Tracer kinetic model-driven registration for dynamic contrast-enhanced MRI time-series data

Area of Science:

• Medical Imaging
• Biophysics
• Radiology

Background:

• Physiological motion during dynamic contrast-enhanced MRI (DCE-MRI) acquisition introduces significant errors in tracer kinetic model fitting.
• Conventional motion correction methods struggle with DCE-MRI due to contrast enhancement altering image content.

Purpose of the Study:

• To develop and validate a novel methodology for motion correction in DCE-MRI data.
• To improve the accuracy of tracer kinetic model fitting by accounting for motion during registration.

Main Methods:

• A tracer kinetic model-driven registration methodology was developed, incorporating a contrast enhancement model.
• The iterative registration procedure focused on a tumor volume of interest (VOI) using 3D translational motion correction.
• The method was tested on a DCE-MRI software phantom and patient datasets with significant motion.

Main Results:

• The methodology accurately removed motion corruption in the DCE-MRI software phantom.
• Model fitting errors were reduced, and localization in 3D parameter maps improved in patient data.
• The approach successfully salvaged clinical trial DCE-MRI datasets affected by motion.

Conclusions:

• Tracer kinetic model-driven registration effectively addresses motion artifacts in DCE-MRI.
• This method enhances the reliability and utility of DCE-MRI data, particularly in clinical trials.
• The developed technique offers a robust solution for motion correction in quantitative MRI analysis.