Reproducibility of intravascular ultrasound radiofrequency data analysis: implications for the design of longitudinal studies