Surgery – Diagnostic testing – Detecting nuclear – electromagnetic – or ultrasonic radiation
Reexamination Certificate
2002-05-10
2003-12-02
Lateef, Marvin M. (Department: 3737)
Surgery
Diagnostic testing
Detecting nuclear, electromagnetic, or ultrasonic radiation
C600S407000, C324S306000, C324S307000
Reexamination Certificate
active
06658280
ABSTRACT:
TECHNICAL FIELD
This invention relates to magnetic resonance imaging.
BACKGROUND
Magnetic resonance (MR) imaging is a useful noninvasive method for imaging the internal components of a wide array of objects. Its noninvasive imaging of tissue in living subjects, especially humans, is highly valued in the medical field.
In its most basic form, MR imaging measures nuclear spin density throughout a sample. In this case, the image intensity is proportional to the number of observed nuclear spins. Practically, either the spin density, T
1
or T
2
of
1
H nuclei is measured. Although such images provide valuable information about an object, these parameters alone may not provide adequate image contrast. Many dissimilar materials have very similar spin densities, T
1
or T
2
and, therefore, such materials are indistinguishable or in other words they lack contrast.
A technique for enhancing contrast is described in “Artery and Vein Separation Using Susceptibility-Dependent Phase in Contrast-Enhanced MRA”, Wang et al., Journal of Magnetic Resonance Imaging, 12:661-670 (2000), the entire contents of which is incorporated herein by reference. In this technique, a magnitude and phase image are obtained using a gradient echo sequence. The magnitude image is operated upon using a mask computed from the phase image.
SUMMARY
In one aspect the invention features a method of MR imaging, including: obtaining a magnitude image, obtaining a phase image, computing a phase image mask using the phase image, applying the phase image mask to the magnitude image a number q times, and selecting q by computing CNR as a function of q.
In another aspect, the invention features a method of MR imaging including: computing CNR as a function of q, SNR, and &Dgr;&phgr;, and selecting q, SNR and &Dgr;&phgr; to produce a desired CNR.
In another aspect the invention features a method of MR imaging, including: obtaining a phase image and applying minimum intensity projection to the phase image.
In another aspect, the invention features a method of MR imaging, including: obtaining a first phase image by selecting a first echo time, obtaining a second phase image by selecting a second echo time, obtaining a predicted phase image by extrapolating the first phase image to the second echo time, and computing a corrected phase image by computing the difference between the predicted phase image and second phase image.
In another aspect the invention features a method of MR imaging including: obtaining a magnitude image, obtaining a phase image, computing a phase image mask using the phase image, applying the phase image mask to the magnitude image a number q times, selecting an acquisition resolution such that the resolution is higher than the size of a feature of interest, and wherein obtaining a magnitude and obtaining a phase image comprises reconstructing a magnitude and a phase image at a lower resolution than the resolution of the acquired data.
Embodiments of the method may include one or more of the following features.
The method may compute CNR as a function of SNR and &Dgr;&phgr;. The method may compute CNR using CNR(q)=SNR*(1−(1−|&Dgr;&phgr;|/&pgr;)
q
)/sqrt(1+q
2
/&pgr;
2
). &Dgr;&phgr; may be the phase difference between water and fat. &Dgr;&phgr; may be the phase difference between tissues with different amounts of iron.
The method may select q by using a function dependent on the size of a feature of interest. The method may select q such that sqrt(A) CNR(q) is greater than a value in the range from about 3 to about 5 where A is the area of the feature of interest measured in square pixels.
The method may include selecting a filter for reducing nonlocal contributions to the phase image, computing a local phase image by filtering the first phase image with the filter to reduce nonlocal contributions on the image, and wherein computing a phase image mask further includes using the local phase image.
Obtaining the magnitude and phase image may include selecting a first echo time that results in partial volume cancellation of features of interest. The method may include obtaining images of veins. The method may include obtaining images of microhemorrage.
The method may include obtaining a second phase image by selecting a second echo time that results in partial volume cancellation of features of interest, computing a corrected phase image by using the first and second phase images, and the computing the phase image mask includes using the corrected phase image.
The method may include selecting an acquisition resolution such that the resolution is higher than the size of a feature of interest, and acquiring magnitude and phase data at the acquisition resolution. Obtaining a magnitude and obtaining a phase image may include reconstructing a magnitude and a phase image at a lower resolution than the resolution of the acquired data. Reconstructing a magnitude and a phase image may include using the magnitude data and the phase data.
The method may include selecting a filter for reducing nonlocal contributions to the phase data, computing local phase data by filtering the phase data with said filter to reduce nonlocal contributions on the phase data, and reconstructing the magnitude and the phase image inlcudes using the magnitude data and the local phase data.
The method may include selecting an echo time to produce the selected SNR or &Dgr;&phgr;. The method may include selecting a set of echo times to produce the selected SNR or &Dgr;&phgr;. The method may compute CNR as CNR(q)=SNR*(1−(1−|&Dgr;&phgr;|/&pgr;)
q
)/sqrt(1+q
2
/&pgr;
2
). The desired CNR may be defined such that sqrt(A) CNR(q) is greater than a value in the range from about 3 to about 5 where A is the area of the feature of interest measured in square pixels. CNR may be dependent on the number of data acquisitions and the method may include selecting a number of data acquisitions to produce the desired CNR.
The method may include for a given total data acquisition time, selecting a number of data acquisitions. The method may compute CNR according to CNR(q)=SNR* sqrt(1/&agr;)*exp((1−&agr;)TE/T
2
*)*(1−(1−|&Dgr;&phgr;|/&pgr;)
q
)/sqrt(1+q
2
/&pgr;
2
) wherein &agr;=&Dgr;&phgr;/&pgr;. SNR and &Dgr;&phgr; may depend on partial volume cancellations.
The method may include selecting a filter for reducing nonlocal contributions to the corrected phase image, and computing a local phase image by filtering the first phase image with said filter to reduce nonlocal contributions on said corrected phase image.
Embodiments may include one or more of the following advantages. An advantage of embodiments is improved contrast in magnetic resonance imaging. Using this method an imaging experiment can be optimized by determining the contrast-to-noise ratios (CNR) of a final image as a function of the signal to noise ratio of the magnitude image (SNR), the phase difference between objects of interest (&Dgr;&phgr;), and the number of times (q) a phase mask is applied to the magnitude image. This permits selection of an optimal value of q to maximize CNR for given experimental conditions. For example, a given phase difference between features of interest may vary between different experiments due to hardware limitations, relaxation times, or susceptibility of the features of interest, yet the analysis determines a value of q that optimizes CNR in a final susceptibility weighted image in each of these cases. Alternatively, or in addition, given a target CNR the analysis provides experimental and processing parameters including echo time and multiplier q that most likely will result in a final image with the target CNR.
The entire contents of U.S. patent application Ser. No. 09/098,651, filed Jun. 17, 1998 and entitled “Application-specific optimization of echo time in MR pulse sequences for investigating materials with susceptibilities different from that of the background in which they are embedded” is incorporated herein by reference.
REFEREN
Fish & Richardson P.C.
Lateef Marvin M.
Pass Barry
LandOfFree
Susceptibility weighted imaging does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Susceptibility weighted imaging, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Susceptibility weighted imaging will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3141402