Model-based blind estimation of kinetic parameters in dynamic contrast enhanced (DCE)-MRI

Area of Science:

• Medical Imaging
• Biophysics
• Pharmacokinetics

Background:

• Dynamic contrast-enhanced MRI (DCE-MRI) is crucial for assessing tissue perfusion and vascularity.
• Accurate estimation of the arterial input function (AIF) is essential for reliable pharmacokinetic modeling in DCE-MRI.
• Current methods for AIF estimation can be complex and may require invasive procedures.

Purpose of the Study:

• To develop and validate a novel, non-invasive method for simultaneously estimating the AIF and pharmacokinetic parameters from DCE-MRI data.
• To assess the accuracy and robustness of the proposed algorithm using computer simulations and clinical data from sarcoma patients.

Main Methods:

• Developed an iterative blind estimation algorithm utilizing a parameterized functional form for AIF modeling.
• Employed k-means clustering to categorize tissue time-concentration measurements into characteristic curves.
• Validated the method through computer simulations assessing noise sensitivity and initial estimate impact.
• Compared estimated pharmacokinetic parameters (Ktrans, kep) with reference values derived from a measured AIF in 12 sarcoma patients.

Main Results:

• The algorithm successfully estimated AIF and pharmacokinetic parameters simultaneously from DCE-MRI data.
• In sarcoma patients, estimated Ktrans and kep values showed no significant difference compared to reference values (P=0.27 and P=0.08, respectively).
• Inclusion of arterial voxels in the blind estimation yielded an AIF comparable to the measured input function.

Conclusions:

• The developed iterative blind estimation algorithm provides accurate and reliable simultaneous estimation of AIF and pharmacokinetic parameters in DCE-MRI.
• This method offers a robust alternative for quantitative analysis of DCE-MRI data, particularly in oncology applications.
• Further investigation is warranted for AIF estimation in specific scenarios, such as when excluding vascular contributions, which may reflect local AIF characteristics.