(19)
(11)EP 2 675 359 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
07.08.2019 Bulletin 2019/32

(21)Application number: 12709374.8

(22)Date of filing:  16.02.2012
(51)International Patent Classification (IPC): 
A61B 6/03(2006.01)
A61B 6/04(2006.01)
(86)International application number:
PCT/IB2012/050723
(87)International publication number:
WO 2012/110982 (23.08.2012 Gazette  2012/34)

(54)

IMAGING SYSTEM SUBJECT SUPPORT MOTION ALGORITHM(S)

BEWEGUNGSALGORITHMEN FÜR DIE PERSONENSTÜTZE EINES BILDGEBUNGSSYSTEMS

ALGORITHME(S) DE MOUVEMENT D'UN SUPPORT DE SUJET DE SYSTÈME D'IMAGERIE


(84)Designated Contracting States:
AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

(30)Priority: 18.02.2011 US 201161444155 P

(43)Date of publication of application:
25.12.2013 Bulletin 2013/52

(73)Proprietor: Koninklijke Philips N.V.
5656 AE Eindhoven (NL)

(72)Inventor:
  • DONG, Shufang
    5656 AE Eindhoven (NL)

(74)Representative: Steffen, Thomas 
Philips Intellectual Property & Standards High Tech Campus 5
5656 AE Eindhoven
5656 AE Eindhoven (NL)


(56)References cited: : 
WO-A1-2010/128417
US-A- 5 598 453
US-A1- 2008 013 674
US-A1- 2009 252 300
US-A1- 2010 166 138
US-B1- 6 381 299
WO-A2-2006/113323
US-A1- 2006 034 419
US-A1- 2008 031 407
US-A1- 2010 134 315
US-A1- 2010 310 040
  
  • David Bowling: "Advances in motion profiling", , 1 January 2008 (2008-01-01), XP055480360, ISBN: 978-1-109-01766-3 Retrieved from the Internet: URL:https://media.proquest.com/media/pq/cl assic/doc/1679799961/fmt/ai/rep/SPDF?_s=iq QsXRua0v3ybsSr5NEfzT1y6IY= [retrieved on 2018-06-01]
  • Bohlke: "Using Input Shaping to minimize residual vibration in flexible space structures", , 1 January 1993 (1993-01-01), XP055480374, Retrieved from the Internet: URL:http://hdl.handle.net/1721.1/36050 [retrieved on 2018-05-31]
  
Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


Description

FIELD OF THE INVENTION



[0001] The following generally relates to translating a subject support of an imaging system in connection with axial and/or perfusion scans of a subject carried by the subject support, and is described with particular application to computed tomography (CT). However, the following is also amenable to subject supports for other imaging modalities such as positron emission tomography (PET), x-radiology, and/or other imaging modalities.

BACKGROUND OF THE INVENTION



[0002] A computed tomography (CT) scanner includes a rotating portion rotatably supported by a stationary portion. The rotating portion supports an x-ray tube, which emits radiation that traverses an examination region and an object or a subject therein, and a detector array that detects radiation traversing the examination region and generates projection data indicative of the detected radiation. A subject support supports the object or subject in the examination region. A reconstructor reconstructs the projection data and generates volumetric image data indicative of the portion of the object or subject in the examination region. One or more images can be generated based on the volumetric image data.

[0003] The subject support generally includes a base, which is affixed to the floor of the examination room and configured to move vertically with respect to the floor, and a tabletop, which is affixed to the base and is configured to translate horizontally, with respect to the base, in and out of the examination region. Both axial and perfusion scans require frequent tabletop translation; an axial scan is an imaging mode in which the tabletop moves the object or subject to each axial position for scanning, but does not move the object or subject during scanning, and a perfusion scan is an imaging mode in which the tabletop cycles the object or subject back and forth in the examination region during scanning.

[0004] Traditionally, tabletop translation for both axial and perfusion scans is accomplished using a trapezoidal or an s-curve motion algorithm. FIGURE 1 illustrates an example s-curve motion algorithm for axial (step and shoot) scanning over three steps. In

[0005] FIGURE 1(A), a y-axis 102 represents tabletop position, in FIGURE 1(B), a y-axis 104 represents tabletop velocity, and in FIGURE 1(C), a y-axis 106 represents tabletop acceleration. In all three figures, an x-axis 108 represents time. As shown in FIGURE 1(C), tabletop acceleration and deceleration for the s-curve motion algorithm includes abrupt variable ramp up and ramp down, for each step.

[0006] FIGURES 2 and 3 respectively illustrate example trapezoidal and s-curve motion algorithms for perfusion (cyclic) scanning. In FIGURE 2(A) and 3(A), the y-axis 102 represents tabletop position, in FIGURE 2(B) and 3(B), the y-axis 104 represents tabletop velocity, and in FIGURE 2(C) and 3(C), the y-axis 106 represents tabletop acceleration, and the x-axis 108 in all six figures represents time. FIGURES 2(C) and 3(C) respectively show tabletop acceleration and deceleration with constant ramp up and ramp down and abrupt variable ramp up and ramp down, during each cycle.

[0007] The acceleration and deceleration profiles of the trapezoidal and s-curve motion algorithms of FIGURES 1-3 generally will result in tabletop vibration and motion of internal organs and tissue of the patient being scanned. More particularly, the tabletop is a lightly damped steel and composite structure with one or multiple resonant frequencies. Upon the moment of trapezoidal or s-curve point to point motion in the horizontal direction, the tabletop resonance can be aroused by the resonant component of the primary motion, resulting in undesired secondary motion or diving board vibration in the vertical direction. High acceleration and jerk due to the primary motion, when used cyclically in perfusion scans, may cause excessive human organ motion, resulting in an unpleasant patient experience.

[0008] Unfortunately, with axial scans, as the tabletop moving and settling time between scans becomes shorter and the intermitted travel becomes longer with wider coverage, motion acceleration and jerk will become greater, and, with perfusion scans, as the coverage becomes wider and the cycle time becomes shorter, perfusion scans can introduce greater secondary vibration. The secondary motion can be mitigated by use of a rigid based structure. However, with the conventional trapezoidal and s-curve algorithms, even with a stiff base structure, the amount of human organ motion and undesirable patient feeling generally cannot be mitigated.

