Single-channel blind estimation of arterial input function and tissue impulse response in DCE-MRI

Area of Science:

• Medical Imaging
• Biophysics
• Pharmacokinetics

Background:

• Dynamic MRI and pharmacokinetic modeling are crucial for estimating capillary perfusion parameters.
• Established methods like curve fitting and nonblind deconvolution rely on accurate arterial input function (AIF) and tissue tracer concentration.
• Nonblind methods are susceptible to AIF measurement or estimation errors.

Purpose of the Study:

• To develop and validate a single-channel blind deconvolution algorithm for dynamic MRI.
• To improve the accuracy of perfusion parameter estimation by mitigating AIF-related errors.
• To present a method that uses only a single tissue tracer concentration function.

Main Methods:

• Developed a single-channel blind deconvolution algorithm.
• The algorithm estimates the AIF and tissue impulse response function from a single tissue tracer concentration function.
• Validated the algorithm using simulations and real mouse data.
• Presented corresponding nonblind and multichannel methods for comparison.

Main Results:

• The single-channel blind deconvolution algorithm successfully estimates AIF and tissue impulse response functions.
• This method reduces errors associated with traditional nonblind deconvolution techniques.
• Validated accuracy through simulations and experimental data from mice.

Conclusions:

• Single-channel blind deconvolution offers a robust alternative for dynamic MRI perfusion analysis.
• The proposed method enhances accuracy by minimizing reliance on external AIF measurements.
• This technique holds promise for more reliable assessment of leaky capillary perfusion.