Image analysis – Applications – Biomedical applications
Reexamination Certificate
2000-05-08
2004-02-24
Rogers, Scott (Department: 2626)
Image analysis
Applications
Biomedical applications
C382S254000, C382S275000, C382S107000, C600S410000, C600S425000
Reexamination Certificate
active
06697507
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to a method of and apparat us for reducing ghost artifact in image data. The invention has particular application in those Magnetic Resonance Imaging (MRI) techniques which are prone to coherent ghost artifact.
BACKGROUND OF THE INVENTION
In conventional Two Dimensional Fourier Transform (2DFT) MRI techniques, one “line”of data in sample space (also known as k-space) is acquired for each application of a radio frequency (rf) pulse. The resulting echo signals are sampled at various points during the evolution of the magnetic gradients in the “read direction”to obtain each line. Incremental increases in the magnetic gradient are applied in the “phase encode direction” in order to read successive lines through the sample. A two dimensional Fourier transformation is applied to the sampled data to obtain the image data.
Under certain circumstances there may be a periodic variation in the sampled data, leading to the appearance of a ghost artifact in the image data. The periodic variation may be in any of the parameters of the sampled data, for example the amplitude or the time of sampling. Any phenomenon that causes a periodic variation in the sampled data may give rise to a ghost artifact. Typical examples are periodic movements in the sample, such as may occur when imaging the heart or lungs, or periodic variations in the operating conditions of the imaging apparatus, which may be due to internal or external influences. In the case of the periodic movement of the heart, a coherent ghost artifact will appear if the heart rate is related to the sampling rate.
In certain MRI techniques, there may be a periodic variation in the sampled data due to the manner in which the data are acquired. Examples of such techniques are Echo Planar Imaging (EPI), Segmented EPI, and Echo Volumar Imaging (EVI).
Echo Planar Imaging differs from 2DFT in that an entire image is acquired from a single rf excitation pulse. To acquire the image, increments of magnetic gradient in the phase encode direction are applied, whilst switching the magnetic field in the read direction between positive and negative. The echo signals are sampled at various points during the evolution of the magnetic gradients to obtain sampled echo data. A two dimensional Fourier transformation is then applied to the sampled data to obtain the image data.
EPI intrinsically involves a periodic variation in the sampled data as a result of the alternate switching of the magnetic field in the read direction between positive and negative. This switching of the magnetic field results in alternate lines in the sampled data requiring time reversal prior to Fourier transformation. Any misalignment between the time reversed lines will result in a coherent ghost appearing in the image, overlapping with the real image.
Segmented EPI works by applying a number of excitation pulses, and acquiring part of the data, known as a segment, following each pulse. Changes in the operating conditions between the different segments, together with the switching of the magnetic field within each segment, may give rise to periodic variations in the sampled data with a periodicity of twice the number of segments. This will give rise to multiple ghost artifacts in the image domain.
For the purpose of simplicity much of the description henceforth will be directed towards EPI, which is highly prone to a single coherent ghost artifact at +Np/2 in resulting images, where Np is the number of points in the phase encode direction. However, the present invention is applicable to any imaging technique in which a coherent or pseudo-coherent ghost artifact occurs, regardless of the origin of the ghost artifact.
Various techniques have been used to reduce or cancel ghost artifact in EPI. Perhaps the simplest method is a manual technique, in which relative time shifts in the sampling points between time reversed lines are adjusted manually, until the ghost disappears. This has the disadvantage that it requires the intervention of a skilled operator. It also requires that the system has real time data acquisition, reconstruction and image display facilities.
An alternative technique for ghost elimination has been proposed, in which a calibrating scan is first done to determine the time shift between the time reversed lines of data. This time shift is then used to correct the errors in subsequent imaging scans. This has the disadvantage of being complex and of increasing the time required for imaging. Furthermore, there may be changes in the time shift between the calibrating scan and the imaging scan, leading to a reduction in the efficiency of the ghost cancelling.
Methods employing additional, redundant, reference scan lines within the imaging sequence have been proposed (Jesmanowicz et al SMRM abstract, 1993, p 1239 and EP 0644 437A, Philips Electronics NV, 1995). As these require extra data sampling these methods prolong the data acquisition and hence the imaging times, which is critical in high speed imaging techniques.
A post processing method of reducing ghost artifact has been proposed by Bruder et al in Magnetic Resonance in Medicine, vol. 23, pp 311-323, (1992). Bruder et al proposed automatic adjusting of the data in the image domain until the ghost substantially disappears. This technique relies on the genuine image and the ghost image being spatially separated, that is, not overlapping in the imaging field. The technique therefore cannot be used if there is a genuine image across the whole of the imaging field. In EPI this is a significant disadvantage, since the size of the imaging field is limited by the ability of the imaging apparatus to generate large and rapidly varying magnetic fields. Consequently, it is highly undesirable to reduce the area allocated to the genuine image even further, by dedicating an area for use in ghost artifact reduction.
A similar technique to that of Bruder et al has been proposed by Buonocore et al in Magnetic Resonance in Medicine, vol. 38, pp 89-100, (1997). The technique of Buonocore et al also relies on analysing the ghost image in isolation and therefore can only be used when there is an area of no overlap between the genuine image and the ghost image. This technique therefore suffers from the same disadvantages as that of Bruder et al.
If employed with imaging techniques such as Segmented EPI, where multiple ghost artifacts occur, the techniques of Bruder et al and Buonocore et al would require the collection of correspondingly more lines dedicated to ghost removal and devoid of genuine image rendering them even less practical.
SUMMARY OF THE INVENTION
It is an aim of the present invention to provide a method of and apparatus for reducing ghost artifact that overcomes or reduces the problems of the prior art.
In one aspect, the invention provides an apparatus for reducing ghost artifact in image data, the apparatus being for use with an imaging apparatus which produces sampled raw image data that may experience a periodic variation giving rise to said ghost artifact, the apparatus comprising image reconstruction means for converting the sampled raw image data into the image data to reconstruct an image, characterized by means for analysing the sampled raw image data alone to determine a correction to reduce the ghost artifact, without requiring additional sampled data beyond that required by said image reconstruction means.
Analysing the sampled data (rather than the image data) to determine the correction can afford the advantage that a correction to reduce ghost artifact may be calculated without requiring extra data to be acquired. At the same time, the present invention can be readily implemented with existing magnetic resonance imaging techniques as it does not necessitate any changes in data sampling procedures. The need for operator interaction in reducing ghost artifact may be avoided. Analysing the sampled data, rather than adjusting the operating conditions, gives the advantage that ghost artifact may be reduced regardless of the origin of the artif
BTG International Limited
Nixon & Vanderhye PC
Rogers Scott
LandOfFree
Ghost artifact reduction does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Ghost artifact reduction, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Ghost artifact reduction will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3277222