Evaluation of highly undersampled contrast-enhanced MR angiography (SPARSE CE-MRA) in intracranial applications
Marcel Gratz1,2, Marc Schlamann3,4, Sophia Göricke4, Stefan Maderwald2, and Harald H. Quick1,2

1High Field and Hybrid MR Imaging, University Hospital Essen, Essen, Germany, 2Erwin L. Hahn Institute for Magnetic Resonance Imaging, University Duisburg-Essen, Essen, Germany, 3Neuroradiology, University Hospital Giessen and Marburg GmbH, Giessen, Germany, 4Department of Diagnostic and Interventional Radiology and Neuroradiology, University Hospital Essen, Essen, Germany

Synopsis

Highly undersampled contrast-enhanced intracranial MR angiography (SPARSE CE-MRA) was evaluated regarding the vascular visibility and image quality in a clinical setting on a 3T MRI system for 23 patients with various pathologies. The overall performance is comparable to TOF MR angiography, yet with much shorter acquisition times and a high-resolution whole-head coverage in both arterial and venous phase. However, a strong dependence of the obtained MRA quality on the bolus timing of the contrast agent was found and needs to be properly addressed by the operators to minimize the arterio-venous overlap in the imagery for best diagnostic quality.

Purpose

Highly undersampled contrast-enhanced MR angiography (SPARSE CE-MRA)1 with high spatial resolution and full head coverage was set up and evaluated in a clinical environment for different intracranial pathologies. Thus, potentials and pitfalls were identified and compared to the performance of conventional time-of-flight (TOF) MR angiography.

Methods

Twenty-three patients, 19 of them with intracranial pathologies, underwent a routine MR Head protocol at 3T (MAGNETOM Skyra, Siemens Healthcare, Erlangen, Germany) including the application of contrast agents, which was extended by three identical SPARSE MRA runs (at the pre-contrast, arterial and venous post-contrast phase). These prototype sequence provide high spatial resolution (0.7mm³ isotropic) and large spatial coverage (field of view: 265 x 232 x 214 mm³) while keeping the acquisition times at a few seconds only (TA = 10s) using an overall acceleration factor of 30 together with a spiral phyllotaxis sampling pattern) as needed particularly for the acquisition of a pure arterial phase. Although an inline reconstruction based on a redundant Haar wavelet based L1-regularized iterative SENSE algorithm with modified FISTA algorithm was available1, the obtained 3D raw datasets were iteratively reconstructed by a custom-written batch script after working hours due to their rather long reconstruction times of about 6 to 8 minutes per dataset for the current setup. Subsequently, all 23 patient datasets were presented to two radiologists and evaluated using 24 parameters and vascular segments on a 5-point ordinary scale (5 – very good vascular visibility/quality, 1 – insufficient visibility/quality). Where available, these results obtained from the SPARSE MRA sequences were compared to the performance of intracranial TOF MR angiography (TA=5 min, field of view: 128 x 117 x 68 mm³, spatial resolution 0.3x0.3x0.5mm³) using identical rating criteria.

Results

The mean overall rating of the SPARSE MRA across all patients was found to be 3.50 ± 1.07. The average rating across different features was 3.56 ± 0.95. Here, segments such as the carotid bifurcation (4.36 ± 0.75) and AVM nidus (4.33 ± 0.52), if applicable, were found to be represented best whereas worst ratings were observed for the ophthalmic artery (2.09 ± 0.81) and superior cerebellar artery (1.87 ± 0.91). Interobserver reliability as provided by Cohen's κ was 0.54, while for 65.1% of the ratings there was full agreement and 34.7% only differed by one point. A significant dependence of the image quality and vascular visibility on the bolus timing was observed. Thus, a separated analysis of the patients with correct and inaccurate timing, respectively, revealed average image quality scores of 3.76 and 3.15 across patients. These rating differences were mainly caused by the difference of the depiction of the arteries in the image data. The average performance of intracranial TOF was rated slightly better than for SPARSE MRA (3.84 ± 0.87 across all patients, 3.54 ± 0.62 across all features).

Discussion and Conclusion

SPARSE MRA sequences in this application provide a very fast overview of the whole cranial vascular system with high spatial resolutions, which proves to be particularly useful for patients with vascular pathologies such as stenosis, fistulae and arterio-venous malformations (AVMs) and in restless patients. Although, intracranial TOF was found to have a slightly better performance, acquisition times of TOF protocols are much longer compared to SPARSE MRA. Moreover the used TOF protocol provides only images of the arterial phase or (in the presence of contrast agents) a mixed arterio-venous phase. Furthermore, within clinically acceptable scan times only a section of the intracranial structure can be acquired with TOF whereas SPARSE MRA provides whole-head coverage. Thus, SPARSE MRA is able to reveal segments such as A.lingualis that are commonly not depicted in TOF datasets. However, special care must be taken for contrast bolus timing to obtain well separated arterial and venous phases, respectively, with a minimum of overlap for best diagnostic image quality.

Acknowledgements

The authors would like to gratefully acknowledge the support of Michaela Schmidt and Dr. Aurelien Stalder, both Siemens Healthcare GmbH, Erlangen, Germany in providing and implementing the sparse MRA sequence and for providing further support and valuable feedback during data acquisition.

References

1. Stalder et al., Magn Reson Med (2014)

Figures

Fig. 1: Ratings for the representation of vascular details in both TOF and SPARSE sequences.

Fig. 2: Patient with a very small AVM (solid arrow) in the cerebellum. Feeding artery is likely the PICA (dashed arrow). The arterial phase (A-C) is zoomed in on the pathology (D-F). The venous phase is depicted in the bottom row (G-I).

Fig. 3: Comparison of SPARSE MRA (A) and TOF datasets (B) of a patient with a small aneurysm (arrow). SPARSE MRA here provides extended angiographic coverage within short acquisition time (10 s) when compared to TOF MRA (5 min).



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
4219