Transforming Magnetic Resonance Spectroscopy into a Clinical Tool

Project Information

TRANSACT

Grant agreement ID: 316679

Closed project

Start date 1 March 2013

End date 28 February 2017

Funded under

• Specific programme "People" implementing the Seventh Framework Programme of the European Community for research, technological development and demonstration activities (2007 to 2013)

Total cost € 3 445 287,12

EU contribution € 3 445 287,12

Coordinated by

KATHOLIEKE UNIVERSITEIT LEUVEN
Belgium

Final Report Summary - TRANSACT (Transforming Magnetic Resonance Spectroscopy into a Clinical Tool)

TRANSACT Marie Curie Initial Training Network No. 316679
Coordinator: Prof. Sabine Van Huffel, KU Leuven, Dept. Electrical Engineering ESAT/STADIUS, Belgium, [e-mail]
Webpage: www.transact-itn.eu

Summary of project objectives

TRANSACT has been an EU-funded FP7-PEOPLE Marie Curie Initial Training Network (ITN) running from March 1, 2013 till February 28, 2017. TRANSACT linked 10 academic and 4 industrial partners across Europe with complementary expertise in basic science, clinical research and information technology. This network trained young researchers in the interdisciplinary field of Magnetic Resonance Spectroscopy. Their research had the aim of Transforming Magnetic Resonance Spectroscopy into a Clinical Tool. At the start of the project, it has been recognised that Magnetic Resonance Spectroscopy and Spectroscopic Imaging (MRS(I)) are unique, indispensable techniques for non-invasive metabolic imaging. Important areas where MRS(I) can make a difference are oncology and neurology, where metabolic changes due to, e.g. tumour formation, can be detected earlier and more sensitively than with morphological imaging modalities alone. However, it was also known that, despite its huge proven potential, MRS(I) was not yet a routine clinical tool operated solely by clinicians, due to lack of standardisation and quality control. The TRANSACT network was set to answer issues such as reliable automation, standardisation and quality control of MRS(I), and to train young scientists in this interdisciplinary field. Specific main objectives have been:
O1. Development of criteria and standards for data quality assessment in MRS and MRSI.
O2. Standardisation, automation, validation and robustness of metabolite quantitation methods.
O3. Exploiting MRS information in multi-modal image quantitation.
O4. Tissue or disease classification based on MRS(I) and multi-modal MR data analysis.
O5. Validation on in vivo data in applications from clinical and preclinical research.

Work performed within the project

Main scientific results achieved
- improved MRS(I) quality control:
- a range of methods for quality control of MR spectra and artefact detection;
- a range of methods for prognosis of metabolite quantitation reliability.
- improved acquisition techniques:
- preclinical MR scanners: improvement and implementation of 1H-MRS, 31P-MRS and 31P magnetisation transfer for 9.4T small animal MR scanner;
- clinical MR scanners: water unsuppressed 3D prostate MRSI without an endorectal coil; optimisation of MEGA-PRESS sequence for quantifying glutathione and GABA in the brain.
- novel post-processing techniques:
- classification techniques based on MRS(I) or multi-parametric MR data;
- modelling and change detection in white matter fiber bundles using DTI data;
- patch-based up-sampling of metabolite images using MRI and brain tissue segmentation.
- new insights in several disease domains:
- identification of brain metabolic changes in hepatic encephalopathy and chronic liver disease;
- prediction of tumour recurrence for glioblastoma patients;
- prostate tumor scoring by fractional tumor content;
- MRI/MRSI-based characterisation of multiple sclerosis clinical forms and lesions;
- identification of significant metabolic changes (e.g. glutamate and glutamine increase) during brain activation using a paradigm;
- neurodegeneration in amyotrophic lateral sclerosis visualised in mice hindbrain; significant differences in absolute concentration of neurotransmitters appearing in hindbrain.
- MRSI and MRI post-processing software:
- new releases of jMRUI (jmrui.eu) documentation/website and plugins: SpectrIm: combined analysis of MR Spectroscopy and Imaging; SpectraClassifier and INTERPRET decision support system;
- MSmetrix (msmetrix.icometrix.com): lesion and tissue segmentation method.

Training activities

TRANSACT organised its training programme as individual research projects, secondments, scientific network-wide workshops (also open to external participants), transferable skills courses, and a Creativity Workshop. The largest training event open to external participants has been the symposium entitled Quality Issues in Clinical MR Spectroscopy, 16-18/06/2015, organised by University of Bern (www.kas.unibe.ch/transact/). This event featured 22 invited speakers, an educative poster session and a contributed poster session, and welcomed 48 external participants. Slides and audio recordings are available for future educational purposes.

Dissemination

- 35 peer-reviewed journal papers published during the project duration.
- contributions at numerous international conferences (> 60 talks and posters).
- 5 scientific workshops open to external participants.
- 3 jMRUI training courses organised in the context of TRANSACT.
- an advanced teaching session devoted to TRANSACT topics, which has taken place at the 33th Annual Scientific Meeting of the European Society for Magnetic Resonance in Medicine and Biology (ESMRMB 2016, September 29 - October 1, 2016, Vienna/AT).
- TRANSACT website (www.transact-itn.eu) maintained and frequently updated.
- videos on project topics: www.transact-itn.eu/index.php/events/outreach-activities

Final results and their potential impact and use

TRANSACT has contributed significantly to the further development of jMRUI (jmrui.eu) a software package for advanced time-domain analysis of magnetic resonance spectroscopy (MRS) and spectroscopic imaging (MRSI) data, free for non-commercial use.