[0009] US5598453A discloses an X-ray apparatus for collecting X-ray transmission data of a subject from a plurality of directions to generate an X-ray transmission image or X-ray CT image of the subject, which includes an X-ray generator for generating an X-ray, an X-ray detector for detecting a transmission X-ray after the X-ray generated by the X-ray generator is transmitted through the subject, a rotation unit for rotating an imaging unit including the X-ray generator and the X-ray detector around the subject, a data collector for converting an output signal of the X-ray detector to a digital signal and collecting the digital signal, a signal processor for subjecting data collected by the data collector to a signal processing operation, a display for displaying thereon as an image the data collected by the data collector and the data subjected by the signal processor to the signal processing operation, and a position change unit for moving a relative position of a rotation center of the imaging unit and the subject in a direction parallel to a rotation plane of the rotation, and wherein the imaging unit is rotated by the rotator around the subject and at the same time the relative position is changed by the position change unit in a direction parallel to the rotation plane to perform X-ray fluoroscopic or radiographic operation or CT scan.

[0010] US6381299B1 discloses that, between an X-ray generating system and an X-ray image detecting system arranged opposite to the X-ray generating system, an examination object supporting system is arranged and contains a straight movement table provided on a rotary table supported by a rotary table supporting member, and an examination object supporting member for supporting the examination object under either a standing position or a sitting position on the straight movement table. While the examination object is rotated by the rotary table, the examination object is continuously moved (reciprocating movement) by the straight movement table along a direction parallel to the rotation plane, and X-ray images of the examination object are acquired along a plurality of directions during both the rotating operation and the moving operation. As a result, both an X-ray tomographic image and a three-dimensional image (stereoscopic image) of a wider area than a viewing field of an X-ray II can be acquired. Since the viewing fields of the X-ray tomographic image and of the 3-dimensional image can be employed, both diagnostic performance and also a diagnostic efficiency with respect to a large organ such as a lung can be improved.

[0011] US2010310040A1 discloses a CT system including a scintillator array having a first length in an axial direction of the CT system, the scintillator array including a plurality of scintillator cells along the first length. The CT system includes a controller configured to repeatedly position a subject fore and aft along the axial direction and over a second length of the CT system, the second length greater than the first length, energize an x-ray source to emit x-rays toward the subject while the subject is being repeatedly positioned, and obtain non-uniform projection data of the subject from the scintillator array, the non-uniform projection data comprising the x-rays received from the x-ray source while the subject is repeatedly positioned.; The CT system includes a computer programmed to reconstruct a plurality of temporally non-uniformly spaced images from the non-uniform projection data, interpolate the plurality of non-uniform images to form a plurality of temporally uniform images, and apply a perfusion map generation process to the plurality of temporally uniform images.

[0012] US2008031407A1 discloses to easily associate X-ray projection data and scanning table z-direction coordinate information with each other. Using set parameters of the operations of a scanning gantry and a scanning table, the association of the X-ray projection data and scanning table z-direction coordinate information with each other is executed. Thereafter, image reconstruction is carried out based on the X-ray projection data to obtain a tomographic image. The operation set parameters are stored as part of the X-ray projection data. Alternatively, they are collectively stored even in the case of files separate from the X-ray projection data.

BRIEF DESCRIPTION OF THE DRAWINGS



[0013] The invention is defined in the independent claims 1 and 4.

[0014] The invention may take form in various components and arrangements of components, and in various steps and arrangements of steps. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention.

FIGURE 1 illustrates a prior art s-curve motion algorithm for axial point to point scanning.

FIGURE 2 illustrates a prior art trapezoidal motion algorithm for perfusion cyclic scanning.

FIGURE 3 illustrates a prior art s-curve motion algorithm for perfusion cyclic scanning.

FIGURE 4 schematically illustrates an example imaging system in connection with a subject support controller for controlling a subject support of the imaging system during scanning based on a subject support motion algorithm.

FIGURE 5 illustrates sinusoidal position, velocity and acceleration motion profiles for a sinusoidal harmonic axial motion algorithm for axial scanning.

FIGURE 6 illustrates simulated tabletop vibration for the sinusoidal harmonic axial motion algorithm of FIGURE 5.

FIGURE 7 illustrates simulated tabletop vibration for the s-curve axial motion algorithm of FIGURE 1.

FIGURE 8 illustrates sinusoidal position, velocity and acceleration motion profiles for a filtered time delay compensated axial motion algorithm for axial scanning.

FIGURE 9 illustrates simulated vibration for the filtered compensation motion algorithm of FIGURE 8 for axial scanning.

FIGURE 10 illustrates sinusoidal position, velocity and acceleration motion profiles for a profile shaping motion algorithm for axial scanning.

FIGURE 11 illustrates simulated vibration for the profile shaping motion algorithm of FIGURE 10 for axial scanning.

FIGURE 12 illustrates sinusoidal position, velocity and acceleration motion profiles for a filtered s-curve motion algorithm for perfusion scanning.

FIGURE 13 illustrates sinusoidal position, velocity and acceleration motion profiles for a sinusoidal harmonic algorithm for perfusion scanning.

FIGURE 14 illustrates an example model for modeling the elastic and viscous nature of soft tissue for the perfusion motion algorithms.

FIGURE 15 shows an example motion profile of a harmonic single step axial scan.

FIGURE 16 shows an example motion profile of a harmonic perfusion scan.

FIGURE 17 illustrates an example method for determining a tabletop motion algorithm for a selected imaging protocol.


