A method to measure renal inner medullary perfusion using MR renography

Area of Science:

• Nephrology and Renal Physiology
• Medical Imaging and Modeling
• Pharmacokinetics

Background:

• The renal medulla is highly vulnerable to perfusion and oxygenation deficits.
• Separate regulation of cortical and medullary perfusion necessitates distinct measurement methods.
• Understanding regional renal perfusion is key to advancing renal pathophysiology insights.

Purpose of the Study:

• To develop and validate a physiologically accurate model for measuring inner medullary perfusion (Fmed) and cortical perfusion (Fcor) independently.
• To assess the stability and accuracy of the proposed 7-compartment renal perfusion model.
• To compare the novel model's performance against a conventional 2-compartment filtration model.

Main Methods:

• A 7-compartment model of renal perfusion was developed.
• An iterative approach was employed to fit 10 free parameters.
• Model stability and accuracy were evaluated using patient data and simulations; cortical perfusion (Fcor) and tubular flow (FT) were compared to a 2-compartment model.

Main Results:

• Average inner medullary perfusion (Fmed) was determined to be 37(23)mL/100 mL/min.
• Fitting stability showed a median coefficient of variation of 0.0% (0.0-5.8%), with outliers up to 81%.
• Simulations indicated an average underestimation of Fmed by approximately 8%; intra-class correlation coefficients for Fcor and FT were 0.87 and 0.63, respectively, when compared to the 2-compartment model.

Conclusions:

• A pharmacokinetic model closely aligned with renal physiology was successfully developed.
• While the model demonstrated vulnerability to overfitting, relatively stable results were achieved, particularly for Fmed.
• The model offers a promising approach for distinct measurement of renal medullary and cortical perfusion.