Reliability of respiratory-gated real-time two-dimensional cine incorporating deep learning reconstruction for the assessment of ventricular function in an adult population; a common mistake.
Journal:
The international journal of cardiovascular imaging
PMID:
36826615
Abstract
Reliability (repeatability or agreement) is assessed by different statistical tests including Pearson r or Spearman rho which is one of the common mistakes in reliability analysis. Pearson r or Spearman rho correlation coefficient only assesses the linearity between two continuous variables. Any shift in the regression line's location or scale, which leads to non-reliability, cannot be detected by these correlation coefficients. In reliability analysis, the individual approach should be considered instead of the global average. To correctly assess the reliability (agreement), for continuous variables, either the Bland -Altman plot or Intra-class correlation coefficient (ICC) absolute measure should be applied.