Method for correcting localization error caused by field inhomogeneity in MRI wireless marker tracking

Abstract

This invention provides a method for correcting the localization errors caused by field inhomogeneity and obtaining position information of a wireless marker in an MRI scanner. In one embodiment, said method comprises the steps of: a) Providing said wireless marker in said workspace; b) Performing a pulse sequence with said MRI scanner, comprising: i) Exciting all spins inside said workspace; ii) Applying a pre-phasing gradient along a frequency encoding; iii) Applying dephasing gradients in two directions orthogonal to said frequency encoding direction; and iv) Applying two readout gradients with same amplitude and opposite polarities along said frequency encoding direction to form a first and a second gradient echo; c) Acquiring a first and a second MR signal using an array of RF receiver coils by sampling said first and second gradient echoes respectively; and d) Reconstructing tracking signals from said first and second MR signals and calculating said position information.

Description

METHOD FOR CORRECTING LOCALIZATION ERROR CAUSED BY FIELD INHOMOGENEITY IN MRI WIRELESS MARKER TRACKINGFIELD OF THE INVENTION

[0001] The present invention relates to systems and methods for wireless marker tracking under MRI system, particularly wireless markers positions in the workspace of an MRI scanner.BACKGROUND OF THE INVENTION

[0002] Wireless marker tracking is an MRI marker tracking method with the use of wireless markers1 1. MRI marker tracking uses radiofrequency (RF) coils as markers and could provide the positions of markers inside MRI scanners. Wired markers4’5are connected to the MRI scanner via cable connection for signal excitation and / or transmission. In comparison, wireless markers interact with MRI system via inductive coupling to the coil system of scanner and hence eliminate the need for cable connections1. Consequently, wireless markers have several advantages over wired markers. First, wireless markers do not need additional receiver channels for cable connection, and therefore, have lower hardware requirements for MRI scanners. Second, the removal of cables with wireless markers reduces the possibility of heating and improves patient safety. In addition, wireless markers are more convenient for patient handling and can improve patient comfort. Therefore, a significant number of works regarding wireless marker tracking have been proposed and developed in recent years.

[0003] Wireless marker tracking can be used in various MRI related areas. First, wireless marker tracking enables retrospective or prospective motion correction. Subject motion during MRI scan can be unavoidable, especially for challenging populations, such as elderly or pediatric patients. It can cause MRI image artifacts and affect the following clinical applications, e.g., lowering clinical diagnosis confidence. Hence, motion tracking and correction is an important MRI topic and advantageous to many MRI applications. With the wireless markers attached to the surfaces of imaging object, the position and motion information of the imaging object inside MRI scanner can be obtained in real time via wireless marker tracking. Accordingly, either retrospective or prospective motion correction can be performed with real-time motion information obtained from the wireless markers. Wireless marker tracking yields advantages over other motion correction methods. For example, optical methods such as cameras, require optical line-of-sight and a relatively complicated crosscalibration procedure for motion tracking and correction6. Navigator methods can be not suitable for some MRI imaging pulse sequences. In comparison, wireless marker tracking is relatively cheap and has no requirements for line-of-sight. In addition, it is suitable for almost all imaging purposes. Therefore, it may benefit the widespread use of motion correction in clinical practices. As reported in many works1’7, wireless marker tracking has demonstrated its uses in prospective motion correction. In addition to motion correction, wireless markertracking could help with MR-guided robot-assisted surgery and interventions8. With wireless markers attached to interventional devices or the robot base, they can be used for device tracking, device localization and robot registration9" , However, obtaining accurate and reliable marker positions is crucial for different applications of wireless marker tracking in MRI.

[0004] Wireless marker tracking requires the use of a tracking pulse sequence to encode and acquire MRI signals that contain marker position information. The currently used tracking pulse sequence, as described in U.S. Pat. No. 10,591,570 B2, applies frequency encoding gradients to encode MR signals. However, during the frequency encoding, the presence of field inhomogeneity can inevitably alter the encoding frequency of the acquired signals and therefore cause local pixel shifts in the obtained tracking signals. This will cause marker localization errors and may affect the performance of some applications relying on wireless marker tracking. Particularly, during MRI intervention, the intervention devices made from certain materials can introduce severe susceptibility effects that distort the magnetic field, therefore causing significant pixel shifts and marker localization errors. Furthermore, magnetic field distortion caused by susceptibility effects is more severe at higher field strengths. Since pixel shifts are proportional to the magnetic field deviation, the leading pixel shifts and marker localization errors are accordingly more significant in a high-field MRI scanner.

