Influence of operator’s professional experience and visual function in the myocardial perfusion scan semiautomatic processing: intra and inter operator variability in quantitative parameters estimation

Abstract

Introduction – Myocardial perfusion scan (MPS) has an important role in the diagnosis, evaluation, and follow-up of patients with coronary artery disease being the semiautomatic processing widely used for routine clinical practice. Since nuclear medicine technologists (NMT) performance might be affected by individual and environmental factors, different professionals that process the same data will provide different estimations of quantitative parameters (QP). Aim – Evaluate the influence of NMT professional experience and visual function on the MPS semiautomatic processing. Analysis of the intra and inter variability considering the function and perfusion QP measured. Methodology – Twenty subjects were selected and categorized into two groups according to their experience with the Quantitative Gated SPECT^TM software: Group A (GA) – NMT ≥600h and Group B (GB) – NMT without experience. All NMTs underwent an orthoptic evaluation and processed 21 MPS five times in a non-consecutive mode. It was assumed to be an altered vision when at least one visual function parameter was found abnormal. The coefficient of variation, %, was calculated to evaluate repeatability and reproducibility. Friedman’s and Wilcoxon's tests were applied to compare QP between operators and to analyze GA and GB performance, respectively, considering the processing of the same MPS. To compare the NMT with the normal and altered vision it was used the Mann-Whitney Test, and to evaluate the visual function influence in each QP was performed the ETA association coefficient. It was assumed statistically significant differences at a significance level of 5%. Results and Discussion – Low intra (<6.59%) and inter (<5.07%) operator variability was achieved. A greater discrepancy in GB was noted, being merely the septal wall (SW) statistically different (z_w=-2.051, p=0.040) when compared with GA. Regarding the influence of visual function, it was observed significant statistical differences only in left ventricle ejection fraction (LVEF) (U=11.5, p=0.012) between NMT with normal and altered vision, contributing to the vision 33.99% for its variation. It denoted more differences between NMT that have a greater incidence of ocular symptomology and diminished Binocular Vision. LVEF was the most consistent QP (1.86%) between operators. Conclusion – MPS is presented as a repeatable and reproducible technique, operator-independent. It was verified professional experience and visual function influence, MPS semiautomatic processing, in the SW and LVEF QP, respectively.