DETAILED DESCRIPTION OF EMBODIMENTS



[0015] FIGURE 4 schematically illustrates an imaging system 400 such as a computed tomography (CT) scanner. The imaging system 400 includes a stationary gantry 402 and a rotating gantry 404, which is rotatably supported by the stationary gantry 402. The rotating gantry 404 rotates around an examination region 406 about a longitudinal or z-axis. A radiation source 408, such as an x-ray tube, is supported by and rotates with the rotating gantry 404 around the examination region 406. The radiation source 408 emits radiation that is collimated by a source collimator to produce a generally fan, wedge, cone, or otherwise shaped radiation beam that traverses the examination region 406. A radiation sensitive detector array 410 includes a one or two dimensional array of detector pixels that respectively detect radiation that traverses the examination region 406 and generates an electrical signal indicative of the detected radiation.

[0016] A subject support 412, such as a couch, supports an object or subject in the examination region 406. The subject support 412 includes a base portion 414 and a tabletop 416, and is configured to position (vertically and/or horizontally) the object or subject, with respect to x, y, and/or z directions, in the examination region 406 for scanning. For axial (step and shoot) scans, this includes moving the tabletop 416 and hence the object or subject to a predetermined position before an axial scan, maintaining a static position during the scan, and moving the tabletop 416 and hence the object or subject to next a position for a next axial when there is another axial scan to perform. This is repeated for each axial scan of the imaging procedure. For perfusion scans, this includes moving the tabletop 416 cyclically and hence the object or subject cyclically in coordination with scanning the object or subject.

[0017] A subject support controller 418 controls the movement of the subject support 412. The illustrated subject support controller 418 includes one or more processors 420 and computer readable storage medium 422 (e.g., physical memory), which stores computer readable instructions, which, when executed by the one or more processors 420 cause the controller 418 to transmit one or more control signals to control the subject support 412. The one or more processors 420 can additionally or alternative execute computer readable instructions carried in a signal, carrier wave or other transitory medium.

[0018] The illustrated storage medium 422 includes computer readable instructions such as one or more subject support motion algorithms 424. Examples of such algorithms include, but are not limited to, a sinusoidal harmonic axial motion algorithm, a feed-forward profile shaping axial motion algorithm, a feed-forward profile filtering with time delay compensation axial motion algorithm, a sinusoidal harmonic perfusion motion algorithm, a filtered perfusion motion algorithm, traditional axial and/or perfusion motion algorithms such as a trapezoidal or s-curve axial and perfusion motion algorithms, a combination of one or more of the above-noted motion algorithms, and/or one or more other motion algorithms.

[0019] As described in greater detail below, at least one of the non-traditional trapezoidal or s-curve axial and perfusion motion algorithms may mitigate at least one of tabletop vibration or internal subject motion, which, where the subject is a patient, may improve the experience of the patient and/or image quality. In one non-limiting instance, this is achieved while maintaining move duration, move distance, average speed, etc. relative to traditional trapezoidal or s-curve motion axial and perfusion motion algorithms, while providing jerk and acceleration motion profiles that minimize the undesired motion.

[0020] A reconstructor 426 reconstructs the signal generated by the detector array 408 and generates volumetric image data indicative of the examination region 406, including the portion of the object or subject therein. A general purpose computing system serves as an operator console 428, and includes an output device such as a display, an input device such as a keyboard, mouse, and/or the like, one or more processor and computer readable storage medium substantially similar to the storage medium 422. The console 428 allows the operator to control operation of the system 400, for example, allowing the operator to select a scan protocol (which may be associated with a default or user specified subject support motion algorithm), select or change the subject support motion algorithm, initiate scanning, etc.

[0021] FIGURES 5, 8 and 10 illustrate a non-limiting set of subject support motion algorithms well-suited for axial scans, FIGURES 7, 9 and 11 illustrate simulated vibration for the set of algorithms, and FIGURE 6 shows simulated vibration for a traditional s-curve axial motion algorithm.

[0022] Generally, a typical axial scan involves multiple steps of rest to rest motion. While the tabletop 416 comes to rest at a particular position, e.g., after each step move in the horizontal direction, inertia force of the step move excites the tabletop 416 and the base 414, causing a diving board like vibration of the tabletop 416, and the object or subject will vibrate in the vertical direction. The illustrated subject support 412 is a lightly damped metal and composite apparatus so resonance is inherent. The weight of the object or subject carried by the subject support 412, when coupled with acceleration, reduces the resonant frequency, and increase the inertia force required to complete the step motion, so the object or subject increases the amplitude of the diving board vibration. From the above, rest to rest moving is an external cause of the diving board vibration, and the low frequency resonance of the subject support 412 is an internal, or system, cause of the diving board vibration. With the illustrated subject support 412, the dominant frequency of the axial scan motion is smaller than the resonant frequency of the subject support even with the heaviest patient load.

[0023] Initially referring to FIGURE 5, an example sinusoidal motion algorithm with the identical time and distance as the s-curve motion algorithm of FIGURE 1 is illustrated.

[0024] In FIGURE 5(A), the y-axis 102 represents tabletop position, in FIGURE 5(B), the y-axis 104 represents tabletop velocity and in FIGURE 5(C), the y-axis 106 represents tabletop acceleration, and the x-axis 108 represents time in all three figures. Note that the tabletop velocity and position profiles in FIGURE 5 are very similar to the tabletop velocity and position profiles of FIGURE 1. However, the tabletop acceleration profiles of the sinusoidal harmonic algorithm of FIGURE 5 is much smoother than the tabletop acceleration profile of FIGURE 1, mitigating undesired motion.

[0025] FIGURES 6 and 7 respectively show simulated tabletop vibration for the s-curve axial motion algorithm of FIGURE 1 and the sinusoidal axial motion algorithm of FIGURE 5. In FIGURE 6 and 7, a y-axis 602 represents tabletop vibration and the x-axis 604 represents position. Note that the tabletop vibration for the sinusoidal harmonic axial motion algorithm of FIGURE 7, during the time periods when the axial scans are acquired (i.e., between 1 and' 2, 3 and 4, and 5 and 6), has a peak residue vibration 702 (e.g., 0.0846) that is about half (50%) of a peak residue vibration 606 (e.g., 1.664) for the s-curve axial motion algorithm of FIGURE 1. In other embodiments, the peak 702 is about 25% to about 75% of the peak 606.

