Reply:

We agree that the differences in technique (2-curve histogram modeling versus minimum ADC region-of-interest analysis) may account for the discrepant conclusions, meriting further investigation in larger datasets.