Technical Note: A safe, cheap, and easy-to-use isotropic diffusion MRI phantom for clinical and multicenter studies

Purpose: Since Diffusion Weighted Imaging (DWI) data acquisition and processing are not standardized, substantial differences in DWI derived measures such as Apparent Diffusion Coefficient (ADC) may arise which are related to the acquisition or MRI processing method, but not to the sample under study. Quality assurance using a standardized test object, or phantom, is a key factor in standardizing DWI across scanners. Methods: Current diffusion phantoms are either complex to use, not available in larger quantities, contain substances unwanted in a clinical environment, or are expensive. A diffusion phantom based on a polyvinylpyrrolidone (PVP) solution, together with a phantom holder, is presented and compared to existing diffusion phantoms for use in clinical DWI scans. An ADC vs. temperature calibration curve was obtained. Results: ADC of the phantom (808 to 857 +/- 0.2 mm(2)/s) is in the same range as ADC values found in brain tissue. ADC measurements are highly reproducible across time with an intra-class correlation coefficient of > 0.8. ADC as function of temperature (in Kelvin) can be estimated as ADC(m)(T) = [expd(-7.09).exp (-2903.8 (1/T - 1/293.55))] with a total uncertainty (95% confidence limit) of +/- 1.7%. Conclusion: We present an isotropic diffusion MRI phantom, together with its temperature calibration curve, that is easy-to-use in a clinical environment, cost-effective, reproducible to produce, and that contains no harmful substances. (C) 2017 American Association of Physicists in Medicine