[0026] With both FIGURE 6 and FIGURE 7, the tabletop vibration dynamics are simulated based on the following transfer function

which has a unit response for DC input, a resonant frequency

and a damping ratio γ = 0.025. Other transfer functions may also be used to characterize the tabletop.

[0027] Experimental results for motion distance of 33.6mm, 39.2mm, 50.5mm, 60mm, 68.6mm, 76.4mm with a load of 90lb, 180lb, 300lb, and 450lb for step and shoot scans are in agreement with the simulated data. By way of example, for a 300lb load, the average vibration reduction of a sinusoidal move is about 47% compared with the s-curved move, and for a 450 lb load, average the vibration reduction of a sinusoidal move is about 46% compared with the s-curved move.

[0028] Turning to FIGURE 8, an example filtered compensation motion algorithm for axial scanning is discussed.

[0029] In this example, a band-stop filter such as a notch filter or the like can be employed where the resonant frequency and damping factor of the subject support 412 are known or can be measured or estimated. Generally, a notch filter is a pole-zero cancelling technology that basically uses new zeros to cancel the resonant poles of the original system, and places new well damped poles to achieve a stable non-oscillatory response. Compared with a low pass filter, a notch filter generally introduces less delay at least in part due to phase lag concentricity to the resonant frequency and has much better vibration squeezing due to the stable poles and cancellation of resonant poles. The time delay to the original motion planning can be accounted for in order to scan a patient in the right rest intervals.

[0030] In FIGURE 8(A), the y-axis 102 represents tabletop position after the time delay is factored in and the original s-curved position profile is filtered with the notch or other band pass filter, in FIGURE 8(B), the y-axis 104 represents tabletop velocity and in FIGURE 8(C), the y-axis 106 represents tabletop acceleration, and the x-axis 108 represents time in all three figures. In FIGURE 9, the y-axis 602 represents tabletop vibration and the x-axis 604 represents position. Note that the simulated vibration has a peak residue vibration 902 of about or on the order of zero.

[0031] In the illustrated embodiment, the time delay compensation factor is expressed in terms of the resonant period

of the couch vertical dynamics. Using

from above, an s-curved step motion of one second, one distance, and one move, and the following example notch filter

to balance the resonance, the time delay introduced is T = 0.2 second. This time delay is accounted for in the profiles of FIGURE 8.

[0032] Turning to FIGURE 10, an example profile shaping motion algorithm for axial scanning is discussed.

[0033] In this example, feed-forward input shaping is implemented by convolving a sequence of impulses, an input shaper, with a desired motion profile to produce a shaped motion profile input that produces self-cancelling command signal to limit the residue vibration of rest to rest move pattern. Similar to the filtered compensation motion algorithm discussed in FIGURE 8, the input is the subject support resonant frequency and damping factor (which is used in this case to compute the gain and time location of the impulse functions), and time delay is accounted for in a digital implementation. However, compared with the filtered compensation motion algorithm, the input shaping filter introduces half of the notch-filter's delay, and has better residue vibration attenuation.

[0034] In FIGURE 10(A), the y-axis 102 represents tabletop position, in FIGURE 10(B), the y-axis 104 represents tabletop velocity and in FIGURE 10(C), the y-axis 106 represents tabletop acceleration, and the x-axis 108 represents time in all three figures. In FIGURE 11, the y-axis 602 represents tabletop vibration and the x-axis 604 represents position. Similar to the notch filtered axial motion algorithm, the simulated vibration for the input shaping axial motion algorithm has a peak residue vibration 1102 of about or on the order of zero.

[0035] For system with transfer function of

the time domain expression of the input shaper can be expressed as M(t)=C1δ(t)+C1δ(t-Td), wherein

and

are coefficients of the input shaper,



is the time delay factor, and δ(t) is the unit impulse function. For a given input motion profile x(t), the input shaped motion profile is xm(t)=x(t)∗M(t), where "" denotes mathematical convolution operation. In one instance, the above time domain convolution can be converted to continuous time transfer function, which is commonly used in simulation, and then the continuous time delay transfer function can be approximated with a continuous rational function. In another instance, the continuous time domain input shaper can be directly converted to discrete the time domain transfer function xm(z)=x(z)(C1+C2z-n) where n is the integer part of the value

with sampling period Δt and C1+C2z-n is the feed-forward digital input shaping filter. Due to the time delay of the input shaping, the original move time has to be shortened Td time and the acceleration is increased.

[0036] FIGURES 12 and 13 illustrate a non-limiting set of subject support motion algorithms well-suited for at least perfusion scans.

[0037] For a perfusion scan, a general concern is patient comfort and the impact of organ motion in horizontal direction on image quality, and since the patient lies on the tabletop 416, patient comfort is affected by the force pattern introduced by the tabletop friction, i.e., the tabletop's acceleration and jerk and its frequency pattern of the primary motion. Generally low frequency and low acceleration will result in lower jerk and better patient feeling.

[0038] FIGURE 12 shows a filtered point-to-point s-curve perfusion motion algorithm. In FIGURE 12(A), the y-axis 102 represents tabletop position, in FIGURE 12(B), the y-axis 104 represents tabletop velocity and in FIGURE 12(C), the y-axis 106 represents tabletop acceleration, and the x-axis 108 represents time in all three figures. FIGURE 13 shows a sinusoidal harmonic perfusion motion algorithm. In FIGURE 13(A), the y-axis 102 represents tabletop position, in FIGURE 13(B), the y-axis 104 represents tabletop velocity and in FIGURE 13(C), the y-axis 106 represents tabletop acceleration, and the x-axis 108 represents time in all three figures. As can seen from FIGURES 12, 13, 2 and 3, the tabletop acceleration of both the filtered point-to-point s-curve perfusion motion algorithm (FIGURE 12) and the sinusoidal harmonic perfusion motion algorithm (FIGURE 13) are smoother and thus result in less jerk and patient discomfort relative to the tabletop acceleration of the trapezoidal perfusion motion algorithm (FIGURE 2) and the point-to-point s-curve perfusion motion algorithm (FIGURE 3).

