Introduction

T₁ and T₂ mapping of abdominal organs and using tissue relaxation times to characterize abdominal pathologies holds considerable promise, but is limited by technical difficulties. Current relaxometry methods involve long scan times and are not repeatable or reproducible.¹ Respiratory motion and B₁ (transmit field) inhomogeneity are additional challenges in the abdomen. MRF has the ability to simultaneously measure multiple tissue properties in a time-efficient manner.² Our previous work demonstrated an accurate T₁ and T₂ quantification method in the abdomen in a single breath-hold, which is an improvement on existing abdominal mapping techniques.³ However, there are limited data available on normative T₁ and T₂ relaxation times of abdominal organs. The purposes of this study were to optimize the abdominal MRF method and establish normative values of abdominal organs at 3T using 2D MRF. A secondary objective was to assess the repeatability of 2D MRF in the abdomen.

Methods

The 2D MRF technique in the abdomen has been described previously.⁴ It is based on fast imaging with steady state free precession technique (MRF-FISP) with a spiral readout.⁴ A total of 2304 time points were acquired for MRF acquisition and the acquisition time for each slice was ~22 sec with breath-hold. Compared to previous abdominal MRF implementation, a higher spatial resolution was achieved (1.25 mm vs. 1.9 mm) with a longer spiral readout (5.5 ms vs. 2.9 ms) and fat suppression module was applied to reduce the effect of fat on spiral blurring. In this IRB-approved study, 22 asymptomatic volunteers (6M/16F) with mean age 29 y.o. (range 20-58 y.o), height 166 cm (range 155-188 cm), and weight 66.2 kg (range 52.2-88.5 kg) were scanned on a 3T MR scanner (Siemens Skyra®) using 16 channel body coil after obtaining written informed consent. Three axial slices were prescribed on breath-hold coronal scouts of upper abdomen. The center slice was positioned at mid-pancreas and top and bottom slices were positioned 5 cm cranial and caudal to the center slice to cover liver, spleen and kidneys, respectively. Each slice was 5 mm thick. Raw data were processed in MATLAB and regions of interest were drawn by a trained radiologist (7 years of radiology experience) for right and left lobes of liver on all three slices, spleen and body of pancreas in center slice and cortex and medulla of the better visualized kidney in the bottom slice. T₁ and T₂ values for each ROI were obtained through a pattern-matching process and were compared to existing literature values for T₁ and T₂ relaxation times in abdominal organs at 3T.^3,5-7 For the repeatability analysis, 4 of the volunteers were scanned a second time between 20 and 50 days after their original scan with identical methodology, followed by a comparison of relaxation times for each organ between the first and second scans.

Results

Figure 1 shows representative T₁, T₂ and M₀ maps from one volunteer. The T₁ and T₂ (mean ± SD) values for liver, spleen, pancreas, kidney cortex, and kidney medulla for 22 healthy volunteers are shown alongside literature values in Table 1. Table 2 shows T₁ and T₂ values measured in the three axial slices in liver for left and right lobes. Results from the repeatability analysis are shown in Figures 2a and 2b.

Discussion

Establishing normative values for abdomen organs represents the first step for further studies in abdomen relaxometry. This work represents the largest number of healthy volunteers scanned to date employing a technique that allows simultaneous mapping of T₁ and T₂ in a time-efficient manner. Most of the T₁ and T₂ values derived from this study are in good agreement with T₁ and T₂ values derived from different volunteers in previous studies.^3,5-7 Compared to the results obtained in the previous 2D abdominal MRF study³, higher T₁ values were noticed in liver, spleen and kidney medulla in the current study. The reason for longer observed T₁ is not clear at this time; further exploration of this phenomenon is underway. The T₁ and T₂ for values for right and left lobes of liver for three axial sections are similar, suggesting intra-organ consistency. Repeat measurements for four volunteers scanned on two separate occasions also demonstrate agreement across the different organs scanned, suggesting that the method is repeatable. Further expansion of this dataset is ongoing.

Conclusion

MRF provides an opportunity to perform abdomen relaxometry and provide a reference for comparison and evaluation of various abdominal pathologies.

Acknowledgements

Research Support: NIH grants 1R01EB016728, 1R01BB017219, 1R01DK098503, T35DK111373, and Siemens Healthineers.