Blind deconvolution estimation of an arterial input function for small animal DCE-MRI

Area of Science:

• Magnetic Resonance Imaging
• Pharmacokinetic Modeling
• Preclinical Research

Background:

• Quantitative dynamic contrast-enhanced (DCE) MRI relies heavily on accurate arterial input function (AIF) determination.
• Preclinical small animal DCE-MRI faces challenges in precise AIF measurement due to artifacts.
• Advanced pharmacokinetic models demand higher accuracy in AIF estimation for parameters like blood flow and permeability-surface area product.

Purpose of the Study:

• To address the challenge of AIF estimation in small animal DCE-MRI using advanced pharmacokinetic models.
• To present a novel method for AIF estimation based on blind deconvolution tailored for small animal physiology.
• To improve the reliability of perfusion parameter estimation in preclinical DCE-MRI.

Main Methods:

• Development of a parametric AIF model specifically designed for small animal physiology.
• Application of multichannel blind deconvolution techniques to estimate AIF parameters.
• Validation using simulated data and real DCE-MRI data with contrast agents of varying molecular weights.

Main Results:

• Blind deconvolution AIF estimation achieved comparable results to using the true AIF under realistic signal-to-noise ratios in simulations.
• Real data evaluation demonstrated consistency with known effects of contrast agent molecular weight.
• The proposed method provides reliable AIF estimates crucial for advanced pharmacokinetic modeling.

Conclusions:

• Multi-channel blind deconvolution with the proposed AIF model offers a reliable approach for small animal DCE-MRI.
• This method enables accurate perfusion parameter estimation even under realistic signal-to-noise conditions.
• The findings support the use of this technique for advanced pharmacokinetic modeling in preclinical research.