Nuclear ejection fraction measurements with a small crystal detector: methodological investigations

Summary. A series of studies were performed with a single crystal detector system, nuclear stethoscope (N.S.) (Bios®) used for measuring left ventricular function (ejection fraction [EF]). Counting capacity, validity, reproducibility, effect of repositioning of the instrument and of different operators were investigated. Counting capacity was found to be approximately linear up to a count rate of 40,000 cps. Validity was studied by comparing measurements in the same subject using the N.S. with those obtained by a gamma-camera computer system with software using a background subtraction similar to that in the N.S. The mean difference in EF (0·01) was not significant. Maximum emptying and filling velocity measurements with the two instruments were also compared r= 0·86 and r= 0·89, respectively, and a coefficient of variation of 18·1% and 17·9%. Reproducibility was determined from duplicate determinations under identical conditions. The standard error of a single determination was 5%. Repositioning by the same operator increased the variation to 9·1% without any systematic difference. Comparison between two operators did not increase the coefficient of variation (8·8%). There was a systematic difference between measurements using the two acquisition modes, ventricular function mode and position monitor mode. Beat-to-beat variation of EF values during 12 s in 25 patients with coronary heart disease studied under standardized resting conditions was 11·9%.