Modeling the residue function in DSC-MRI simulations: analytical approximation to in vivo data

Area of Science:

• Medical Imaging
• Biophysics
• Computational Modeling

Background:

• Dynamic susceptibility contrast (DSC) MRI perfusion quantification commonly uses an exponential residue function.
• This assumption may be inaccurate in pathological conditions where microvascular hemodynamics are altered.
• Inaccurate models can lead to inappropriate estimations of DSC-MRI quantification accuracy.

Purpose of the Study:

• To characterize in vivo residue function variations in normal and infarcted tissue.
• To compare the accuracy of different models for residue function approximation in DSC simulations.
• To determine the most appropriate model for DSC-MRI simulations in atherosclerotic disease.

Main Methods:

• In vivo residue functions were measured in patients with atherosclerotic disease.
• A nonparametric Control Point Interpolation method was used for robust residue function characterization.
• Observed residue functions were approximated using exponential, bi-exponential, Lorentzian, Fermi, and a Vascular Model.

Main Results:

• The bi-exponential function provided the lowest approximation error for in vivo residue functions.
• This finding was consistent for both normal and infarcted tissue samples.
• The bi-exponential model demonstrated superior performance over other tested models.

Conclusions:

• A bi-exponential model is recommended for future numerical simulations of DSC-MRI.
• Replacing the traditional exponential function with a bi-exponential model will improve simulation accuracy.
• This improvement is crucial for reliable perfusion quantification in various tissue conditions.