Skip to main content

Rapid Adaptive Nonlinear Gradient Encoding for Magnetic Resonance Imaging

Final Report Summary - RANGEMRI (Rapid Adaptive Nonlinear Gradient Encoding for Magnetic Resonance Imaging)

Based on a principle of “Rapid Adaptive Nonlinear Gradient Encoding” (RANGE), the goal of the project was to develop novel spatial encoding methodology. This included the development of theoretical foundations, new hardware and prospective applications for human medical imaging especially including localized imaging.

The first task of the project aimed at designing and implementing the technological backbone for the RANGE principle –the novel highly-integrated actively-shielded matrix gradient coil. During the process of the coil design, several metrics have been developed to access the performance of different matrix coils, which helped us to select a suitable number of coil elements of the matrix coil. Later on, the novel design method of single coil element has been developed based on the requirement of local non-linear encoding and eddy current reduction. Using the designs of coil elements, the 84-channel prototype of the matrix gradient coil has been implemented. These results have been presented in numerous conference abstracts, several peer-reviewed publications and three published European/US patent applications.

In order to operate the matrix coil using existing 12-channel commercial gradient amplifiers at our institute, an optimization approach has been implemented to obtain groups of coil elements with only a limited restriction of flexibility. Moreover, a 12-channel driver system integrated with multi-channel switches for the matrix coil has been developed. An RANGE MR console has been created to control the novel hardware in combination with the original clinical scanner. Imaging results have demonstrated that the integrated system including the matrix coil is able to generate both local nonlinear and global linear fields and provide an unprecedented and ultimate flexibility in generating rapidly switched localised encoding fields for human MR imaging. These results have been presented in numerous conference abstracts, several peer-reviewed publications and one published European/US patent application.

Although the majority of the results in the first task are targeted at the successful implementation of the matrix coil, two of such results can readily be employed beyond the original scope of the project. One such example is the performance measures developed within the project for the analysis of the novel multi-channel matrix coils, which can be applied to optimisation of the traditional linear gradient systems with only three channels. As a result of an initial study a concept of a truly symmetric gradient coil has been created, leading to a patent application. Another example is the development of the RANGE console. The RANGE MR console was developed as an extension of the of an open-source MR simulation toolkit JEMRIS and was originally intended to be exclusively used within the project. It turned out however, that the console is very useful for a rapid prototyping of MR pulse sequences and many of the members of the Freiburg MR group are really interested to start using it. Furthermore, the RANGE console has also become a hardware-independent cross-platform MR prototyping tool.

The application-oriented research on localized imaging was performed using phase pre-modulation and excitation and encoding in curved coordinates. Some clinically-relevant in-vivo applications using the developed methodology have been performed to demonstrate the power of localized imaging. For example, the result of a zoomed cardiac MRI shows that the scan time can be reduced by a speed of up to 60%. Besides the imaging encoding aspects, the new matrix coil can also be used for shimming. A preliminary evaluation shows that the coil can outperform 5th-order spherical harmonic (SH) shimming in the human brain. These results have been presented numerous conference abstracts and several peer-reviewed publications.