[0005] It would therefore be desirable to provide a method for fast wireless marker tracking under MRI system that can eliminate the marker localization errors caused by field inhomogeneity; that is applicable to wireless markers used for MRI marker tracking, to benefit the development of applications with wireless marker tracking, such as interventional MRI and motion correction techniques.SUMMARY OF THE INVENTION

[0006] This invention provides a method for obtaining position information of a wireless marker in a workspace of an MRI scanner. In one embodiment, said method comprises the steps of: a) Providing said wireless marker in said workspace; b) Performing a pulse sequence with said MRI scanner, wherein said pulse sequence comprises: i) Exciting all spins inside said workspace; ii) Applying a pre -phasing gradient along a frequency encoding direction to prepare for generation of gradient echoes; iii) Applying dephasing gradients in two directions orthogonal to said frequency encoding direction to reduce intensity of signals arising from sources other than said wireless marker; and iv) Applying two readout gradients with same amplitude and opposite polarities along said frequency encoding direction to form a first gradient echo and a second gradient echo; c) Acquiring a first MR signal and a second MR signal using an array of RF receiver coils by sampling said first and second gradient echoes respectively; and d) Reconstructing tracking signals from said first and second MR signals and calculating said position information from said tracking signals.BRIEF DESCRIPTION OF THE FIGURES

[0007] Figure. 1 shows a flowchart describing the steps of obtaining accurate wireless marker positions unaffected by field inhomogeneity with method proposed in this invention.

[0008] Figure. 2A shows an exemplary dual-echo fast tracking pulse sequence with full echo acquisition.

[0009] Figure. 2B shows an exemplary dual-echo fast tracking pulse sequence with partial echo acquisition.

[0010] Figure. 3 is an example of a one-dimensional projection signal, in which three peaks indicate the position of three markers respectively.

[0011] Figure. 4A shows an exemplary tracking pulse sequence for obtaining the three- dimensional positions of wireless marker with full echo readouts.

[0012] Figure. 4B shows an exemplary tracking pulse sequence for obtaining the three- dimensional positions of wireless marker with partial echo readouts.

[0013] Figure. 5A shows the marker position measurements and errors of marker A measured from the first gradient echo (Echo 1), the second gradient echo (Echo 2), and the average of positions from the two echoes.

[0014] Figure. 5B shows the marker position measurements and errors of marker B measured from the first gradient echo (Echo 1), the second gradient echo (Echo 2), and the average of positions from the two echoes.

[0015] Figure. 5C illustrates the measurements errors from Echo 1, Echo 2, and the proposed average marker position with marker A.

[0016] Figure. 5D illustrates the measurements errors from Echo 1, Echo 2, and the proposed average marker position with marker B.

[0017] Figure. 6A illustrates the tracking signal obtained from the first gradient echo with full echo readouts.

[0018] Figure. 6B illustrates the tracking signal obtained from the first gradient echo with partial echo readouts.

[0019] Figure. 6C shows the marker position measurements and peak-to-noise ratio (PNR) values of two markers with full echo and partial echo readouts respectively.DETAILED DESCRIPTION OF THE INVENTION

[0020] The present invention provides a method for obtaining the accuracy positions of wireless markers under MRI system in a short duration. The method includes the use of a dualecho fast tracking pulse sequence with full or partial echo readouts to acquire MRI signals, and subsequent calculation of wireless marker positions unaffected by field inhomogeneity. The dual -echo fast tracking pulse sequence includes one set of pulses for each direction along whichmarker positions are to be obtained. Each set of pulses includes a radiofrequency (RF) pulse that excites all spins inside the workspace of the MRI scanner. It also includes a pre-phasing gradient followed by two readout gradients that have opposite polarities but the same amplitude, to create two consecutive gradient echoes during the signals sampling and data acquisition. The pre-phasing and readout gradients are applied along the direction which the markers ID positions are needed to measure, and that direction is also the frequency-encoding direction. Two consecutive tracking signals can therefore be respectively reconstructed by Fourier transforming the MRI signals acquired from the two gradient echoes. Because the two readout gradients have opposite polarities but the same amplitude, the local pixel shifts caused by field inhomogeneity existing in the two tracking signals are also opposites. Therefore, the accurate marker positions can be calculated by averaging the positions obtained from the two tracking signals and therefore correcting the localization error due to distorted magnetic field. In addition, the use with partial echo readouts can potentially reduce the duration of both prephasing and readout gradients, thereby achieving a shortened repetition time (i.e., TR) and improved racking rate for fast wireless marker tracking. Furthermore, the duel-echo fast tracking pulse sequence also includes dephasing gradients along the other two orthogonal directions to reduce the noise signals arising from the background (i.e., imaging object).

[0021] The foregoing and other aspects and advantages of the invention will appear from the following description. In the description, reference is made to the accompanying drawings that form a part hereof, and in which there is shown by way of illustration a preferred embodiment of the invention. Such embodiment does not necessarily represent the full scope of the invention, however, and reference is therefore made to the claims and herein for interpreting the scope of the invention.

