3347

En route to multiphasic anthropomorphic MR phantoms: A new mold-based approach applying gel-based preparation to real MR-datasets geometries
Adriano Troia1, Umberto Zanovello1, Luca Zilberti1, Matteo Cencini2, Michela Tosetti2,3, David Kilian4, Martina Capozza5, Wolfgang Kilian6, and Tuğba Dışpınar Gezer7
1INRIM, Turin, Italy, 2IRCCS Stella Maris, Pisa, Italy, 3IMAGO7 Foundation, Pisa, Italy, 4Centre for Translational Bone, Joint and Soft Tissue Research, Faculty of Medicine and University Hospital TUD, Dresden, Germany, 5Department of Molecular Biotechnology and Health Sciences UNITO, Turin, Italy, 6Physikalisch-Technische Bundesanstalt, Braunschweig, Germany, 7National Metrology Institute, TUBITAK, Istanbul, Turkey

Synopsis

Anthropomorphic phantoms for MRI imaging are rapidly evolving also thanks to the increase of 3D printing technologies. Their realization may represent an important support for the implementation of advanced quantitative imaging technique such as Magnetic Resonance Fingerprinting (MRF)or Electrical Properties Tomography (EPT). Even if formulations of gel based tissue-mimicking materials have been widely explored, generally they lacked in achieving the simultaneous tuning of electrical and relaxation properties of the mimicked tissues. In this study, customized moulds of white and grey matter are combined to realize brain-like gel phantoms in which relaxation times, conductivity and permittivity have been measured.

Purpose

Phantoms for quantitative MRI have been largely studied. However, most of them take the form of simple vials or bottles filled with salt solutions or gels with tuneable relaxation properties. As a consequence, they don’t address other aspects, such as heterogeneity or complex geometries, typical of real tissues. Here we aim at realizing heterogeneous phantoms, with tuneable relaxation and electrical properties, exploiting the fabrication of 3D printed mouldsin order to obtain anthropomorphicbrain-like structures.This approach will allow to realize volumetric phantoms of realistic size with the possibility to test quantitative MR-based imaging technique like MRF and EPT.

Methods

Different types of naturally-derived polysaccharides and polymers have been used to realize heterogeneous phantom in order to simulate white and grey matter. Starting from the results reported in reference [1], combined blend of Agar and Gellan Gum have been used as a matrix in which T1 and T2 relaxation times can be modulated independently by varying the amount of Agar and GdCl3. Gellan gum has been used to increase the strength and mechanical stability of the gel, as well as to limit the diffusion phenomena of the paramagnetic salt due to its electrostatistically crosslinked gelation mechanism in the presence of divalent ions (i.eGd). To modulate the electrical properties, different amount of NaCl and Polyvinylpyrrolidone (PVP) have been used in the above composition, while, NaN3 has been added as preservative. Negative moulds of white (WM) and grey matter (GM) have been realised using soft and elastic polymer like RTV silicone or cryogel based on Polyvinyl alcohol. The polysaccharide-based solution has been poured in the moulds, using a step by step pouring technique, to produce a heterogeneous gel-based structure on the scale of human brain.MRI measurements have been performed with two different MRI scanners. Small homogenous phantoms (15ml vials) have been characterized in a 3T Bruker BioSpecscanner, equipped with a 72 mm 1H coil, to investigate the role of Agar and GdCl3. Raw data have been analysed with the Bruker Image sequence analysis tool. A "T1 saturation recovery" sequence has been applied to extractT1 values, whilst the "T2 vtr" sequence allowed T2 determination. The anthropomorphic phantoms have been characterized in a GE HDxt 1.5T in which data have been acquired using a 3D SSFP-MRF technique [2]. Acquisition schedule consisted of a variable flip angle train (length: 880 pulses) preceded by an adiabatic inversion pulse (TE/TR=0.5/8.5ms). Each frame has been sampled with 55 interleaves of a 3D spiral projection trajectory. At the end of each frame, a spoiler gradient has been applied to obtain a 4π dephasing along the z-direction. The electrical properties have been measured using a dielectric measurement system(SPEAG DAK) based on an open-ended coaxial probe technique. The probe has been positioned on the flat surface of cylindrical samples(diameter:10 cm, height: 5 cm) and connected to a network analyzer (VNA). By means of a commercial software, the conversion between the complex S-parameters, measured by the VNA, and the dielectric values of the samples (i.e. electrical conductivity and dielectric permittivity) has been performed.

Results

In Fig. 1an example of T1 and T2 parametric map is reported, together with the role of Agar and GdCl3 concentration. Further measurements confirmed that the variation of these components allows to independently tune T1 and T2 values within the reported range. In Fig. 2 the T1 and T2 parametric map for the brain-like phantom is reported. No presence of diffusion phenomena has been observed after several months and, as it is possible to note from the figure, the structure of the heterogeneous phantom clearly traces the complex geometry of WM and GM compartments. In Fig. 3, an example of dielectric measurements is shown. In this case, electrical conductivity and relative permittivity have been modulated by means of NaCl and PVP.

Discussion & Outlook

Preparation of phantoms using gel-based tissue mimicking materials have been realized by means of different polysaccharide and additives for tuning their relaxation and electrical properties. With the aid of 3D printed moulds, it was possible to realize heterogeneous gel-based anthropomorphic structures. MRI measurements, performed on this phantom, showed distinct relaxation parameters comparable to those reported in literature for WM and GM demonstrating the feasibility of this approach. .Further preparations will be investigated in order to tune all the parameters in a single anthropomorphic phantom, taking into account the characterization data provided by small samples.

Acknowledgements

The results here presented have been developed in the framework of the 18HLT05 QUIERO Project. This project has received funding from the EMPIR programme co-financed by the Participating States and from the European Union’s Horizon 2020 research and innovation programme.

References

1]K. Yoshimura, Development of a Tissue-Equivalent MRI Phantom Using Carrageenan Gel, Magnetic Resonance in Medicine 50:1011–1017,(2003)

[2] Pedro A. Gómez [Rapid three-dimensional multiparametric MRI with quantitative transient-state imaging, Scientific Reports volume 10, Article number: 13769 (2020).

[3] Debra F. McGivney, SVD Compression for Magnetic Resonance Fingerprinting in the Time Domain, IEEE Trans Med Imaging. Dec; 33(12): 2311–2322 (2014)

[4]M. Weigel, Extended Phase Graphs: Dephasing, RF Pulses,and Echoes - Pure and Simple, Journal Of Magnetic Resonance Imaging 41:266–295 (2015)

Figures

Figure 1: Parametric images of small vials containing Gellan Gum solution (1.5% in weight) with different amount of Agar and GdCl3

Figure 2: Parametric images of heterogeneous brain-like structured phantom measured at 1.5 T.For the reconstruction,K-space data was compressed as proposed by McGivney et al. [3] and transformed to the image domain by using a Non-Uniform Fourier Transform.M0/T1/T2 maps of the object were computed using dot-product pattern matching with a dictionary of signal evolution obtained using EPG formalism [4].

Figure 3:Electrical properties of two phantoms simulating WM (on the left) and GM (on the right).The conductivity and relative permittivity values around 128 MHz are close to those reported in the IT'IS Foundation database for WM and GM.

Proc. Intl. Soc. Mag. Reson. Med. 29 (2021)
3347