[0039] For the filtered point-to-point s-curve perfusion motion algorithm (FIGURE 12), a first order low pass filter (e.g.,

) or higher order filter can be applied to the non-harmonic motion to alleviate the acceleration and jerk level of the non-harmonic motion to obtain better patient feeling. Note that the cut-off frequency ωσ of the low pass filter has to be carefully chosen. Too high cut-off frequency, the low pass filter has no motion smoothing effect, while too low cut-off frequency will produce too much motion coverage reduction and phase lag. For perfusion scanning, the phase lag is not an issue (like it is in axial scanning, as described above) as long as it is repetitive and constant, but the excessive coverage reduction has to be compensated. Around four (4) times cut-off frequency of the perfusion cycle frequency can be used to the low pass filter to balance the coverage cutting and motion smoothing effect. The effect of the low pass filter on the point to point s-curve perfusion motion phase lag and amplitude cutting can be seen in FIGURE 12 where the cut-off frequency is 2 Hz for a 0.5 Hz perfusion cycle (i.e., a one (1) second end-to-end move, or a period of two (2) seconds).

[0040] FIGURE 14 shows an example model that can be used to model the elastic and viscous nature of soft tissue for the perfusion motion algorithms.

[0041] The model assumes human organ as a passive, stable, and low pass systems, which may or may not have resonance, and is not sensitive to high frequency excitation. In this model, the relative human organ motion "X-Y" will be small if the organ position X can follow the tabletop position Y closely, where 'm' is the mass of the organ, 'k' is the stiffness of the tissue, and 'c' is the damping coefficient of the organ to external perturbation. Based on this model, for a harmonic perfusion motion profile, which only has one very low frequency component, the motion of the organ will follow the tabletop motion closely without extra motion mismatch or lag, while the motion mismatch will be much larger for the same amplitude and frequency non-harmonic motion profile because of its inherent high-frequency harmonic component of the non-sinusoidal motion profile according to the Fourier transformation theory, and human organ dynamics will filter the high frequency out. Of course, other models can additionally and/or alternatively be used.

[0042] Experiments verified the impact of the different perfusion motion profiles in terms of water ripple in the half-filled fish tank that is placed on the couch tabletop. Four kinds of motion profile: back and forth trapezoidal motion so triangle speed, point to point trapezoidal motion with 50ms delay, point to point trapezoidal motion with 255ms smoothing factor and 50ms delay so basically a s-curve motion, and back and forth sinusoidal motion were tested with cycle time around 1 second and cycle distance around 160mm. The results show all other motion profiles except sinusoidal motion cause very dramatic water ripple. Other experiments showed the impact of the smaller acceleration level with 1.5 seconds and 2 seconds cycle time sinusoidal motion. With these experiments, barely any water ripple could be seen in case of 160 mm travel and 1.5 seconds cycle time sinusoidal motion. The experiments further showed that sinusoidal motion with even longer coverage (higher average speed and acceleration) still achieves much smaller water ripple. Further experiments included human subjects for determining true human feeling for the sinusoidal perfusion motion algorithm profile with 180mm coverage of 1 second cycle time that lasts 1 minute, ends up 886mm/s2 peak acceleration and 282mm/s peak speed. The results showed no dizziness or unpleasant feeling for the subjects. The long coverage and short cycle time combination with other perfusion motion algorithms is a concern for point to point back and forth perfusion scan.

[0043] With respect to the example motion algorithms discussed herein, the sinusoidal motion algorithms have a 90 degree phase shift between the velocity and acceleration curves, which may result in relatively smaller drive power capacity compared with the other simultaneous peak speed and peak acceleration motion profile, because speed corresponds to voltage for motor drive, acceleration means torque so drive current, and power is derived by speed times force or by voltage times current. Smaller power drive means cost saving opportunity. Furthermore, different from the zero speed, zero acceleration at beginning and end of point to point s-curve back and forth perfusion, the initial and end speed of the sinusoidal motion at the beginning and end is zero, but the sinusoidal acceleration is peak at both ends. This is a unique feature of the sinusoidal perfusion motion profile. If optimal or desired sinusoidal perfusion motion cannot be implemented with a particular subject support and/or motion controller, the sub-optimal motion profile still satisfies a 90 degree phase shift of speed and acceleration curve to get the smallest acceleration level and smoothest acceleration change to shore up patient feeling.

[0044] The following provides a general summary of various non-limiting motion algorithms discussed herein.

[0045] For feed-forward input shaping (FIGURE 8) and band pass filtering (FIGURE 10) axial motion algorithms, the time delay is factored in first to calculate the acceleration profile, then the acceleration is integrated to get speed, which is integrated to get position, and then the profile is sampled and digitally shaped or filtered to generate the modified motion profile for couch primary motion input. For feed-forward low pass filtering of the perfusion motion profile (FIGURE 12), the low pass filter cut-off frequency is calculated first, and then the original profile is digitally sampled and low pass filtered to generate the filtered motion profile.

[0046] The harmonic motion algorithms for the axial scan (FIGURE 5) and perfusion scan (FIGURE 13) are similar in theory but different in implementation. The pattern of axial scan move includes move duration, stop duration for scan, then next move duration, and stop duration, and more dependent on the move steps. The pattern of perfusion scan includes move, then back move, then move, and more dependent on the perfusion cycles, and no intermittent stop duration. The harmonic single step point to point move of axial scan includes a full wave of acceleration cycle to begin and end the move with initial zero jerk, acceleration and speed, which is essential to minimize the secondary motion. The harmonic cyclic move of perfusion scan includes a half wave of acceleration and speed cycle due to the repetitive feature of the motion.

[0047] A non-limiting harmonic motion profile for an axial scan can be based on,





and, wherein A represents a single step move distance, T represents the move period, s represents the real time position, υ represents the real time speed, α represents s the real time acceleration, and t represents the time index. FIGURE 15 shows an example motion profile (acceleration, velocity and position) for a single step harmonic point to point motion algorithm for an axial scan with A = 50 mm and T = 1 s.