[0022] This invention provides a method for obtaining wireless marker positions unaffected by field inhomogeneity inside the workspace of an MRI scanner with a short acquisition duration. In one embodiment, the steps of the method comprises: a) Directing an MRI system to perform a pulse sequence comprising: Applying a nonselective radio frequency (RF) pulse with small flip angle to excite all spins inside the workspace of the MRI scanner; Applying one pre -phasing gradient along the frequency encoding direction following the application of the RF pulse to prepare for the generation of gradient echoes; Applying two readout gradients with the same amplitude and opposite polarities following the pre-phasing gradient in the frequency encoding direction, in order to form two gradient echoes and acquire MR signals containing marker position information with opposite errors caused by field inhomogeneity; Applying dephasing gradients in the other two orthogonal directions before the applications of readout gradients, in order to reduce the intensity of signals arising from sources other than markers; b) Acquiring MR signals with an array of RF receiver coils by sampling the formed echo signals; c) Reconstructing tracking signals and calculating accurate subpixel marker positions from the tracking signals.

[0023] In one embodiment, step a) further comprises applying partial echo readouts during gradient echoes to reduce the duration of the dual-echo tracking pulse sequence and improve its tracking rate.

[0024] In one embodiment, step c) includes applying one dimensional Fourier transformation or one dimensional partial Fourier reconstruction for full echo or partial echo readouts to the acquired MR signals and obtaining tracking signals.

[0025] In one embodiment, step c) includes finding the pixel which is nearest to the marker through peak detection with the tracking signal and calculating the subpixel marker position using subpixel interpolation.

[0026] In one embodiment, step c) further includes averaging the calculated subpixel positions signals of the first and second gradient echoes.

[0027] Systems and methods for fast wireless marker tracking under magnetic resonance imaging system, in which a dual -echo gradient echo sequence with full or partial echo readouts is utilized to obtain two sequential MRI tracking signals encoded by two gradients with opposite polarities. Consequently, the wireless marker localization errors induced by field inhomogeneity in two tracking signals are also opposite. Therefore, accurate wireless marker positions without field inhomogeneity induced errors can be calculated by averaging the positions obtained from these two tracking signals. The use of partial echo readouts can reduce the duration of the tracking sequence and improve its tracking rate. This dual -echo tracking pulse sequence can enable accurate wireless marker tracking with a high tracking rate, and therefore would benefit the applications of wireless marker tracking, such as interventional MRI and motion correction techniques.

[0028] This invention provides a method for obtaining position information of a wireless marker in a workspace of an MRI scanner. In one embodiment, said method comprises the steps of: a) Providing said wireless marker in said workspace; b) Performing a pulse sequence with said MRI scanner, wherein said pulse sequence comprises: i) Exciting all spins inside said workspace; ii) Applying a pre -phasing gradient along a frequency encoding direction to prepare for generation of gradient echoes; iii) Applying dephasing gradients in two directions orthogonal to said frequency encoding direction to reduce intensity of signals arising from sources other than said wireless marker; and iv) Applying two readout gradients with same amplitude and opposite polarities along said frequency encoding direction to form a first gradient echo and a second gradient echo; c) Acquiring a first MR signal and a second MR signal using an array of RF receiver coils by sampling said first and second gradient echoes respectively; and d) Reconstructing tracking signals from said first and second MR signals and calculating said position information from said tracking signals.

[0029] In one embodiment, said two readout gradients are applied with opposite polarity.

[0030] In one embodiment, each readout gradient has double area compared to said prephasing gradient for full echo readouts.

[0031] In one embodiment, said method further comprises applying partial echo readouts during said first and second gradient echoes.

[0032] In one embodiment, said step (d) comprises applying one dimensional Fourier transformation to said first and second MR signals for full echo readouts.

[0033] In one embodiment, said step (d) comprises applying one dimensional partial Fourier reconstruction to said first and second MR signals for partial echo readouts.

[0034] In one embodiment, said step (d) comprises finding a pixel nearest to said wireless marker through peak detection with said tracking signals and calculating said position information using subpixel interpolation.

[0035] In one embodiment, said position information comprises subpixel position.

[0036] In one embodiment, said subpixel position is obtained by averaging a first subpixel position obtained from said first gradient echo and a second subpixel position from said second gradient echo.

[0037] In one embodiment, said wireless marker is placed along said frequency encoding direction.

[0038] In one embodiment, said step b(i) comprises applying a nonselective radio frequency (RF) pulse with small flip angle.

[0039] In one embodiment, said small flip angle is 1 to 10 degrees.

[0040] In one embodiment, said method further comprises repeating step (b) with one or more different frequency encoding directions to obtain position information from another dimension.

[0041] In one embodiment, said method further comprises repeating step (b) with a second frequency encoding direction and a third frequency encoding direction to obtain 3-dimenional position information; wherein said frequency encoding direction, said second frequency encoding direction and said third frequency encoding direction are orthogonal to each other.

