ERETIC based in vivo H MRSI quantification

Introduction Quantification of metabolite concentrations from spectroscopic images has been performed by a variety of approaches: Absolute quantification based on internal water referencing [1] is the current method of choice. It accounts for B1 variations across the field of view but imposes problems when assessing diseases that lead to alteration of tissue water concentration [2] and relaxation behavior. Methods using external references for absolute quantification [3] often suffer from susceptibility problems and resonances of interest being obscured by the reference signal. Just like the phantom replacement methods, whose main drawback is the need for an additional carefully calibrated scan, methods using external references have never been established for routine clinical diagnostics, although they were successfully used for many scientific studies. The ERETIC (Electric REfrence To acess In vivo Concentrations) method was introduced by Barantin et al. [4] and since then has been proven to be a reliable and accurate method for the assessment of metabolite concentrations in vitro [4, 5] and in vivo in single voxel MRS [4, 6]. The advantage of this approach is its insensitivity to changes in coil loading conditions, receiver gain settings and data processing. Also, the quantification can be performed simultaneously with the acquisition of the spectra. Unlike internal reference standards, the ERETIC method is also applicable in lesions or in the presence of pathological disorders and can be combined with localization without scan time prolongation. In this work a modified ERETIC method has been developed to enable quantification of metabolite concentrations in spectroscopic images (MRSI). Together with the acquisition of a B1 map, an accurate quantification of absolute metabolite concentrations can be achieved. Materials and Methods All experiments were performed on a Philips Achieva 3T human MRI scanner (Philips Healthcare, Best, Netherlands) with a modified commercial transmit/receive birdcage proton coil [6]. Two-dimensional (2D) spin-echo MRSI localized by OVS [7] data were obtained from a 12 mm transversal slice (FOV: 170 mm x 190 mm, 18x20 voxels) that was planned on a T1weighted FFE image (see Fig. 1). ERETIC Layout and Setting: The ERETIC Setup for in vivo 1H MRS is described in detail and has been validated in [6]. In the present work the ERETIC pulse shape was set to Voigt; the frequency offset was chosen to be zero ppm, because no other metabolite signals are expected there. The amplitude, phase and decay time constant of the ERETIC signal were adjusted to obtain a peak-shape similar to the N-acetyl aspartate (NAA) peak. ERETIC and MRSI: In case of MRSI, a defined phase as well as a defined spatial response function and hence stable and comparable signal amplitudes can only be achieved, when the ERETIC signal is placed in a predefined voxel. For this purpose an additional variable phase was calculated for every phase encoding step and added to the ERETIC signal to imitate phase encoding. The additional phase offset was determined by calculating the phase encoding gradient zeroth moment in the chosen voxel. Quantification: Unsmoothed spectra including the ERETIC peak were fitted using the AMARES approach of jMRUI [8] after water removal by HLSVD filtering. In order to determine absolute concentrations of NAA, creatine (Cre) and choline (Cho) in vivo, the ERETIC signal was calibrated in vitro using the a phantom with physiological brain metabolite levels. The in vivo metabolite concentration is then calculated according to [6]