Dynamic contrast-enhanced MRI: Study of inter-software accuracy and reproducibility using simulated and clinical data

Area of Science:

• Medical Imaging
• Radiology
• Biophysics

Background:

• Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is crucial for assessing tissue perfusion and vascularity.
• Pharmacokinetic parameters derived from DCE-MRI, such as K(trans) and ve, provide quantitative insights into tissue characteristics.
• Standardization of DCE-MRI analysis software is essential for reliable and reproducible clinical interpretation.

Purpose of the Study:

• To evaluate the accuracy and reproducibility of pharmacokinetic parameter measurements across five different analysis software packages (SPs) for DCE-MRI.
• To compare the performance of these SPs using both simulated and clinical DCE-MRI datasets.
• To identify potential sources of error and variability in DCE-MRI data analysis.

Main Methods:

• Retrospective analysis of 36 DCE-MRI scans from 27 rectal cancer patients and simulated DCE-MRI data.
• Quantitative analysis of pharmacokinetic parameters including volume transfer constant (K(trans)), extravascular extracellular volume fraction (ve), and initial area under the gadolinium curve (iAUGC).
• Statistical evaluation using Modified Bland-Altman analysis, intraclass correlation coefficients (ICCs), and within-subject coefficients of variation.

Main Results:

• Significant measurement errors were observed for all pharmacokinetic parameters across all tested SPs using simulated data.
• Intraclass correlation coefficients indicated moderate agreement for simulated data (K(trans): 0.50, ve: 0.67, iAUGC: 0.77).
• Very poor agreement was found between SPs for clinical data (K(trans): 0.10, ve: 0.16, iAUGC: 0.21), highlighting substantial inter-software variability.

Conclusions:

• The tested perfusion analysis software packages for DCE-MRI exhibit significant errors and poor inter-software reproducibility.
• The variability in measurements can lead to unreliable pharmacokinetic parameter estimations in clinical practice.
• Further standardization and validation of DCE-MRI analysis tools are necessary to improve the accuracy and consistency of cancer imaging assessments.