[0042] Described here are systems and methods for fast wireless marker tracking under MRI system. In general, a designed dual -echo fast tracking pulse sequence is utilized to acquire two consecutive tracking signals which contain embedded wireless marker position information along the one projection direction. Two tracking signals containing opposite localization errors caused by field inhomogeneity are obtained for marker tracking along each direction. Therefore, by averaging the marker positions calculated from the two consecutive tracking signals, wireless marker positions unaffected by field inhomogeneity can be obtained. In addition, partial echo readouts are applied to shorten the duration of the dual-echo fast tracking pulse sequence for improving tracking rate. The proposed reliable and accurate fast wireless marker tracking method may benefit the applications of wireless marker tracking such as interventional MRI and motion correction techniques.

[0043] Field inhomogeneity, also referred to as off-resonance effect12, is a common problem that exists in MRI systems, and it can result from several factors. First, hardware and systemimperfection can cause the inhomogeneous static main magnetic field in MRI scanner. Second, the magnetic susceptibility property of imaging objects can distort the magnetic field. Magnetic susceptibility either increases or decreases the local magnetic field strength. Since the imaging object can include components with different magnetic susceptibility (e.g., different tissues), susceptibility effects can cause inhomogeneous magnetic field. Field inhomogeneity is often unavoidable in MRI, and it can induce deviations in the rotating frequencies of spins that are the sources of MRI signals, therefore causing image artifacts due to spatial encoding errors.

[0044] The previously used fast wireless marker tracking pulse sequence, as described in U.S. Pat. No. 10,591,570 B2, acquires only one gradient echo for obtaining ID projection signals that contains the information about marker positions along one single direction. Specifically, gradient echo uses frequency encoding to encode MRI signals of the spins at different spatial locations. The underlying principle of frequency encoding is to assign different rotating frequencies to spins located at different positions along the direction to be projected. Afterward, the spins positions can be calculated and derived by analyzing the frequency spectrum of the acquired signal. However, field inhomogeneity causes deviations in spins rotating frequencies, therefore leading to pixel shifts in the calculated tracking signals, and marker localization errors. A higher magnetic field can induce larger field inhomogeneity13, and hence the marker localization errors can be more severe under the MRI scanners with higher field strengths. These marker localization errors can reduce the performance of applications relying on wireless marker tracking, such as interventional MRI and MRI motion correction. Therefore, removing marker localization errors induced by field inhomogeneity is valuable and highly meaningful for different wireless marker tracking applications.

[0045] The systems and methods of the present invention differ from the conventional fast wireless marker tracking method described in U.S. Pat. No. 10,591,570 B2, in that the conventional method produces one single gradient echo for encoding MRI signals and obtaining marker positions along one direction, whereas the systems and methods of the present invention utilize dual gradient echoes produced by two opposite readout gradients for obtaining marker positions along each direction. In addition, partial echo readouts are used for data acquisition to shorten the duration of the dual-echo fast tracking pulse sequence. Accordingly, in each excitation of the conventional method, only one MR signal is obtained and used for the calculation of marker positions, while in the systems and methods of the present invention, two consecutive MR signals are obtained from the dual gradient echoes, and both are used for calculating marker positions. Because the dual gradient echoes are generated by two readout gradients with opposite polarities but the same strength, the marker localization errors existing in the two consecutively acquired MR signals are also opposite. Accordingly, in contrast to the marker positions obtained in the conventional method are contaminated by the pixels shifts induced by field inhomogeneity, systems and methods of the present invention can obtain marker positions unaffected by field inhomogeneity via averaging the marker positionsobtained during the two echoes. Moreover, the use of partial echo readouts can further reduce the tracking time and improves the tracking rate for fast wireless marker tracking with dualecho fast tracking pulse sequence.

[0046] Referring to Figure. 1, a flowchart describing the steps of one embodiment of the invention is illustrated. First MRI signals are acquired using the proposed dual-echo fast tracking pulse sequence. Second the MR signals are Fourier transformed to ID projection MR signals in which the peaks indicate the projected marker positions along the encoding direction. Third, subpixel marker positions are calculated from the obtained ID projection signals. Last, the desirable marker positions unaffected by field inhomogeneity are calculated by averaging the subpixel marker positions obtained from the two consecutively acquired gradient echoes. Details of these steps are described below.

[0047] An example of a dual-echo fast tracking pulse sequence of the present invention is illustrated in Figure. 2A. The pulse sequence uses full echo acquisition. It includes a nonselective radio frequency (RF) excitation pulse 201 that is played out in the beginning, in order to produce transverse magnetization and therefore generate MR signals. The nonselective RF excitation pulse can excite all spins inside the MRI workspace. Here the MRI workspace refers to the space where the coil systems of the MRI scanner work so that signal excitation and acquisition can be performed. Therefore, spins near or inside the wireless markers can be excited as long as markers are placed inside the MRI workspace. The flip angle of the RF excitation pulse is usually less than 10 degrees for the fast tracking pulse sequence. As one example, the flip angle can be 1 degree. This is because under usual conditions, there are other signal sources, such as human tissues, which can also generate MRI signals, in addition to the spins near markers. The use of an ultra-small flip angle (i.e., 1 degree) is to minimize the intensity of signals arising from these sources and therefore reduce their interference to marker signals.

