Personal tools
You are here: Home Research site Jobs Minimizing geometric distortion in Diffusion Weighted MRI with 2D RF pulses

Minimizing geometric distortion in Diffusion Weighted MRI with 2D RF pulses

Master project


Diffusion Weighted (DW) MRI, a pair of a dephasing and rephasing gradients is used to probe for diffusion of water in human tissue. Spins that are displaced during the interval between the two gradients will not be properly rephased. This will cause signal loss reflecting the local tissue diffusion characteristics. This can be used to find tumors. Tumors have a high cell density which restricts diffusion motion of water. See figure 1a. Although diffusion weighted images show good contrast between tumor and surrounding tissue, tumor delineation is hampered by geometric distortions. These disturbances can result in a shift in geometric position, shearing, compression or stretching of the organ of interest. These geometric distortions find their origin in the single shot echo planar (ssEPI) sequence commonly used for DW-MRI. Althoughh EPI is fast it has the major disadvantage that it shows large geometric irregularities that arise from small distortions in the static magnetic field. An example is shown in figure 1c where the trachea induces a magnetic field distortion which appears to shift the spine from its real position. An additional, undesired effect of these magnetic field distortions around air cavities is intra slice signal dephasing which results in a poor contrast-to-noise in the vicinity of air cavities. These two phenomena are the main challenges for DW-MRI.

Figure 1a: DW MRI shows the position and extent of an oropharyngeal tumor (arrow). b:) MRI image of neck without (b) and with (c) geometric distortion. The trachea and spine on the undistorted image are overlaid on the distorted (c) image.

Goal of the project

The goal of this project is to address the geometric distortion and signal dephasing issues of DW-MRI.
This work is part of a larger project to create a DW MRI technique with a high geometric fidelity and imaging quality such that it can be used for tumour delineation for radiotherapy treatment. The application area will be head and neck cancers as standard DW-MRI is very challenging in the head and neck due to the presence of many air cavities which distort the B0 magnetic field.


To reduce the geometric distortion of ssEPI sequences, the echo train length has to be shortened. A promising way to achieve this is to use 2D spatially selective excitation pulses (2D-RF). Using such a specially tailored RF pulse, the field-of-view in the phase encoding direction can be reduced, e.g. by exciting only the tumour region. An example is shown in figure 2 where a specially designed RF pulse was employed to excite a X-shaped pattern in a phantom. In this way a much lower echo train length can be employed. In addition, the signal dephasing can also be repaired with 2D-RF in combination with the availability of several RF transmit channel.

In the Utrecht Medical Centre, a new 3 Tesla MR system (Philips Healthcare) has been recently installed. This system is capable of transmitting with two individual RF transmit channels which would ideal for the use of multi-channel 2D-RF pulses. The goal of this master project is to use the extra RF transmit capabilities of this system to implement 2D RF pulses for (1) echo train shortening and (2) reduction of signal dephasing in DW-MRI.

Figure 2a: An example of an 2D-RF pulse. Note how different the waveform is from standard sinc RF pulses. b:) These 2D RF pulses can achieve excitation of arbitrary shaped such as here the letter “X” in a cylinder.

The work of the student will consists of (1) a numerical part: RF pulse simulations using an electromagnetic model of the scanner and patient and (2) an experimental part consisting of experimental testing on the scanner and analysis and evaluation of the acquired images (Matlab). The focus will be on the second experimental part. The student will be supervised by MR physicists dr. Nico van den Berg and dr. Hans Hoogduin and work closely with PhD student Alessandro Sbrizzi who develops RF pulse design algorithms. The work will be technically supported by Philips Healthcare.

Research Environment

In department of Radiotherapy and Radiology of the University Medical Centre Utrecht a large infrastructure with respect to MRI is present. Human MRI scanners at various field strength (1 to 7 Tesla) are available and a large knowledge base is present with respect to MRI physics and hardware.
The new Philips 3T dual transmit channel MR system is one of the most advanced clinical MR system available in the world. As such, Philips Healthcare has a clear interest in this project and will be involved in the implementation. The student will be embedded in team of other PhD students and Master students working on Radiotherapy related MRI topics. The candidate will acquire in the project a broad general knowledge about MR physics. The candidate should have enthusiasm for MR physics and preferably some prior basic MR knowledge.


For further information contact dr. Nico van den Berg (supervisor) 

dr. Nico van den Berg
MR physicist 
+31 88 7553136