Purpose: To quantify the errors involved in calculating dynamic parameters (K(trans) and ve) from dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) scans, and to develop alternative analyses to improve accuracy or increase processing speed.
Materials and methods: This paper presents three different ways of handling the discrete samples of the arterial input and tissue response data with increasing fidelity, with which this continuous arterial input function (AIF) is represented. Also, a new noniterative approach to parameter estimation was developed from one used previously for analysis of radioactive tracer concentrations in radioangiography. The analysis methods were tested using simulated data.
Results: The more sophisticated schemes for data processing give more accurate parameter estimates when data are sparsely sampled, at least for the AIF that we modeled. The noniterative algorithm is very rapid in execution, but was more susceptible to measurement errors.
Conclusion: The improved algorithms presented should be useful when the AIF and tissue response are sparsely sampled. The noniterative approach may be suitable for semiquantitative visualization, or where the AIF and tissue response are sampled accurately and with a small time interval between samples.
Copyright 2004 Wiley-Liss, Inc.