Synopsis

Measurement of hepatic arterial perfusion is important for assessment of chronic liver diseases and characterization of liver lesions. The aim of this study was to investigate the capability of a 3D-PCASL sequence for measurement of arterial, portal-venous and global perfusion of the liver under free respiration. It is demonstrated that the presented method provides high quality ASL perfusion images of the liver under free breathing conditions. Results in a patient with hepatocellular carcinoma clearly indicate high and quantitatively measurable arterial perfusion of lesions, although direct measurement of the clearly lower ASL signal of arterial perfusion in normal liver tissue is usually limited by background noise.

Purpose

Diagnosis of various liver diseases (e.g., adenoma, carcinoma, metastases) often involves MRI examinations with contrast media for assessment of the perfusion status. Arterial spin labeling (ASL) MRI has been demonstrated to be a promising and non-invasive alternative approach for cerebral blood flow measurements.¹ However, with transfer of the ASL technique to liver perfusion imaging, there are certainly more challenges due to the physiological motion during the respiratory cycle and effects such as cardiac pulsations and peristalsis of the bowel.² Furthermore, perfusion of the liver is unique because this organ receives blood via two distinct circulatory routes: arterial and portal-venous. The separation of two perfusion fractions is important for the evaluation of liver diseases, especially for the diagnosis of hepatocellular carcinoma.³ The separation of two perfusion fractions using ASL is difficult because of the complex geometry of the blood supplying vessels. Recently, 2D pseudo-continous ASL (PCASL) sequences were applied for separate measurements of arterial and portal-venous fractions of the human liver perfusion.^4,5

3D ASL sequences are the prefered technique since they provide nearly optimal SNR and are relative insensitive to magnetic field inhomogeneities.⁶ Furthermore, compared with 2D multislice acquisition, the 3D-ASL sequences allow for significantly better background suppression (BS).⁷

Thus, the scope of this preliminary study was to investigate the capability of a segmented 3D-PCASL sequence for the separate measurement of arterial and portal-venous perfusion in healthy volunteers under free respiration. In addition, the method was applied in a patient with hepatocellular carcinoma and compared with the contrast enhanced MRI results.

Methods

Five healthy volunteers were examined on a 3T MR scanner (MAGNETOM Prisma^Fit, Siemens Healthcare). The perfusion of the liver was measured by using a prototype PCASL sequence with a 3D single-shot Turbo Gradient Spin Echo (TGSE) data acquisition. Post labeling delay (PLD) and tag duration (TD) were set to 1.5s and tagging flip angle was 20°. A series of four non-selective hyperbolic secant inversion pulses was used for BS.⁸ Data acquisition was performed with following parameters: TR, 5500-7000ms; TE, 24.6ms; bandwidth, 3065Hz/Pixel; matrix (phase×readout×slice), 48×96×24; FoV, 192×384×144mm³; voxel-size, 4×4×6mm³. Twenty eight label-control image pairs and a M0 scan were acquired within 5:16-6:39min under free breathing condition.

In order to separate arterial and portal-venous perfusion fractions, various orientations and positions of the tagging plane were used.⁹ Arterial perfusion was measured by tagging the aorta in an axial plane, cranial to the liver (Figure 1, plane 1). For labeling of the global liver perfusion, a tagging plane crossing portal vein and hepatic artery was applied (plane 2), and portal-venous perfusion was assessed by positioning a tagging plane below the hepatic artery (plane 3).

In a patient with hepatocellular carcinoma, arterial and venous phase images were acquired after administration of contrast media using a 3D-VIBE (volumetric interpolated breath-hold examination) T1-weighted gradient echo sequence.

Image analysis was performed on a standalone PC using MATLAB (MathWorks, Natick, MA, USA). Original ASL images (M₀,M_ctrl,M_tag) were 4-fold interpolated and perfusion-weighted images ((M_ctrl-M_tag), (M_ctrl-M_tag)/M₀) were calculated. VIBE data was reshaped to match the slice thickness and position of ASL images.

Results

PCASL images of the liver with high quality could be obtained for all volunteers under free respiration. Figure 2 shows exemplary six slices of 3D-PCASL data set of a healthy volunteer: arterial (A), global (B) and portal-venous (C) perfusion-weighted images. The arterial perfusion fraction of the liver led to signal intensities close to the noise level, whereas the global and portal-venous perfusion showed clearly higher signal intensity. Proton-density (A), control (B), labeling (C) and perfusion-weighted subtraction (D) images of the global liver blood flow of another healthy volunteer are shown in Figure 3.

In Figure 4, PCASL perfusion images of a patient with liver cirrhosis are displayed. Arterial perfusion fraction (A) and global perfusion (B) images of the liver are in good agreement with T1-weighted arterial (C) and portal-venous (D) phase images acquired by a 3D-VIBE sequence after contrast agent injection.

Discussion

In the presented study the feasibility of a 3D-PCASL technique for perfusion imaging of the liver could be demonstrated. The combination of the PCASL sequence with 3D-TGSE readout provides acquisition of high quality perfusion images of the liver under free respiration. Arterial, portal-venous and global perfusion of the liver could be measured by appropriate selection of the tagging plane position.

Our results indicate that although a direct measurement of a relatively low arterial perfusion signal of normal liver could be limited by background noise, this approach may nevertheless be beneficial in patients with highly vascularized liver lesions. An efficient background suppression as well as motion correction is essential for high image quality.

Acknowledgements

References

1. Luh WM et al. QUIPSS II with thin-slice TI1 periodic saturation: a method for improving accuracy of quantitative perfusion imaging using pulsed arterial spin labeling. Magn Reson Med 1999;41:1246-54.
2. Robson PM et al. Strategies for reducing respiratory motion artifacts in renal perfusion imaging with arterial spin labeling. Magn Reson Med 2009;61:1374-87.
3. Pandharipande PV et al. Perfusion imaging of the liver: current challenges and future goals. Radiology 2005;234:661-73.
4. Schalkx HJ et al. Arterial and portal venous liver perfusion using selective spin labelling MRI. Eur Radiol 2015;25:1529-40.
5. Martirosian P et al. Spatial-temporal perfusion patterns of the human liver assessed by pseudo-continuous arterial spin labeling MRI. Z Med Phys 2018. doi: 10.1016/j.zemedi.2018.08.004.
6. Alsop DC, et al. Recommended implementation of arterial spin-labeled perfusion MRI for clinical applications: a consensus of the ISMRM perfusion study group and the European consortium for ASL in dementia. Magn Reson Med 2015;73:102–16.
7. Ye FQ, et al. Noise reduction in 3D perfusion imaging by attenuating the static signal in arterial spin tagging (ASSIST). Magn Reson Med 2000;44:92–100.
8. Robson PM et al. Volumetric arterial spin-labeled perfusion imaging of the kidneys with a three-dimensional fast spin echo acquisition. Acad Radiology 2016;23:144-54.
9. Martirosian P et al. Measurement of liver perfusion using pseudo-continuous arterial spin labeling with background suppression: approaches to separate portal-venous and arterial perfusion. In Proceedings of the 24th Annual Meeting of ISMRM, Singapore, 2016. p. 2944.