Servizio Comunitario di Informazione in materia di Ricerca e Sviluppo - CORDIS


EXPAT Sintesi della relazione

Project ID: 309674
Finanziato nell'ambito di: FP7-IDEAS-ERC
Paese: France

Mid-Term Report Summary - EXPAT (Exploring the human brain using magnetic resonance imaging and parallel transmission at ultra-high field)

The EXPAT project consists of solving important RF pulse design problems in parallel transmission in MRI at 7 T and at 11.7 T using new and unconventional approaches. After developing the methodology to measure the subject-based RF field maps quickly and accurately (few minutes), constrained-optimization algorithms have been established to return RF pulses able to greatly mitigate the RF inhomogeneity problem at ultra-high field both in 3D and 2D, under explicit hardware and safety constraints, and in very reasonable computation times. This furthermore includes the optimization of the k-space trajectory jointly with the one of the RF pulses. Special attention has been paid to deal with the temperature aspects instead of the ubiquitously used, but less relevant, specific absorption rate. It has thereby been shown that it is now possible to handle the temperature directly in the RF pulse design, once a bio-heat model is thoroughly validated, and that gain in RF pulse performance as a result is within reach. The key concepts explored in the project have been validated experimentally at 7T via the successful implementation of key MRI sequences such as the GRE, MPRAGE, spin-echo and SPACE sequences. Moreover, by nature the last two sequences called for new approaches to solve the pulse design problem given the refocusing aspect of the pulses, i.e. where the state of the magnetization prior to the pulse presumably is unknown. The problems were solved by using tools and techniques inspired and adapted from the quantum computing literature. In vivo experiments in the laboratory are now performed with full parallel transmission capabilities where the SAR constraints are now evaluated by taking into account the phase of the RF waveforms and of the electric fields, and where dedicated hardware monitors the correct behavior of the sequence. The evaluation of the SAR thus is much less conservative and boosts the RF pulse and MR sequence performance. This however required the experimental validation of the electromagnetic simulations performed to calculate the SAR.

Reported by