Reproducibility and variability of global and regional dyssynchrony parameters derived from phase analysis of gated myocardial perfusion SPECT

Aim. The aim of the study was to assess the reproducibility (R) and variability of quantitative parameters derived from Cedars-Sinai QGS software for phase analysis of G-SPECT in 39 coronary artery disease (CAD) patients with or without severe heart-failure (HF). Methods. Phase standard deviation (s) and entropy (e) are global parameters that quantify dyssynchrony. Regional measures of dyssynchrony were also computed including differences between the lateral wall versus the septal wall (Dm _w) and differences between the averages of midventricular lateral versus midventricular septal segments (Dm ₅). Results. Global parameters e and s and regional parameters Dm _w and Dm _s exhibit excellent values of R=0.92, 0.99, 0.99 and 0.96, respectively. In regional parameters Dm _w and Dm _s there is a significant variability in individual scoring assessed by a standard error of measurement of =9.17 and 21.7, respectively. The box plots of e in patients with or without HF do not show any significant superimposition, while the box plots of s and Dm _w show a partial overlap mainly due to the significant variability of s and Dm _w within patients with HF. Conversely the box plots of Dm _s in patients with or without HF show a significant overlap due to the great variability of Dm _s within patients with HF. Conclusion. Regional parameters derived from phase analysis of G-SPECT studies are not useful in the individual assessment of dyssynchrony in CAD patients either due to large variability in individual scoring and to a large heterogeneity in HF patients. Global parameters (e and s) exhibit both an excellent reproducibility. Nonetheless, e seems to perform better than s in individual assessment of dyssynchrony due to a better separation between HF and non HF patients. Using the QGS software approach no manual intervention is necessary to ensure a good reproducibility of global parameters.