Two-compartment modeling of tissue microcirculation revisited.

Area of Science:

• Biomedical Engineering
• Medical Imaging Analysis
• Pharmacokinetics

Background:

• Conventional two-compartment modeling in DCE imaging assumes instantaneous contrast agent mixing, which is physiologically unrealistic.
• This assumption introduces bias in tracer kinetic analysis, particularly overestimating blood flow.

Purpose of the Study:

• To develop and validate a modified two-compartment model for DCE imaging analysis.
• To characterize and correct for the bias inherent in conventional compartment modeling.
• To improve the accuracy of estimating microcirculatory parameters from DCE data.

Main Methods:

• A modified lumped two-compartment exchange model was derived from a distributed-parameter model accounting for spatial gradients.
• A formula was developed to compute bias-corrected flow using estimated apparent flow and permeability-surface area product.
• Noise-free DCE curves simulated from an axially distributed reference model were analyzed to evaluate accuracy.

Main Results:

• The modified model demonstrated structural identifiability from tissue residue data.
• Bias-corrected flow estimation showed a significant improvement, with deviations of (11.2 ± 6.4)% compared to (105 ± 21)% without correction.
• Other parameters like permeability-surface area product, vascular, and interstitial volumes were accurately estimated with minimal deviations after bias correction.

Conclusions:

• Physiologically relevant tissue parameters are accurately estimable using the modified two-compartment model with bias correction.
• The accuracy of bias-corrected flow is comparable to other unbiased parameters, offering a significant advantage over conventional modeling.
• This approach enhances tracer kinetic analysis in both preclinical and clinical DCE imaging studies.