MR-based semi-automated quantification of renal functional parameters with a two-compartment model--an interobserver...

Area of Science:

• Medical Imaging
• Renal Physiology
• Quantitative MRI

Background:

• Accurate quantification of renal perfusion and filtration is crucial for diagnosing and managing kidney diseases.
• Semi-automated approaches using mathematical models offer potential for standardized analysis of MRI data.
• Assessing interobserver agreement is vital for validating the reliability of quantitative MRI methods.

Purpose of the Study:

• To evaluate the consistency between two independent observers in quantifying renal perfusion and filtration parameters.
• To assess the interobserver agreement of a semi-automated, two-compartment model analysis for MR-based renal function parameters.
• To identify parameters with high and low interobserver variability for potential clinical application.

Main Methods:

• Twelve patients underwent dynamic contrast-enhanced MRI (3.0 T) with Gd-BOPTA.
• Two observers manually defined regions of interest (ROIs) for arterial input function and tissue-air interface.
• A pixel-by-pixel two-compartment model analysis generated maps of plasma flow (FP), plasma mean transit time (TP), tubular flow (FT), and tubular mean transit time (TT).

Main Results:

• High correlation coefficients were observed between observers for all parameters (0.78–0.90, p<0.05).
• Significant systematic differences were found for plasma flow (FP) between observers (p=0.004).
• Plasma mean transit time (TP) demonstrated the highest stability between observers.

Conclusions:

• The semi-automated two-compartment model shows good interobserver agreement for most renal perfusion and filtration parameters.
• A significant systematic difference in plasma flow (FP) necessitates further refinement for clinical routine use.
• Reducing residual variability is essential to enhance the robustness and clinical applicability of this quantitative MRI method.