[0048] After the excitation of all spins inside MRI workspace, MR signals are to be encoded and acquired using gradient echoes. A pre-phasing gradient 202 is applied first along the direction which marker positions are to be obtained. This is the frequency-encoding direction, which is the x-direction shown in Figure. 2A. This pre-phasing gradient moves the first sampled k-space point away from the k-space center, to prepare for gradient echo generation. After the pre-phasing gradient, the first readout gradient 203 is applied along the same frequency-encoding direction. To generate the first gradient echo, readout gradient 203 is applied with opposite polarity and double area compared to the pre-phasing gradient 202. During the gradient echo acquisition, readout gradient 203 assigns spins along the encoding direction (the x-direction) with linearly-varying rotating frequencies to achieve frequency encoding. Assume the amplitude of gradient 203 is G. since its polarity is positive as shown in Figure.2A, the acquired MR signal S(t) can be expressed as:Formula (1)Where p is the proton density, y is the gyromagnetic ratio, x represents the location of spin along the frequency-encoding direction, and AB represents the variation in the magnetic field strength induced by field inhomogeneity. Equation (1) can then be transformed to the following equation:Formula (2) and with kx= yGt Formula (3)The following equation can be obtained:Formula (4)Define Axxas Formula (5)ThenFormula (6)Where Axxrepresents the pixel shift existing in this signals, i.e., marker localization error. According to Equation (5), the pixel shift is determined by both the magnetic field deviation AB and the readout gradient strength G .

[0049] The second readout gradient 204 is applied following the first readout gradient 203, with the same amplitude and an opposite polarity to gradient 203. It produces the second gradient echo and allows for the second frequency encoding of MR signals. Similarly, the pixel shift contained in the signals acquired during this second gradient echo Ax2can be derived as:ABAX2= -G = — AxxFormula (7)Since the second readout gradient has a polarity opposite to the first readout gradient, pixel shifts exisiting in the two acquired signals are opposite for the same local inhomogeneous magnetic field.

[0050] In addition, the fast tracking pulse sequence includes two dephasing gradients 205 and 206 applied in the other two orthogonal directions. For example, as shown in Figure. 2A, they are applied in the y-direction and the z-direction respectively. Dephasing gradients 205 and 206 are applied before the readout gradients to add various phase changes to spins located at different positions along these two directions, therefore reducing their resultant signal intensity. For example, dephasing gradient 205 is applied in the y-direction, and therefore it causes additional various phase changes to the spins located at different y positions. Accordingly, the resultant signal strength of these spins decreases. Because the size of markers are relatively small compared to other signal sources inside MRI workspace, the dephasing gradients can be applied with specified areas to decrease signals from other sources with largesize while at the same time preserving signals generated from wireless markers. Under such conditions, spins around markers still have close phases, and therefore their resultant signal barely decreases.

[0051] Figure. 2B illustrates another example of the dual -echo fast tracking pulse sequence with partial echo readouts, in which the curved dash lines indicate the signals acquired. With other parameters kept the same, the pre-phasing gradient 207 has a smaller area than the prephasing gradient 202 in Figure. 2A to enable partial echo acquisition. In addition, becasue data points to be acquired during each gradient echo are reduced with partial echo acquisition, the durations of readout gradients 208 and 209 are also reduced compared to gradient 203 and 204 for full echo acquisition. The echo times of the first and the second gradient echo are reduced accordingly. Even though the amount of acquired data points decreases, the shortened echo times for two echoes can partially compensate the loss of signal-to-noise ratio (SNR) and still provide enough signal intensity for wireless marker tracking. Gradients 208 and 209 still have the same amplitude and the opposite polarities, and hence the pixel shifts contained in the two consecutively acquired signals keep opposite with the same inhomogeneous field.

[0052] With the dual-echo fast tracking pulse sequence, two one -dimensional MRI signals are consecutively acquired during each repetition time (TR). As indicated by Equation (6), these signals are k-space signals, and they need to be Fourier transformed to spatial domain for obtaining marker position information, as follows:Formula (8)Where F represents one-dimensional Fourier transformation, and / (%) is the obtained onedimensional (ID) projection signal (i.e., tracking signal). For data acquired with partial echo readouts, F represents one-dimensional partial Fourier reconstruction. An example of ID projection signal is illustrated in Figure. 3. Peaks in the ID projection signal indicate the projected positions of markers because most signal is generated from the spins around the markers. In Figure. 3, three peaks indicate the positions of three wireless markers, respectively.

[0053] Accordingly, subpixel marker positions can be calculated from the ID projection signals. First, the pixel with the maximum signal intensity in the tracking signal is obtained via peak detection, and the true position of the marker is around this pixel with peak signal. Afterward, subpixel marker positions are calculated with interpolation algorithms. One example of the interpolation algorithm is the intensity linear interpolation (ILI) method14, in which marker positions are calculated as the center of two positions with half-maximum intensity values. Third, the final marker position x is calculated by: Formula (9)where x1and x2represent the subpixel marker positions obtained from the first and second gradient echoes, respectively. This position is accordingly not affected by the marker localization error induced by field inhomogeneity.

