Surgery – Diagnostic testing – Detecting nuclear – electromagnetic – or ultrasonic radiation
Reexamination Certificate
1999-05-21
2002-08-13
Smith, Ruth S. (Department: 3737)
Surgery
Diagnostic testing
Detecting nuclear, electromagnetic, or ultrasonic radiation
C600S413000, C324S309000
Reexamination Certificate
active
06434412
ABSTRACT:
BACKGROUND OF THE INVENTION
The invention relates to MR imaging, and more particularly relates to MR imaging of the heart. In its most immediate sense, the invention relates to MR cine images of the heart such as are used to assess cardiac function and anatomy.
When cardiac function is to be evaluated using cine MR imaging, the image should distinguish clearly between the blood and the myocardium. Conventionally, this is accomplished by using a spoiled gradient-echo MR pulse sequence in which inflowing blood has a more intense signal than myocardium does.
Such a conventional technique can produce unsatisfactory results when the repetition time TR is decreased (as when a cardiac function study is to be carried out during a single breath hold). During a very short repetition time, only a small volume of blood can flow into a slice of interest. For this reason, when the repetition time is short so that one RF pulse follows quickly after another, only a small fraction of the blood within a slice of interest is replaced by inflowing blood. The blood that remains in the slice therefore becomes saturated, diminishing the difference between the signal intensity of the blood and the signal intensity of the myocardium. This problem is particularly acute in standard long-axis and four-chamber views of the heart. In these views, blood flow is primarily parallel to the slice plane, and little or no saturated blood is replaced between one RF pulse and the next RF pulse.
At short repetition times, certain MR pulse sequences (e.g. sequences referred to as “TrueFISP” sequences) produce image contrast as a function of the T
2
/T
1
ratio (rather than relying upon blood flow for contrast). Such MR pulse sequences would therefore be appropriate candidates for use in cardiac function studies (since the T
2
/T
1
ratio of blood significantly exceeds the T
2
/T
1
ratio of the myocardium).
Accordingly, it would be advantageous to provide a method for conducting a cardiac imaging MR study, and particularly a cardiac functional MR study, in which the contrast between the blood and the myocardium remained significant at short repetition times TR, whereby the study could be conducted during a shorter breath hold (or, alternatively, conducted with higher temporal or spatial resolution than using MR sequences in which the repetition time TR is longer).
The invention proceeds from the realization that when a TrueFISP-type of MR pulse sequence is used to image the heart and the blood therein, segmenting the MR pulse sequence not only reduces (and possibly eliminates) artifacts caused by respiratory motion but also reduces the acquisition time so that the acquisition can be completed within a reasonable breath-hold (on the order of ten to twenty heartbeats). This makes it possible to produce MR cine images having adequate spatial and temporal resolution in a relatively short period of time.
As used in this patent application and as understood by persons skilled in the art, a “segmented” MR pulse sequence is one that causes the final k-space matrix to be built up from sub-matrices comprising groups of lines of MR data, the lines in each group being distributed over more than one cardiac cycle. To understand the meaning of this term, two examples of unsegmented MR pulse sequences will be compared with a segmented MR pulse sequence.
In the first instance, let it be assumed that a particular k-space matrix is made up of 165 lines of MR data taken at different phase-encodings and that a typical unsegmented MR pulse sequence is used to acquire the MR data. In this instance, one line of data would typically be acquired during each cardiac cycle, so that 165 heartbeats would be required to complete the image acquisition. Few if any patients could hold their breath for such a long time. The result would be an MR acquisition having a very long duration and producing a reconstructed image having a very high temporal resolution.
In a second instance, let it be assumed that all 165 lines of MR data were to be acquired during a single cardiac cycle (i.e. during a single heartbeat), using an unsegmented MR pulse sequence. In this instance, the time needed to acquire each image would be too long to accurately depict cardiac function, even if the repetition time TR were chosen to be very short. In this instance, the MR acquisition would have a very short duration and would produce a reconstructed image having a very low temporal resolution.
In the third and last instance, let it be assumed that a segmented MR pulse sequence were to be used to acquire the 165 lines of MR data. In this instance, some (e.g. eleven) lines of MR data would be acquired during one heartbeat and the rest acquired during other heartbeats (i.e. eleven per heartbeat during fourteen more heartbeats so that a total of 165 lines of MR data would ba acquired over fifteen heartbeats). This MR acquisition would have a moderate duration and would produce a reconstructed image having a moderate temporal resolution.
In accordance with the invention, a segmented but otherwise conventional TrueFISP-type cine MR pulse sequence is used to image the patient's heart and the blood therein. (Advantageously but not necessarily, this is done during a single breath-hold.) Initially, the magnetizations of the heart and blood are brought to a steady-state. Then, lines of MR data are acquired using the segmented TrueFISP-type MR pulse sequence. (During this data acquisition, the magnetizations of the heart and blood remain in steady-state, because this is an intrinsic characteristic of TrueFISP-type MR pulse sequences.)
The MR pulse sequence may be of the two-dimensional type or of the three-dimensional type. In one preferred embodiment, MR data acquisition is gated to the patient's cardiac cycle. In accordance with this embodiment, the steady-state magnetizations of the heart and blood are maintained regardless whether lines of MR data are being acquired. In another preferred embodiment, MR data acquisition continues without interruption while the patient's cardiac cycle is monitored, and the data are retrospectively gated to form multiple images, each representing the state of the heart at a different time within a composite cardiac cycle.
REFERENCES:
patent: 5273040 (1993-12-01), Apicella et al.
patent: 5337000 (1994-08-01), Bruder
patent: 5447155 (1995-09-01), NessAiver et al.
patent: 5545992 (1996-08-01), Foo
patent: 5668474 (1997-09-01), Heid
patent: 5997883 (1999-12-01), Epstein et al.
patent: 6034528 (2000-03-01), Heid
patent: 6073041 (2000-06-01), Hu et al.
patent: 6078175 (2000-06-01), Foo
Westbrook et al., MRI in Practice, Blackwell Scientific Publications, 1993, pp. 102-131 and 200-203.*
Heid, O. “TrueFisp Cardiac Fluoroscopy” 1997, Abstract Proceedings of the ISMRM, p. 320.
Bundy Jeffrey M.
Finn J. Paul
Kroeker Randall
Simonetti Orlando P.
Jay Mark H.
Siemens Medical Systems Inc.
Smith Ruth S.
LandOfFree
Cardiac cine imaging with a short repetition time and high... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Cardiac cine imaging with a short repetition time and high..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Cardiac cine imaging with a short repetition time and high... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2909250