Synopsis

Keywords: Phantoms, Relaxometry

A potential low-cost quantitative MR image assessment phantom was investigated by calculating relaxation times and the magnetization transfer ratio over multiple scans. An imaging protocol comprised of an inversion recovery sequence with varying inversion times, a spin echo with varying echo times, and an Enhanced Fast Gradient Echo 3D sequence with and without a magnetization transfer pulse, was developed and used to acquire images of the phantom, a block of store-bought jelly. The imaging data was used to calculate a T1 time of 109±1ms, a T2 time of 12.8±0.2ms, and an MTR value of 17.4±0.2%.

Introduction

With the continued expansion of quantitative MR imaging, which can measure material information beneficial to diagnostic imaging^1,2, methods to understand inter- and intra-system repeatability and stability are becoming essential additions to the workflow of an MR scanner. While commercial T₁ and T₂ relaxation phantoms are available³, and the preparation of gels, such as agar, have been extensively documented⁴, these options are often costly or require precise manufacturing measurements to ensure stability. This study aimed to investigate the use of store-bought packaged jelly as a potential phantom for imaging characteristics, first proposed by Rua et al⁵. As well as being low-cost and readily available, food goods within the United Kingdom are regulated by the Food Standards Agency (FSA), providing a level of consistency to each product. T₁ and T₂ relaxation times, and magnetization transfer ratios (MTRs) of the jelly phantom were calculated over multiple sessions to assess the stability of these parameters.

Methods

MRI was performed on a 3T GE Signa Premier (GE Healthcare, Chicago, Illinois), using a 19 channel Head Coil with the jelly packet (Hartley’s, Histon, England) taped to a Braino sphere, as shown in Figure 1.
2D inversion recovery images were acquired using inversion times (TI): 50, 60, 75, 100, 200, 250, and 500 ms. T₂ was characterised using a spin echo sequence with echo times (TE): 10, 15, 20, 40, 60, and 80 ms. The following imaging parameters were used for all relaxation data acquired: Number of slices=1, acquisition matrix size = 128 x 128, reconstructed matrix size= 256 x 256, slice thickness = 5 mm, field of view (FOV) = 256 mm, and bandwidth= 122 Hz/pixel. MTR values were calculated using an Enhanced Fast Gradient Echo 3D sequence (EFGRE3D) with and without a fermi magnetization transfer pulse of duration 8 ms, offset by 1200 Hz, and a flip angle of 200°. The imaging parameters of the EFGRE3D sequence were: number of slices = 86, slice thickness = 2 mm, FOV = 256 mm, acquisition matrix size = 320 x 320, bandwidth = 244 Hz/Pixel, TI = 32 ms, and TE = 2.77 ms.Voxel-wise curve fitting was performed on the resultant relaxation data to allow the creation of T₁ and T₂ maps respectively of the phantom in MATLAB (2022a, MathWorks, Cambridge, MA), displayed in Figure 2. Equations 1 and 2 were used for fitting the T₁ and T₂ data respectively, where M_xy is the signal intensity at each voxel location, and M₀ is the maximum signal. MTR values for the phantom were calculated using Equation 3, where SI_MTon and SI_MToff are the signal intensities with the magnetization pulse turned on and off respectively.

(Equation 1:) $$M_{xy}=\vert{a-(b\times\exp{(\frac{-TI}{T_1}}))}\vert$$

(Equation 2:) $$M_{xy}=M_0\times(\exp{(\frac{-TE}{T_2})})+c$$

(Equation 3:) $$MTR=100\times\frac{SI_{MToff}-SI_{MTon}}{SI_{MToff}}$$


For consistency in the creation of each region of interest (ROI), used to calculate the overall characteristic values, automated segmentation tools were developed in FIJI⁶ to create ROIs accurately around the phantom, with a uniform 3mm reduction applied to the ROIs to avoid impact from partial volume effect at the boundaries. The relaxation data was acquired in four separate sessions over the period of one month, while the MTR data was acquired during a single session with all components re-setup between scans. The same jelly phantom and MRI scanner was used for all data collection. MTR values were calculated over five slices throughout the jelly to assess any variations in results, with the data averaged to generate the final MTR %.

Results

The average T₁ and T₂ relaxation times calculated in this study for the jelly phantom were 109±1 ms and 12.8±0.2 ms respectively. The average MTR value for the jelly phantom was calculated as 17.4±0.2 %. While the inter-slice variation in the mean was low, larger variations in MTR values were observed within each studied slice, with an average standard deviation of 5.4. The full set of T₁ and T₂ times, and MTR values for each dataset is provided in Table 1.

Discussion

Over the course of 1 month, the coefficient of variation for the T₁ and T₂ times calculated was 1.43% and 2.46% respectively. No trends were observed with the calculated relaxation times over time suggesting good phantom stability. A comparison with estimated errors from components within a commercially available characteristics phantom with similar relaxation times³ suggests jelly could be a very cost-effective phantom; however, further investigation is needed to assess inter-scanner variations, including different scanner vendors, temperature dependencies, and potential differences between manufacturer batches of jelly. MTR values also provided high consistency when comparing the average values obtained over the complete phantom; however, individual slice measurements displayed a larger variation in results. Further investigation highlighted that this was a potential result of noise, with a gaussian distribution observed around the averaged MTR value, as shown in Figure 3. Further studies are being conducted to assess the MTR stability over longer time periods.

Conclusions

The calculated T₁ and T₂ relaxation times, and MTR % values, suggest that store-bought jelly is suitable for use as a pre-made image characteristics phantom. While further studies are still required, the current results indicate potential for the use of pre-packaged food products in quantitative MR imaging assessment.

Acknowledgements

This work was supported by the National Institute for Health and Care Research, and the British Heart Foundation [Grant Number RG/F/21/110035]