[0054] Markers positions along one direction can be obtained by one set of dual-echo fast tracking pulse sequence as shown in Figures. 2A and 2B. In Figures. 2A and 2B, the frequency encoding direction is the x-direction, and therefore the acquired marker position is along the x-direction. Nevertheless, with the proposed method the frequency encoding direction can be applied in any specified direction in the 3D MRI workspace, to obtain the corresponding marker positions. Furthermore, for some related applications, the three- dimensional positions of markers inside the MRI scanner are more desirable and useful. Three sets of the proposed dual-echo fast tracking pulse sequence with three different orthogonal frequency encoding directions can be used for fast marker tracking for this purpose. As shown in Figure. 4A and Figure. 4B, three sets of fast tracking pulse sequence are applied consecutively. The first set is applied with frequency encoding direction along the x-direction, the second along the y-direction, and the third along the z-direction. Six one -dimensional MR signals are acquired during three excitations (i.e., three TRs), and afterward the positions of markers along the three orthogonal directions can be calculated.

[0055] The invention will be better understood by reference to the Experimental Details which follow, but those skilled in the art will readily appreciate that the specific experiments described are only for illustrative purpose and are not meant to limit the invention as described herein, which is defined by the claims that follow thereafter.

[0056] Throughout this application, various references or publications are cited. Disclosures of these references or publications in their entireties are hereby incorporated by reference into this application in order to more fully describe the state of the art to which this invention pertains. It is to be noted that the transitional term “comprising”, which is synonymous with “including”, “containing” or “characterized by”, is inclusive or open-ended and does not exclude additional, un-recited elements or method steps.

[0057] EXAMPLE 1

[0058] Experiments were performed on a 1.5 T MRI scanner (Artist, General Electric Healthcare) to evaluate the correction of marker localization errors for the present invention. Two tiny omnidirectional wireless markers3 proposed by Kwok et al. were used as the sources for fast wireless marker tracking. The two markers were placed at different locations along the superior-inferior (SI) direction and referred to as marker A and marker B, respectively. A bottle of water was also placed inside the MRI scanner workspace to provide the signal source other than markers, for simulating the presence of background signals in certain conditions, such as the human tissue during MRI intervention. The dual-echo fast tracking pulse sequence shown in Figure. 2A was used for signal acquisition of ID projection signals. In order to obtain marker positions along the SI direction of MRI scanner, the frequency-encoding direction was set as the SI direction. A 2D spoiled gradient echo imaging (SPGR) pulse sequence with its phaseencoding direction set as the SI direction was also performed for obtaining the gold standard of marker positions. It is because in two-dimensional (2D) gradient echo imaging, fieldinhomogeneity induced pixel shifts only exist in the frequency encoding direction, but not in the phase -encoding direction.

[0059] Both the dual-echo fast tracking pulse sequence of the present invention and the 2D SPGR pulse sequence were performed with three different receiver bandwidths for the data sampling: ±31.25 kHz, ±50 kHz, and ±62.5 kHz, for evaluating the marker localization errors under different receiver bandwidths. With each receiver bandwidth, the two pulse sequences were performed twice. The obtained marker positions from the 2D SPGR pulse sequence were averaged as the gold standard marker positions. Field-of-view (FOV) was set as 240 mm and kept the same for all experiments. The flip angle for both pulse sequences was 1 degree. A quadrature body coil and a 19-channel phase-array head coil were used for inductive coupling to the wireless markers and receiving MR signals, respectively.

[0060] Marker positions were successfully obtained as shown as in Figures. 5A to 5D. Figure. 5A shows the marker positions of marker A measured from the first gradient echo (Echo 1), the second gradient echo (Echo 2), and the average of positions from the two echoes. The results were from the first experiment. Gold standard positions were calculated from the corresponding 2D SPGR data, and localization errors were calculated as the deviations between measurements and gold standard values. First, it can be observed that the localization errors corresponding to data from Echo 1 and Echo 2 had opposite signs for all receiver bandwidths. This is consistent with the theory that the two readout gradients with opposite polarities give rise to pixel shifts with opposite directions in relation to field inhomogeneity. Second, for marker positions obtained from Echo 1 or Echo 2, the localization errors became larger when using a lower receiver bandwidth. This agrees with the theory that the size of pixel shifts induced by field inhomogeneity is inversely proportional to the receiver bandwidth15. With the same FOV, the decrease of receiver bandwidth could reduce the strength of the readout gradient. Therefore, according to Equation 5, the pixel shift caused by field inhomogeneity was also increased. These results indicate that the localization errors in the marker positions measured from Echo 1 and Echo 2 are mainly affected by field inhomogeneity, and the degree of localization errors obeys the pixel shift calculation in Equation 5. Moreover, it is apparent that with all receiver bandwidths, averaging the subpixel positions from Echo 1 and Echo 2 effectively reduced errors in marker position measurements. Despite the absolute values of errors from Echo 1 or Echo 2 ranging from 0.176 mm to 0.651 mm, localization errors in the average marker positions were all smaller than 0.1 mm for all tested receiver bandwidths. Similar results can be observed in Figure. 5B, which demonstrates the first experimental results for marker B.

