Electricity: measuring and testing – Particle precession resonance – Using a nuclear resonance spectrometer system
Reexamination Certificate
2000-11-30
2003-01-21
Arana, Louis (Department: 2862)
Electricity: measuring and testing
Particle precession resonance
Using a nuclear resonance spectrometer system
C324S307000
Reexamination Certificate
active
06509736
ABSTRACT:
BACKGROUND OF THE INVENTION
This invention relates to magnetic resonance imaging (MRI) systems and more particularly to such systems used for acquiring data and reconstructing data for two- or three-dimensional images.
Magnetic Resonance imaging (MRI) involves acquiring data in the spatial frequency domain referred to as k-space, and transforming the data into the spatial domain prior to viewing. The acquired k-space data samples have both magnitude and phase components. The Fourier transform of the k-space data is the MRI image.
Cartesian sampling in k-space followed by inverse discrete Fourier transform (IDFT) represents a commonly-used magnetic resonance imaging scheme. While the IDFT reconstruction is generally realized using the well-known and highly efficient fast Fourier transform (FFT) algorithm, reconstruction latency can still be significant. Reconstruction latency refers to the interval from the time of data acquisition to the time of actual visualization of the corresponding image.
With a modern MR imaging system run in gated time-resolved, interleaved multi-slice or volumetric imaging mode, depending on the data set size the latency may be on the order of tens of seconds, which may seriously hamper the use of the system for real-time or concurrent monitoring and diagnosis. The conventional practice of separating the FFT-reconstruction from the acquisition of a complete set of k-space samples is a main contributor to the latency. Further, the problem is generally aggravated when spatiotemporal coverage/resolution increases, because the time required to complete an FFT increases as the number of data points increase.
With the MR system run in fluoroscopy mode, the reconstruction latency degrades the system's real-time performance and leads to low image frame rate. In this case, computation redundancy may be another major contributor to the latency. To achieve a smoother depiction of imaged dynamics for example, the known technique of sliding-window reconstruction attempts to increase the number of reconstructed images through sharing raw data between reconstructed images. When this technique is applied in Cartesian-sampling based MR fluoroscopy, data acquisition constantly loops through the phase encodes, resulting in a fully refreshed k-space data frame every T
traverse
seconds (T
traverse
=time required for a complete k-space traversing). Image reconstruction, on the other hand, repeatedly applies FFT to a sliding window of the most recent full set of phase encodes, producing an image every T
compute
seconds (T
compute
=time required for the FFT computation). While T
traverse
determines the temporal resolution of the fluoroscopic images, T
compute
determines the upper limit of the rate at which one can slide the reconstruction window and hence the frame-rate of the fluoroscopic images. The fluoroscopy's real-time performance will thus be degraded if the FFT's are carried out slowly, because not only will the latency be significant, but also will the frame-rate be low. What is needed is further efficiency in reconstructing images to further cut down acquisition-to-visualization latency. Further needed is the ability to share intermediate results and to most efficiently eliminate redundant computations.
BRIEF SUMMARY OF THE INVENTION
A computationally efficient and latency-minimized method of data acquisition and reconstruction for use with a Magnetic Resonance Imaging (MRI) system comprises traversing acquired k-space data in a plurality of segments and computing sub-images for each of the segments. Thereafter, the sub-images are incrementally summed to form intermediate images for use in monitoring and diagnosis in said MRI system.
Another computationally efficient method of data acquisition and reconstruction comprises equivalently evaluating a Fourier transform of acquired k-space data substantially immediately after data acquisition and incrementally reconstructing an image with the equivalently evaluated Fourier transform for the respective view.
REFERENCES:
patent: 4651095 (1987-03-01), Eastwood et al.
patent: 4830012 (1989-05-01), Riederer
patent: 4958282 (1990-09-01), Barjhoux
patent: 5164671 (1992-11-01), Goldberg
patent: 5275164 (1994-01-01), Maeda et al.
patent: 6097977 (2000-08-01), Collick et al.
patent: 6278273 (2001-08-01), Riederer et al.
patent: 6320378 (2001-11-01), Maier et al.
Arana Louis
Breedlove Jill M.
General Electric Company
Testa Jean K.
LandOfFree
Method for MRI data acquisition and reconstruction with... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for MRI data acquisition and reconstruction with..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for MRI data acquisition and reconstruction with... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3060294