[0048] A non-limiting harmonic motion profile for a perfusion scan can be based on



and

wherein A represents cyclic move (end to end distance) distance, T represents the (end to end time) move period, s represents the real time position, υ represents the real time speed, α represents s the real time acceleration, and t represents the time index. FIGURE 16 shows an example motion profile (acceleration, velocity and position) for a harmonic perfusion scan with A = 1 mm and T = 1 s.

[0049] To command a harmonic motion, the subject support controller 418 can generate a motion reference profile in advance, which is an array of position and time pairs. Where the controllers 418 is not capable of the computing a sinusoidal function online, a look up table and be generated and stored for the controller 418. By way of example, the point to point harmonic move trajectory of unit distance and unit period with 100 (or some other number) samples can be determined offline dependent on the controller's memory size. In one instance, the trajectory is an array of 100 components. In this instance, each component can be represented as two parameters, a position count parameter and a time index parameter. Such trajectory data can be saved into the storage 422 of the controller 418 as a lookup table. If the desired move period is X rather than unit second, then the time index is multiplied by X. If the desired move distance is Y rather than unit length, then the position count is multiplied by Y. Two harmonic motion modes can be created, one for single step move, and the other for the cyclic move. Each harmonic motion function will have a set of functional variables of move distance and move time.

[0050] Although the motion algorithms are primarily discussed in connection with CT, it is to be understood that one or more of them can be employed in connection with one or more other imaging modalities such as PET, X-ray, etc. In one instance, the one or more algorithms can be used to mitigate and/or minimize the induced secondary vibration and/or human organ motion and to improve patient feeling. The one or more algorithms can also be used with other single step or multiple steps or back and forth point to point positioning scenarios with or without rest time in between which concern the unnecessary residue motion.

[0051] FIGURE 17 illustrates an example method for determining a tabletop motion algorithm for a selected imaging protocol.

[0052] It is to be appreciated that the ordering of the acts in the methods described herein is not limiting. As such, other orderings are contemplated herein. In addition, one or more acts may be omitted and/or one or more additional acts may be included.

[0053] At 1702, a signal indicative of a user selected imaging protocol is received. The selected imaging protocol may be an axial or a perfusion scan.

[0054] At 1704, it is determined whether the selected protocol is an axial or perfusion scan protocol.

[0055] At 1706, where the selected protocol is for an axial scan, it is determined whether the patient's weight is known and available. Such information may be part of the electronic patient information, entered by a clinician, and/or dynamically measured, for example, via a scale or the like incorporated into the subject support 412 or otherwise.

[0056] At 1708, where the selected protocol is for an axial scan and the patient's weight is known, a resonant frequency and damping factor of the subject support 412 are determined based at least on the patient's weight. This may include mapping the patient's weight to a look up table (LUT) that cross references patient weight with subject support resonant frequency and damping factor. The LUT can be stored in the storage 422 or other storage.

[0057] At 1710, where the resonant frequency and damping factor are determined, the processor 420 identifies one of the feed-forward input shaping (FIGURE 8) or the band pass filtered (FIGURE 10) axial motion algorithms.

[0058] At 1712, where the selected protocol is for an axial scan and the patient's weight is not known, the processor 420 identifies the sinusoidal harmonic axial scan motion algorithm (FIGURE 5).

[0059] At 1714, where the selected protocol is for a perfusion scan, the processor 420 identifies one of the feed-forward low pass filtered perfusion motion algorithm (FIGURE 12) or the perfusion harmonic motion algorithm (FIGURE 13).

[0060] At 1716, a profile for the identified tabletop motion algorithm is obtained or generated, and loaded by the subject support controller 418. The identified motion profile can dynamically generated, retrieved from storage 422 or other storage, derive, or otherwise obtained.

[0061] At 1718, the scan is performed using the profile to control the motion of the tabletop 416.

[0062] The invention has been described herein with reference to the various embodiments. Modifications and alterations may occur to others upon reading the description herein. It is intended that the invention be construed as including all such modifications and alterations insofar as they come within the scope of the appended claims.


Claims

1. An imaging system (400), comprising:

a subject support (412) configured to carry a subject being imaged in an examination region (406) of the imaging system (400), wherein the subject support (412) comprises a tabletop (416); and

a subject support controller (418) that is configured to position the subject support (412) in the examination region (406) for scanning the subject based on a motion algorithm (424), for a given translation duration and a given translation distance and further to identify the motion algorithm based on a type of scan;

characterized in that the motion algorithm (424) corresponds to a feed-forward input shaping axial motion algorithm or a time compensated band pass filtered s-curve axial motion algorithm or a low-pass filtered s-curve perfusion motion algorithm;

wherein the feed-forward input shaping axial motion algorithm is implemented by convolving a sequence of impulses with a desired motion profile to produce a shaped motion profile input;

wherein the time-compensated band pass filtered s-curve axial motion algorithm is implemented by employing a band-stop filter;

wherein for one of said axial motion algorithms the motion algorithm is based on a resonant frequency and damping factor of the subject support (412), which are determined based on a weight carried by the subject support (412); and

wherein for the low-pass filtered s-curve perfusion motion algorithm the subject support controller (418) is configured to generate a motion profile of the motion algorithm by sampling and low pass filtering a predetermined perfusion motion profile based on a predetermined cut-off frequency.


 
2. The imaging system (400) of claim 1, wherein for one of said axial motion algorithms the subject support (412) includes a scale for determining the weight of the subject when the subject is on the subject support (412).
 
3. The imaging system (400) of any of claims 1 to 2, wherein for one of said axial motion algorithms the subject support controller (418) is configured to dynamically generate a motion profile of the motion algorithm (424) by adding a time delay into a tabletop position profile and input shaping or band pass filtering the time-delayed tabletop position profile.
 
