Distributed capillary adiabatic tissue homogeneity model in parametric multi-channel blind AIF estimation using...

Area of Science:

• Medical Imaging
• Biophysics
• Quantitative MRI

Background:

• Accurate estimation of the arterial input function (AIF) is crucial for quantitative dynamic contrast-enhanced (DCE) MRI.
• Traditional AIF estimation methods often neglect critical factors like contrast dispersion, partial volume effects, and flow artifacts, or patient variability.

Purpose of the Study:

• To address the challenges in AIF estimation for quantitative DCE MRI.
• To introduce and evaluate an advanced multi-channel blind deconvolution method for improved AIF estimation.

Main Methods:

• Employed multi-channel blind deconvolution, bypassing the need for direct arterial sampling.
• Utilized the distributed capillary adiabatic tissue homogeneity (DCATH) model for a more realistic impulse residue function.
• Investigated an alternative AIF model and various AIF-scaling techniques.

Main Results:

• The proposed method demonstrated consistent results on both synthetic and clinical data (renal cell carcinoma patients).
• Performance was evaluated concerning the number of tissue regions and signal-to-noise ratio.
• Initial clinical data analysis suggests more reliable and noise-resilient perfusion parameter estimation.

Conclusions:

• Multi-channel blind deconvolution, particularly with the DCATH model, presents a promising approach for AIF estimation in clinical settings.
• This method offers a potential solution to overcome limitations of conventional AIF estimation techniques.