[0061] Line charts in Figure. 5C and Figure. 5D demonstrate the corresponding tracking errors measured from Echo 1, Echo 2, and the proposed average method with marker A and marker B, respectively. The dashed lines indicate the localization errors measured from Echo 1 in two experiments, while the dotted lines indicate errors from Echo 2. The solid lines showthe localization errors measured from the proposed average method. It is apparent in the line charts that the average marker positions of the proposed invention show insignificant errors than the positions measured from either Echo 1 or Echo 2. Moreover, while the localization errors in Echo 1 and Echo 2 were smaller when using a higher receiver bandwidth, the localization errors in the average marker positions maintain a similarly low level for all receiver bandwidths. As discussed above, the pixel shifts caused by field inhomogeneity are inversely proportional to the receiver bandwidth. Therefore, results showing no such trends in the average marker position errors suggest that the proposed invention can effectively correct the tracking errors caused by field inhomogeneity.

[0062] In conclusion, experimental results suggest the effectiveness of the proposed invention in correcting the field inhomogeneity induced errors in fast wireless marker tracking.

[0063] EXAMPLE 2

[0064] Experiments were performed on a 1.5 T MRI scanner (Artist, General Electric Healthcare) to validate the effectiveness of the proposed dual-echo fast tracking pulse sequence with partial echo readouts for improving the tracking rate of dual-echo fast tracking pulse sequence. Two tiny omnidirectional wireless markers3 proposed by Kwok et al. were used as the sources for fast wireless marker tracking. The two markers were placed at different locations along the superior-inferior (SI) direction and referred to as marker A and marker B respectively. A bottle of water was also placed inside the MRI scanner workspace to provide the signal source other than markers, for simulating the background signals in certain conditions, such as in the presence of human tissue during MRI intervention. The dual-echo fast tracking pulse sequence with either full echo readouts or partial echo readouts, as shown in Figure. 2A and Figure. 2B, respectively, were used for signal acquisition. The frequencyencoding direction was set as the SI direction of MRI scanner, and receiver bandwidth was ±31.25 kHz. Other parameters include: field of view (FOV) = 240 mm; flip angle = 1 degree; and 240 sampling points with full echo acquisition, while 144 sampling points with partial echo acquisition for each gradient echo. In the obtained tracking signals, peaks indicate the position and signals of markers, and the peak-to-noise ratio (PNR) was calculated for each marker in each tracking signal. PNR was calculated as the ratio of the intensity of the peak value and the background signal.

[0065] The duration of the tracking pulse sequence with full echo readouts was 11.8ms for the tracking along one single direction, while the duration with partial echo readouts was 9.23ms. Accordingly, the use of the partial echo acquisition improve the tracking rate by 27.84%. Figure. 6A and Figure. 6B show the comparison between the tracking signals obtained from full echo readout and partial echo readout, respectively. It can be observed that with the use of the partial echo readouts, markers still have sufficiently higher signal intensity than background noise. Figure. 6C illustrates the marker position measurements and peak-to- noise ratio (PNR) values of two markers with full echo and partial echo readouts, respectively.It can be observed that with comparable PNRs, marker positions obtained from full echo readouts and partial echo readouts were comparable. The maximum difference between positions measured from two types of readouts were 0.12 mm. In conclusion, the use of partial echo acquisition could improve the tracking rate by around 27.84% while preserving high tracking accuracy.