4. A method, comprising:

determining (1704) a type of a scan for a subject based on a selected scan protocol of the scan, wherein the type is one of an axial scan or a perfusion scan;

identifying (1710, 1712, 1714) a motion algorithm of interest (424), based on the type of scan, for a tabletop (416) of a subject support (412) of an imaging system (400) used to perform the scan, ;

generating or retrieving (1716) a tabletop motion profile for the algorithm (424);

loading (1716) the tabletop motion profile into a subject support controller (418); and

employing (1718) the subject support controller (418) to control the tabletop (416) during the scan of the subject based on the loaded tabletop motion profile;

characterized in that the motion algorithm of interest (424) comprises a feed-forward input shaping axial motion algorithm or a time compensated band pass filtered s-curve axial motion algorithm or a low-pass filtered s-curve perfusion motion algorithm;

wherein the feed-forward input shaping axial motion algorithm is implemented by convolving a sequence of impulses with a desired motion profile to produce a shaped motion profile input;

wherein the time-compensated band pass filtered s-curve axial motion algorithm is implemented by employing a band-stop filter;

wherein for one of sais axial motion algorithms the motion algorithm is based on a resonant frequency and damping factor of the subject support (412), which are determined based on a weight carried by the subject support (412); and

wherein the low-pass filtered s-curve perfusion motion algorithm is implemented by generating a motion profile of the motion algorithm by sampling and low pass filtering a predetermined profusion motion profile based on a predetermined cut-off frequency.


 
5. The method of claim 4, wherein for one of said axial motion algorithms the method further comprises:

receiving (1702) a signal indicative of a weight of the subject;

determining (1708) a resonant frequency and a damping ratio of the subject support (412) based on the received weight; and

generating the motion algorithm (424) based on the resonant frequency and the damping, wherein the scan is an axial scan and the motion algorithm (424) is a feed-forward input shaping axial motion algorithm or a time-compensated band pass filtered s-curve axial motion algorithm.


 
6. The method of claim 5, wherein the motion algorithm (424) includes a time delay that compensates for a time delay introduced by the shaping or filtering.
 


Ansprüche

1. Bildgebungssystem (400), umfassend:

einen Objektträger (412), der zum Tragen eines Objekts konfiguriert ist, das in einem Untersuchungsbereich (406) des Bildgebungssystems (400) abgebildet wird, wobei der Objektträger (412)
eine Tischplatte (416) umfasst; und

eine Objektträger-Steuerung (418), die konfiguriert ist, um den Objektträger (412) in dem Untersuchungsbereich (406) zu positionieren, um das Objekt basierend auf einem Bewegungsalgorithmus (424) für eine gegebene Übersetzungsdauer und eine gegebene Übersetzungsdistanz abzutasten und um weiter den Bewegungsalgorithmus basierend auf einem Abtasttyp zu identifizieren;

dadurch gekennzeichnet, dass der Bewegungsalgorithmus (424) einem Axialbewegungsalgorithmus zur Vorwärtskopplungseingabeformung oder einem zeitkompensierten bandpassgefilterten S-Kurven-Axialbewegungsalgorithmus oder einem tiefpassgefilterten S-Kurven-Perfusionsbewegungsalgorithmus entspricht;

wobei der Axialbewegungsalgorithmus zur Vorwärtskopplungseingabeformung implementiert wird, indem eine Sequenz von Impulsen mit einem gewünschten Bewegungsprofil konvolviert wird, um eine geformte Bewegungsprofileingabe zu erzeugen;

wobei der zeitkompensierte bandpassgefilterte S-Kurven-Axialbewegungsalgorithmus unter Einsetzen eines Bandsperrfilters implementiert wird;

wobei für einen der Axialbewegungsalgorithmen der Bewegungsalgorithmus auf einer Resonanzfrequenz und einem Dämpfungsfaktor des Objektträgers (412) basiert, die basierend auf einem Gewicht, das von dem Objektträger (412) getragen wird, bestimmt werden; und

wobei für den tiefpassgefilterten S-Kurven-Perfusionsbewegungsalgorithmus die Objektträger-Steuerung (418) konfiguriert ist, um ein Bewegungsprofil des Bewegungsalgorithmus durch Abfragen und Tiefpassfiltern eines vorbestimmten Perfusionsbewegungsprofils basierend auf einer vorbestimmten Grenzfrequenz zu erzeugen.


 
2. Bildgebungssystem (400) nach Anspruch 1, wobei für einen der Axialbewegungsalgorithmen der Objektträger (412) eine Skala zum Bestimmen des Gewichts des Objekts aufweist, wenn das Objekt auf dem Objektträger (412) ist.
 
3. Bildgebungssystem (400) nach einem der Ansprüche 1 bis 2, wobei für einen der Axialbewegungsalgorithmen die Objektträger-Steuerung (418) konfiguriert ist, um dynamisch ein Bewegungsprofil des Bewegungsalgorithmus (424) durch Hinzufügen einer Zeitverzögerung in ein Tischplattenpositionsprofil und Eingabeformung oder Bandpassfilterung des zeitverzögerten Tischplattenpositionsprofils zu erzeugen.
 
4. Verfahren, umfassend:

Bestimmen (1704) eines Abtasttyps für ein Objekt basierend auf einem ausgewählten Abtastprotokoll der Abtastung, wobei der Typ eine Axialabtastung oder eine Perfusionsabtastung ist;

Identifizieren (1710, 1712, 1714) eines Bewegungsalgorithmus von Interesse (424) basierend auf dem Abtasttyp für eine Tischplatte (416) eines Objektträgers (412) eines Bildgebungssystems (400), das zum Durchführen der Abtastung verwendet wird;

