Invention:
The present invention is a new MRI acquisition that does not suffer from the noise-bandwidth trade-off commonly seen in current MRI measurement devices. This trade-off measuring seen in large range of inhomogeneity values will suffer from high levels of noise resulting in unreliable estimates of measurement. Accurate results are only guaranteed over a small range of inhomogeneity values. The method collects three Gradient-Recalled Echo (GRE) images at three different echo times. That is, one additional GRE acquisition (i.e. less than half a second longer), in comparison to traditional field map estimation methods. The associated estimation algorithm (processing software) is optimally tuned to this engineered acquisition and returns field map estimates that have a pre-determined accuracy over this arbitrary bandwidth. These acquisition and processing parameters are jointly optimized offline based on the tissue/material of interest.
Background:
Rapid MRI acquisitions with long read out times suffer from artifacts due to inherent magnetic field imperfections, or inhomogeneities. These inhomogeneities are a function of the hardware (magnets), electronics (switching gradient of the MR) and, more importantly, the underlying physiology (air/bone/fluid interfaces, chemical composition, etc). The resulting artifacts, such as signal loss and geometric distortion worsen with higher field MRI machines.
Application:
- Can be incorporated into the acquisition methods of most commercial MRI scanners
- The product could one day be developed on an existing commercial MRI machine and sold as a software upgrade to institutions owning the MRI machine and/or having an agreement with the manufacturer
Advantages:
- Proposed method does not suffer from the noise-bandwidth trade-off
- Can estimate very accurate field inhomogeneity maps over an arbitrary range of values
- Method can be optimized to the imaged material or tissue of interest
- Method does NOT require a phase unwrapping routine
Status: issued U.S. Patent #9,733,329
