Grzegorz Chadzynski

Alumni of the Department High-Field Magnetic Resonance

Main Focus

Proton magnetic resonance spectroscopy (1H MRS) utilizes the nuclear magnetic resonance phenomenon to identify the chemical composition of the tissue under investigation. 1H MRS can be realized with two approaches: single voxel spectroscopy (SVS) and chemical shift imaging (CSI). In the first method only one voxel is excited and therefore only one spectrum is obtained. The second approach consist in excitation of an extended region composed of multiple voxels, thus it provides spatial information about concentration of the metabolites.

In-vivo CSI of the human brain allows simultaneous detection of several metabolites, which play an important role in various biochemical processes. The most important are: N-acetyl-aspartate (associated with neurons and is used as a neuronal marker), creatine/phosphocreatine (energy buffer, provides the constant levels of ATP), choline (cell membrane marker), glutamine and glutamate (involved in neurotransmission), inositol (involved in intracellular signaling pathways), lactate (marker of ischemia and hypoxia); and macromolecules and lipids (can provide information about the cell membrane damage).

CSI at ultra-high magnetic fields (9.4 T) can benefit from increased both: signal to noise ratio and spectral resolution. The resonances of typical metabolites are better separated, so the overlap between the adjacent peaks is reduced. This allows more robust quantification of the measured metabolites. However, there are several limitations, such as: shorter T2 and longer T1 relaxation times of metabolites, higher SAR, lower available transmit B1 field, higher B0 and B1 inhomogeneities and increased spectral linewidth.

Characterizaton of the biochemical composition of the human brain using proton MRS/MRSI at 9.4 T.

Within this project we already demonstrated the feasibility of CSI at 9.4 T. The results obtained so far showed, that even the use of conventional acquisition strategies as stimulated echo acquisition mode (STEAM) enables quantification of not only the major metabolites (N-acetylaspartate – NAA, choline – Cho, creatine – Cr, or myo-inositol – Ins), but also those which are heavily overlapping (glutamine – Gln, glutamate – Glu, N- acetylaspartylglutamate – NAAG or taurine – Tau), or have small concentrations (glycine – Gly, glutathione – GSH or phosphorylethanolamine – PE) [1].

Further enhancements of the quantification procedure could be achieved with the assessment of macromolecule signals, which may be done with an inversion recovery technique (Fig. 1). The measured macromolecule resonances could be incorporated into the quantification procedure, which will not only provide more comprehensive information about the state of the investigated tissue, but may also improve its overall stability.

Assessement of human brain tumors (gliomas) by means of proton MRSI at 9.4 T

Improved signal-to-noise ratio (SNR) and spectral resolution offered by ultra-high field may be of high interest when using CSI for the assessment of human brain tumors. Recent studies performed at lower field strength (3T), showed that using CSI it is possible to verify the presence of 2-hydroxyglutarate (2HG) [2]. Detection of this particular metabolite is of high importance as it is associated with the mutations in isocitrate dehydrogenase (IDH) which occurs frequently in grade II and III gliomas. Moreover, it was demonstrated that 2HG has a potential to be a prognostic and diagnostic biomarker [2].

The main focus of this project is the assessment of the human brain gliomas. We were able to demonstrate that at 9.4 T it is possible to detect 2HG without using spectral editing techniques [3]. The results obtained so far were confirmed by a histopathological examination. Apart of the evaluation of 2HG, due to improved spectral resolution offered by ultra-high magnetic fields, it was also possible to detect the changes in other metabolites, such as Glu, Gln, Tau, Ins (Fig 2). This additional information may be used for further spectral analysis with the methods based on the analysis of metabolite coordinates, such as the Orthonormal Discriminant Vector method (ODV) [4, 5].

Development of fast and efficient acquisition techniques for the purpose of proton MRSI at 9.4 T

Increased chemical shift displacement error and signal losses due to faster T2 relaxation are two major limitations of proton CSI at ultra-high field strength. Previous studies demonstrated that CSI based on free induction decay acquisition (CSI-FID) is highly promising as it addressed the abovementioned issues [6]. Yet, spectra measured with this technique suffer from phase distortions, caused by truncation of the first FID points (due to the acquisition delay necessary for the RF pulse and for rephasing and phase encoding gradients), and from lipid contaminations (arising from subcutaneous fatty tissue). The first issue can be minimized by reconstructing the missing FID points by using an autoregressive model [7], while the second requires either additional outer volume suppression [6] or optimization of the point spread function (increasing the spatial resolution and spatial Hamming filter) [8].

