by the body surface area. These volumes are used to calculate the stroke volume index (SVI) and cardiac index (Cl) for the left ventricle by the following equations:
SVI (total) = EDVI – ESVI Eq 1
Cl (total) = SVI (total) x Heart Rate Eq 2
Regurgitant fraction (RF) was obtained at rest and after maximal exercise. The raw clinical data were filtered using a 9-point temporal and 9-point smoothing algorithm. Standard computer software was then employed to produce phase and amplitude images. The amplitude image was visually inspected, and the left ventricle (LV) and right ventricle (RV) are outlined to best separate the ventricles from the atria. These outlines are visually compared with the first frame of the smoothed raw data, and obvious discrepancies are corrected manually by using a joystick. The fixed LV and RV regions of interest are than applied to the raw data. The maximum and minimum counts are obtained for both the RV and LV and are used to calculate the stroke counts for each ventricle. The following equations were used:
In our laboratory, regurgitant fraction of 25 percent represents the upper limit of normal (approximately two standard deviations above the mean); values between 25-30 percent are considered “borderline,” 31-40 percent are consistent with mild mitral regurgitation, 41-50 percent are consistent with moderate MR, and values exceeding 50 percent are considered severe MR. The interobserver variability for determination of RF using these methods is less than 5 percent when performed by experienced physicians (Hauser M, Gibbons R, unpublished data).