Motivation: Diffusion weighted MRI (DWI) has potential as a quantitative imaging biomarker in oncology, but apparent diffusion coefficient (ADC) measurements vary across time, scanners, and protocols. Goal(s): Demonstrate feasibility of SI traceable in-vivo ADC calibration using reference standards embedded in the scanner couch. Approach: Nine subjects who had previously undergone clinical mpMRI for suspected prostate cancer underwent repeat DWI scans with a co-patient phantom containing reference diffusivity vials. Calibration functions were constructed to correct bias and calculate uncertainty. Results: Calibration functions were close to unity. Calibrated ADC maps and uncertainty were generated. Calibration slightly reduced variability in repeat scans. Impact: This work demonstrates the first in-vivo SI traceable ADC calibration, enabling bias correction and uncertainty quantification on a voxel-wise basis. This is a key step towards establishing ADC as a reliable quantitative imaging biomarker for multicenter oncology studies.
Bolton et al. (Tue,) studied this question.