Our initial results demonstrated that CSI-FID acquisition at high spatial resolution minimized the contaminations with lipid signals. Even in voxels closely located to the skull (Fig.2) lipid contamination does not negatively affect the main part of the spectrum (between 2 and 4 ppm) (Fig. 3). However, higher spatial resolution prolongs the total acquisition time, so it becomes necessary to optimize the time efficiency of the CSI-FID sequence. This can be achieved with shortening the water suppression scheme, as this is the most time consuming part of the sequence. Our preliminary results indicated that the modified water suppression (4 gaussian RF pulses, inter-pulse delay of 10 ms and the total duration of 61 ms) not only allows shorter total acquisition time (from the initial 21 to 15 minutes) but also is more robust against B1 field variations (Fig.4).

The efficiency of the CSI-FID sequence could be further improved by shortening the acquisition delay. This will reduce the influence of T2 relaxation and J-coupling effects and allow minimization of phase distortions, as the truncation of the FID signal will be smaller.

References:

1. Chadzynski GL, et al. Magn Reson Mater Phys DOI: 10.1007/s10334-014-0460-5 (2014).

2. Choi C, et al. Nat Med 18: 624-630 (2012).

3. Chadzynski GL, et al. Proceedings ISMRM 2013, pp. 961.

4. Hagberg G, et al. Magn Reson Med 34: 242-252 (1995).

5. Roser W, et al. Magn Reson Mater Phys 5: 179-183 (1997).

6. Henning A, et al. NMR Biomed 22: 683-696 (2009).

7. Chadzynski GL, et al. Proceedings ISMRM 2014, pp 2904.

8. Bogner W, et al, NMR Biomed 25: 873-882 (2012).

Curriculum Vitae

Education:

June 2012 – present

Max Planck Institute for Biological Cybernetics, Department of High-field Magnetic Resonance, Tuebingen, Germany

Post doc

April 2009 – June 2012

Eberhard Karls Universität Tübingen, Germany

Dr. sc. hum

'Development of CSI without water suppression for the purpose of clinical applications in the human brain'

October 2005 – September 2007 (4 semesters)

University of Silesia, Katowice, Poland, Faculty of Mathematics, Physics and Chemistry,

Department of Medical Physics

Master Degree in Experimental Physics, Major in Medical Physics

'Building and Testing of a Measurement Site for the ‘Mag-Force’ hyperthermia in-vitro experiments'

October 2002 – July 2005 (6 semesters)

University of Silesia, Katowice, Poland, Faculty of Mathematics, Physics and Chemistry,

Department of Medical Physics

Bachelor of Science Degree in Medical Physics

'Measurement of the Natural Dose of Ionizing Radiation with the Termoluminescent Dosimetry Method'

Professional Experience:

October 2011 - Present

Department of High-Field Magnetic Resonance,

Max Planck Institute for Biological Cybernetics,

Spemannstraße 41, 72076 Tübingen, Germany

Post-doctoral fellow (full time scientific work)

October 2011 – Present

Department of Biomedical Magnetic Resonance,

University Hospital Tübingen,

Hoppe-Seyler-Strasse 3, 72076 Tübingen, Germany

Post-doctoral fellow (full time scientific work)

September 2009 – September 2011

MR Research Group, Department of Diagnostic and Interventional Neuroradiology,

Universitätsklinikum Tübingen,

Hoppe-Seyler-Strasse 3, 72076 Tübingen, Germany

Scientific co- worker (full time scientific work),

position founded by Deutsche Forschung Gesellschaft (DFG KL 1073/7-1)

January 2009 – August 2009

MR Research Group, Department of Diagnostic and Interventional Neuroradiology,

Universitätsklinikum Tübingen,

Hoppe-Seyler-Strasse 3, 72076 Tübingen, Germany

Scientific co- worker (full time scientific work)

October 2007 – December 2008

Section for Experimental MRI of CNS, Department of Diagnostic and Interventional Neuroradiology,

Universitätsklinikum Tübingen,

Hoppe-Seyler-Strasse 3, 72076 Tübingen, Germany

Scientific co- worker (full time scientific work),

position founded by EU MARIE CURIE Host Fellowship,

Early Stage Training ‘Multimodal Biomedical Imaging and Clinical Application’ (MEST-CT-2004-007264)

July 2004 – July 2004

Silesian Center of Diagnostic Imaging HELIMED, Panewnicka 65, 40-760 Katowice, Poland

Practical Training (CT)

July 2004 – July 2004

Silesian Center of Diagnostic Imaging HELIMED, Panewnicka 65, 40-760 Katowice, Poland

Practical Training (MRI)

Grants:

May 2014 - May 2016:

"A novel approach to assess an extended biochemical profile of the human brain by the means of fast and efficient in-vivo proton Magnetic Resonance Spectroscopic Imaging  at 9.4 Tesla"

Fortüne junior grant (Das Forschungsprogramm der Tübinger Medizinischen Fakultät, project no. F 1358006.1)

2 years E13, full-time.

Go to Editor View