Hepatobiliary/GI Application Tips: 3D Magnetic Resonance Cholangiopancreatography (3D-MRCP) With Prospective Acquisition Correction (PACE)

3D Magnetic Resonance Cholangiopancreatography (3D-MRCP) with Prospective Acquisition Correction (PACE)
(Sample Case)

In a variety of MRI applications, motion can adversely affect image quality. For some patients, even the shortest breath-hold duration might be too demanding; or, patients may be unable to follow breathing commands due to impairments in mental status or hearing impairment. Since patient motion cannot be controlled sufficiently in all cases, image acquisition strategies have been developed to maintain image quality despite motion. Correction of motion effects at the postprocessing stage is one approach, however, it is better to acquire a good data set in the first place. The techniques for coping with motion use respiratory triggered image acquisition and are collectively termed PACE (Prospective Acquisition Correction). Respiratory triggering reduces motion artifacts by synchronizing anatomical data acquisition with the respiratory cycle, thus, allowing for imaging while the patient is breathing freely. Corresponding to the spatial dimensions of the dataset used for calculating the adjustment, PACE techniques are termed 1D-PACE, 2D-PACE, and 3D-PACE. The first two are used mainly to deal with breathing motion, while the third one is applied for motion adjustment in neurological studies.

The fastest method of detecting motion is the 1D PACE technique, also known as a “navigator” technique. Navigator triggering uses a navigator (i.e. MR signals) to monitor the respiratory motion. This technique is different than the older respiratory triggering techniques, which used a respiratory belt to retrieve the patient's breathing pattern. Navigator triggering relies on real time tracking of diaphragmatic motion to ensure consistent positioning of the imaging slice. A typical application for the use of this technique is during an MRCP examination.

Current MRCP techniques that can produce ERCP-like images employ heavily T2-weighted sequences that depict the fluid-containing biliary tree and pancreatic duct. The heavy T2 weighting is achieved by very long echo time (TE), with 2D turbo spin echo (TSE) sequences being the most commonly performed techniques. The 3D technique has potential advantages over 2D imaging, due to its capacity to provide thinner sections, no gaps, and a higher signal-to-noise ratio (SNR). Postprocessing manipulation of the images including multiplanar reconstructions (MPR), maximum intensity projection (MIP) or volume rendering can be performed. Unlimited projectional views in any arbitrary plane can be obtained. With the recent technical advances in MR imaging, such as faster gradients and especially the application of parallel acquisition technique (PAT), marked reduction in acquisition time can be achieved, which in turn makes it possible to obtain a higher spatial resolution with preserved SNR within a reasonable amount of scan time.

The navigator-triggered PACE sequence can be split into two parts. The first part is short “learning phase” where the breathing of the patient is analyzed by a trigger algorithm to learn about the patient's breathing pattern and the central position of an “acceptance window” is calculated automatically. The second part is the imaging phase during which the gated acquisition begins and image slices are acquired only when the diaphragm position falls within the acceptance window. Here, the slice positions of different scans can also be aligned based on information about the position of the diaphragm. The information about the diaphragm position allows the operator to monitor the breathing pattern of the patient online. Furthermore, acquisition of slices during different breath-holds can be aligned in order to compensate for imperfect reproducibility of the breath-hold position. In this way, gaps between slices or overlaps are avoided.

To plan a navigator triggered measurements proceed as follows:

The 3D MRCP sequence we perform is a heavily T2-weighted 3D TSE restore sequence. Due to the extremely long TE (TR(msec)/TE (msec)/flip angle = 1300/680/180°), fluid is the main signal source. A high turbo factor (>100) is used so that only 2 echo trains are needed to fill each partition. A -90° RF pulse (so called ‘restore pulse') at the end of the echo train is applied to flip the transverse magnetization back into longitudinal direction, which shortens the spin relaxation time. Zero filling and partial-Fourier of 6/8 in slice selection direction is performed to achieve an interpolated slice thickness of 1 mm. A stack of slices (52-60 slices) is acquired for diagnostic purpose. Using PAT with acceleration factor 2, the acquisition time is further reduced. The typical acquisition time is 1-2 minutes with the navigator triggered acquisition techniques and depends on the patient's respiratory pattern (length of each respiratory cycle).