Quantitative diffusion MRI has many important research and clinical applications. To assess the accuracy and reproducibility of these techniques across sites and vendors, diffusion phantoms are needed as a tool to enable testing under highly controlled conditions. Phantoms are needed for technique development, data harmonization in multi-center trials and quality assurance in both research and clinical applications. Ideally, a diffusion phantom should provide desirable signal behavior that include single-peak NMR spectrum, Gaussian diffusion, and a wide range of tunable apparent diffusion coefficient (ADC) values at a well-controlled temperature (eg: 0°C in an ice-water bath). A recently proposed phantom based on acetone-D₂O satisfies these desirable properties by using MR-invisible D₂O to modulate the ADC of acetone (which provides the signal)^[1]. However, this phantom may be limited by possible exchange between deuterium (from D₂O) and hydrogen (from acetone), giving rise to water signal after prolonged storage. Further, the relatively long T1 of acetone limits the signal-to-noise ratio (SNR) of the acetone-D₂O phantom. In this work, we developed and validated a novel diffusion phantom based on acetone-H₂O mixtures doped with MnCl2 in order to provide desirable signal behavior over a wide range of ADC. In this phantom, water modulates the ADC of acetone, while MnCl₂ serves the dual purpose of eliminating water signal (through T2 shortening) and shortening acetone T1, while maintaining long acetone T2. The combination of short T1 and long T2 of acetone is ideal for high SNR performance.

Phantom Construction: In order to test the ability of the novel acetone- H₂O-MnCl₂ phantom to eliminate water signal and shorten the T1 of acetone, while not altering the diffusion behavior of acetone, we built a matrix of acetone-H₂O phantoms with H₂O concentration of 5%,20%,40%(v/v) and MnCl₂ molar concentration of (4,2,1,0.5,0.25,0.125,0.0625mM). A 45ml cylindrical vial of such mixture was made for each combination. The combination of 5% H2O and 4mM MnCl₂ was excluded due to limited solubility of MnCl₂ at this H₂O concentration.

MR Spectroscopy and Imaging: The R2 (=1/T2) of water and acetone were measured using a multi-TE stimulated echo acquisition mode (STEAM) single-voxel spectroscopy sequence^[2]. Multiple sets of echo times ranging from 10-810ms were implemented to cover the wide range of R2 variation that water and acetone signals experience. To assess the diffusion behavior of the acetone-water-MnCl₂ phantom with different concentrations of MnCl₂, diffusion-weighted echo-planar imaging (DW-EPI) experiments were performed, with vials aligned in S/I direction. Images were acquired in axial plane, frequency encoding in R/L direction and diffusion direction in A/P, with b=50,100,150,200,300,400,500,750,1000,1500s/mm². To measure T1 of the phantoms, FSE-IR was performed with TR=9000ms,TI=100,200,400,800,1200,1600ms.

Data Processing and Analysis: R2 for both acetone and water were measured from the spectroscopy data using an offline joint-fitting algorithm^[3]. ADC maps were calculated from the DW-EPI acquisitions, ADC was measured in each vial by averaging ADCs of voxels in a single region-of-interest (ROI) in a central slice. T1 maps were generated from FSE-IR images, followed by ROI-T1 begin generated for each vial in the same way as ADC.

Figure 1 shows R2water and R2acetone for increasing concentrations on MnCl2, demonstrating the rapid increase of R2water but much slower increase of R2acetone. This differential effect of MnCl₂ on water signal and acetone is highly desirable: the high r2 relaxivity of MnCl₂ in water eliminates water signal at TE values used for diffusion MRI. Meanwhile, the modest r1 and r2 relaxivity of MnCl₂ on acetone leads to favorable shortening of T1 of acetone and minimal T2 shortening of acetone, both ideal for high SNR performance. This behavior is confirmed in Figure 2, which shows DW images and ADC maps in several vials with increasing MnCl₂ concentration and constant water concentration (40%). This figure demonstrates the elimination of water signal through T2 shortening as MnCl₂ concentration increases. In addition, the acetone signal increases with increasing MnCl₂, demonstrates the T1 shortening effects of MnCl₂ on acetone.

Figure 3 demonstrates that ADC of acetone can be modulated over a wide range by varying the water concentration, independently of MnCl₂ concentration.

We have proposed a novel diffusion MRI phantom based on acetone-water mixtures doped with MnCl₂. We validated this phantom using both relaxometry and quantitative diffusion measurements. By overcoming the limitations of a previously proposed acetone-D₂O phantom, the proposed phantom may provide improved stability (ie: extended shelf-life) and increased SNR in DW-EPI. Future work will evaluate the long-term stability of the novel phantom. In summary, the proposed acetone-H₂O-MnCl₂ phantom may have applications in the technical development and quality assurance of quantitative diffusion MRI.