Increased robustness in reference region model analysis of DCE MRI using two-step constrained approaches.

Area of Science:

• Medical Imaging
• Biophysics

Background:

• Dynamic contrast-enhanced MRI (DCE-MRI) is vital for tumor perfusion quantification.
• Reference region models (RRMs) offer an alternative to arterial input function-dependent methods.
• A key RRM parameter is linked to the reference region across all voxels.

Purpose of the Study:

• To develop and evaluate a two-step constrained reference region model (CRRM) approach for DCE-MRI.
• To improve the accuracy and precision of tumor perfusion quantification.
• To leverage the shared parameter constraint within RRMs.

Main Methods:

• Three constrained RRM (CRRM) methods were developed.
• Simulations assessed CRRM accuracy and precision under varying noise and temporal resolutions.
• CRRM was validated using a virtual phantom and in vivo data from breast cancer and soft tissue sarcoma.

Main Results:

• CRRM demonstrated improved precision and accuracy, particularly at low signal-to-noise ratios in simulations.
• The virtual phantom showed CRRM successfully fitted voxels with kinetics similar to reference tissue.
• In vivo data revealed reduced variability and better agreement with the Tofts model using CRRM.

Conclusions:

• The proposed two-step CRRM fitting approach effectively reduces variability in DCE-MRI perfusion estimates.
• CRRM offers a more robust method for quantifying tumor perfusion.
• This technique enhances the reliability of DCE-MRI in clinical settings.