Quantifying Perfusion Properties with DCE-MRI Using a Dictionary Matching Approach

Abstract

perfusion properties can be estimated from pharmacokinetic models applied to Dce-MRi data using curve fitting algorithms; however, these suffer from drawbacks including the local minimum problem and substantial computational time. Here, a dictionary matching approach is proposed as an alternative. Curve fitting and dictionary matching were applied to simulated data using the dual-input single-compartment model with known perfusion property values and 5 in vivo Dce-MRi datasets. In simulation at SNR 60 dB, the dictionary estimate had a mean percent error of 0.4-1.0% for arterial fraction, 0.5-1.4% for distribution volume, and 0.0% for mean transit time. The curve fitting estimate had a mean percent error of 1.1-2.1% for arterial fraction, 0.5-1.3% for distribution volume, and 0.2-1.8% for mean transit time. In vivo, dictionary matching and curve fitting showed no statistically significant differences in any of the perfusion property measurements in any of the 10 ROIs between the methods. In vivo, the dictionary method performed over 140-fold faster than curve fitting, obtaining whole volume perfusion maps in just over 10 s. This study establishes the feasibility of using a dictionary matching approach as a new and faster way of estimating perfusion properties from pharmacokinetic models in Dce-MRi. Dynamic contrast-enhanced (DCE) MRI data can be used with a variety of pharmacokinetic models to estimate perfusion properties through either an ROI-based or voxel-based analysis. However, many models are complex, and thus determination of fitted variables can be challenging, particularly when employing voxel-wise analysis as several thousand voxels must be evaluated to estimate many model parameters simultaneously, which poses significant computational burdens 1-5. Curve fitting algorithms are used to estimate properties of interest 6-8 , but this approach has many potential drawbacks. For instance, these algorithms can be extremely computationally expensive and require many hours to process just one dataset, which has led to the exploration of alternatives such as the linear least squares method 9-11. Curve fitting algorithms often also have numerous configuration options including initial property guesses, property bounds, algorithm choice, tolerances, and cost functions. The large number of available configuration options and underreporting of precise configurations causes difficulties in replicating perfusion modeling results across institutions. Lastly, curve fitting algorithms are vulnerable to converging on local minima 11. When this occurs, the best fit to the model is not optimally identified, resulting in an inaccurate estimate of perfusion properties. Furthermore, these occurrences may be difficult to identify. These drawbacks have significant consequences. Heye, et al. 12 have shown that there is considerable variation in perfusion properties quantified across analysis platforms such as Tissue4D (Siemens,