[0066] Reference:1. Ooi MB, Aksoy M, Maclaren J, Watkins RD, Bammer R. Prospective motion correction using inductively coupled wireless RF coils: Prospective Motion Correction Using Wireless Markers. Magn Reson Med. 2013;70(3):639-647. doi: 10.1002 / mrm.248452. Cheung CL, Ho JDL, Vardhanabhuti V, Chang HC, Kwok KW. Design and fabrication of wireless multilayer tracking marker for intraoperative MRI-guided interventions. IEEE / ASME Transactions on Mechatronics . 2020;25(2): 1016-1025.3. Cheung CL, Wu M, Fang G, et al. Omnidirectional Monolithic Marker for Intraoperative MR-based Positional Sensing in Closed MRI. IEEE Transactions on Medical Imaging. Published online 2023.4. Ooi MB, Krueger S, Thomas WJ, Swaminathan SV, Brown TR. Prospective real-time correction for arbitrary head motion using active markers. Magnetic Resonance in Medicine: An Official Journal of the International Society for Magnetic Resonance in Medicine. 2009;62(4):943-954.5. Qin L, Schmidt EJ, Tse ZTH, et al. Prospective motion correction using tracking coils. Magnetic Resonance in Med. 2013;69(3):749-759. doi: 10.1002 / mrm.243106. Maclaren J, Herbst M, Speck O, Zaitsev M. Prospective motion correction in brain imaging: a review. Magnetic resonance in medicine. 2013;69(3):621-636.7. Sengupta S, Tadanki S, Gore JC, Welch EB. Prospective real-time head motion correction using inductively coupled wireless NMR probes: Head Motion Correction Using Wireless NMR Probes. Magn Reson Med. 2014;72(4):971-985. doi: 10.1002 / mrm.250018. Su H, Kwok KW, Cleary K, et al. State of the art and future opportunities in MRI-guided robot-assisted surgery and interventions. Proceedings of the IEEE. 2022;110(7):968-992.9. He Z, Dong Z, Fang G, et al. Design of a percutaneous MRI-guided needle robot with soft fluid-driven actuator. IEEE Robotics and Automation Letters. 2020;5(2):2100-2107.10. Fang G, Chow MC, Ho JD, et al. Soft robotic manipulator for intraoperative MRI-guided transoral laser microsurgery. Science Robotics. 2021;6(57):eabg5575.He Z. Dai J, Ho JD, et al. Interactive Multi-Stage Robotic Positioner for Intra-Operative MRI-Guided Stereotactic Neurosurgery. Advanced Science. Published online 2023:2305495. Haskell MW, Nielsen J, Noll DC. Off-resonance artifact correction for MRI: A review. NMR in Biomedicine. 2023;36(5):e4867. Farahani K, Sinha U, Sinha S, Chiu LC, Lufkin RB. Effect of field strength on susceptibility artifacts in magnetic resonance imaging. Computerized Medical Imaging and Graphics. 1990;14(6):409-413. Rea M, McRobbie D, Elhawary H, Tse ZT, Lamperth M, Young I. Sub-pixel localisation of passive micro-coil fiducial markers in interventional MRI. Magnetic Resonance Materials in Physics, Biology and Medicine. 2009; 22:71-76. Manson EN, Inkoom S, Mumuni AN. Impact of Magnetic Field Inhomogeneity on the Quality of Magnetic Resonance Images and Compensation Techniques: A Review. RMI. 2022;Volume 15:43-56. doi: 10.2147 / RMI.S369491

Claims

What is claimed is:

1. A method for obtaining position information of a wireless marker in a workspace of an MRI scanner, comprising the steps of: a) Providing said wireless marker in said workspace; b) Performing a pulse sequence with said MRI scanner, wherein said pulse sequence comprises: i. Exciting all spins inside said workspace; ii. Applying a pre-phasing gradient along a frequency encoding direction to prepare for generation of gradient echoes; iii. Applying dephasing gradients in two directions orthogonal to said frequency encoding direction to reduce intensity of signals arising from sources other than said wireless marker; and iv. Applying two readout gradients with same amplitude and opposite polarities along said frequency encoding direction to form a first gradient echo and a second gradient echo; c) Acquiring a first MR signal and a second MR signal using an array of RF receiver coils by sampling said first and second gradient echoes respectively; and d) Reconstructing tracking signals from said first and second MR signals and calculating said position information from said tracking signals.

2. The method of claim 1, wherein said two readout gradients are applied with opposite polarity.

3. The method of claim 2, wherein each readout gradient has double area compared to said pre-phasing gradient for full echo readouts.

4. The method of claim 1, further comprising applying partial echo readouts during said first and second gradient echoes.

5. The method of claim 1, wherein said step (d) comprises applying one dimensional Fourier transformation to said first and second MR signals for full echo readouts.

6. The method of claim 1, wherein said step (d) comprises applying one dimensional partial Fourier reconstruction to said first and second MR signals for partial echo readouts.

7. The method of claim 1, wherein said step (d) comprises finding a pixel nearest to said wireless marker through peak detection with said tracking signals and calculating said position information using subpixel interpolation.

8. The method of claim 7, wherein said position information comprises subpixel position.

9. The method of claim 8, wherein said subpixel position is obtained by averaging a first subpixel position obtained from said first gradient echo and a second subpixel position from said second gradient echo.

10. The method of claim 1, wherein said wireless marker is placed along said frequency encoding direction.

11. The method of claim 1, wherein said step b(i) comprises applying a nonselective radio frequency (RF) pulse with small flip angle.

12. The method of claim 11, wherein said small flip angle is 1 to 10 degrees.

13. The method of claim 1, further comprising repeating step (b) with one or more different frequency encoding directions to obtain position information from another dimension.

14. The method of claim 1, further comprising repeating step (b) with a second frequency encoding direction and a third frequency encoding direction to obtain 3-dimenional position information; wherein said frequency encoding direction, said second frequency encoding direction and said third frequency encoding direction are orthogonal to each other.

Images