Erzeugen oder Abrufen (1716) eines Tischbewegungsprofils für den Algorithmus (424); Laden (1716) des Tischplattenbewegungsprofils in eine Objektträger-Steuerung (418);

und

Einsetzen (1718) der Objektträger-Steuerung (418), um die Tischplatte (416) während des Abtastens des Objekts basierend auf dem Bewegungsprofil der geladenen Tischplatte zu steuern;

dadurch gekennzeichnet, dass der Bewegungsalgorithmus von Interesse (424) einen axialen Bewegungsalgorithmus zur Vorwärtskopplungseingabeformung oder einen zeitkompensierten bandpassgefilterten S-Kurven-Axialbewegungsalgorithmus oder einen tiefpassgefilterten S-Kurven-Perfusionsbewegungsalgorithmus umfasst;

wobei der Axialbewegungsalgorithmus zur Vorwärtskopplungseingabeformung implementiert wird, indem eine Sequenz von Impulsen mit einem gewünschten Bewegungsprofil konvolviert wird, um eine geformte Bewegungsprofileingabe zu erzeugen;

wobei der zeitkompensierte bandpassgefilterte S-Kurven-Axialbewegungsalgorithmus unter Einsetzen eines Bandsperrfilters implementiert wird;

wobei für einen der Axialbewegungsalgorithmen der Bewegungsalgorithmus auf einer Resonanzfrequenz und einem Dämpfungsfaktor des Objektträgers (412) basiert, die basierend auf einem Gewicht, das von dem Objektträger (412) getragen wird, bestimmt werden; und

wobei der tiefpassgefilterte S-Kurven-Perfusionsbewegungsalgorithmus durch Erzeugen eines Bewegungsprofils des Bewegungsalgorithmus durch Abfragen und Tiefpassfiltern eines vorbestimmten Perfusionsbewegungsprofils basierend auf einer vorbestimmten Grenzfrequenz implementiert wird.


 
5. Verfahren nach Anspruch 4, wobei für einen der Axialbewegungsalgorithmen das Verfahren weiter umfasst:

Empfangen (1702) eines Signals, das ein Gewicht des Objekts anzeigt; Bestimmen (1708) einer Resonanzfrequenz und eines Dämpfungsverhältnisses des Objektträgers (412) basierend auf dem empfangenen Gewicht; und

Erzeugen des Bewegungsalgorithmus (424) basierend auf der Resonanzfrequenz und der Dämpfung, wobei die Abtastung eine axiale Abtastung ist und der Bewegungsalgorithmus (424) ein Axialbewegungsalgorithmus mit Vorwärtskopplungseingabeformung oder ein zeitkompensierter, bandpassgefilterter S-Kurven-Axialbewegungsalgorithmus ist.


 
6. Verfahren nach Anspruch 5, wobei der Bewegungsalgorithmus (424) eine Zeitverzögerung aufweist, die eine Zeitverzögerung kompensiert, die durch das Formen oder Filtern eingeführt wird.
 


Revendications

1. Système d'imagerie (400) comprenant :

un support de sujet (412) configuré pour porter un sujet imagé dans une région d'examen (406) du système d'imagerie (400), dans lequel le support de sujet (412) comprend un plateau (416) ; et

un dispositif de commande de support de sujet (418) qui est configuré pour positionner le support de sujet (412) dans la région d'examen (406) pour balayer le sujet sur la base d'un algorithme de mouvement (424), pour une durée de déplacement en translation et une distance de déplacement en translation données, et en outre pour identifier l'algorithme de mouvement sur la base d'un type de balayage ;

caractérisé en ce que l'algorithme de mouvement (424) correspond à un algorithme de mouvement axial de mise en forme d'entrée prédictive ou à un algorithme de mouvement axial à courbe en s filtré passe-bande compensée dans le temps ou à un algorithme de mouvement de perfusion à courbe en s filtré passe-bas ;

dans lequel l'algorithme de mouvement axial de mise en forme d'entrée prédictive est mis en oeuvre en convoluant une séquence d'impulsions avec un profil de mouvement souhaité pour produire une entrée de profil de mouvement mise en forme ;

dans lequel l'algorithme de mouvement axial à courbe en S filtré passe-bande compensé dans le temps est mis en oeuvre en utilisant un filtre coupe-bande ;

dans lequel, pour l'un desdits algorithmes de mouvement axial, l'algorithme de mouvement est basé sur une fréquence de résonance et un facteur d'amortissement du support de sujet (412), qui sont déterminés en fonction d'un poids supporté par le support de sujet (412) ; et

dans lequel, pour l'algorithme de mouvement de perfusion à courbe en S filtré passe-bas, le dispositif de commande de support de sujet (418) est configuré pour générer un profil de mouvement de l'algorithme de mouvement en échantillonnant et en filtrant passe-bas un profil de mouvement de perfusion prédéterminé sur la base d'une fréquence de coupure prédéterminée.


 
2. Système d'imagerie (400) selon la revendication 1, dans lequel pour l'un desdits algorithmes de mouvement axial, le support de sujet (412) inclut une échelle permettant de déterminer le poids du sujet lorsque le sujet se trouve sur le support de sujet (412).
 
3. Système d'imagerie (400) selon l'une quelconque des revendications 1 à 2, dans lequel pour l'un desdits algorithmes de mouvement axial, le dispositif de commande de support de sujet (418) est configuré pour générer de manière dynamique un profil de mouvement de l'algorithme de mouvement (424) en ajoutant un retard temporel dans un profil de position de plateau et en mettant en forme d'entrée ou en filtrant passe-bande le profil de position de plateau temporisé.
 
4. Procédé, comprenant les étapes consistant à :

déterminer (1704) un type de balayage pour un sujet sur la base d'un protocole de balayage sélectionné du balayage, dans lequel le type est un parmi un balayage axial ou un balayage de perfusion ;

identifier (1710, 1712, 1714) un algorithme de mouvement d'intérêt (424), sur la base du type de balayage, pour un plateau (416) d'un support sujet (412) d'un système d'imagerie (400) utilisé pour effectuer le balayage, ;

générer ou récupérer (1716) un profil de mouvement de plateau pour l'algorithme (424) ;

charger (1716) le profil de mouvement de plateau dans un dispositif de commande de support de sujet (418) ; et

utiliser (1718) le dispositif de commande de support de sujet (418) pour commander le plateau (416) pendant le balayage du sujet sur la base du profil de mouvement de plateau chargé ;

caractérisé en ce que l'algorithme de mouvement d'intérêt (424) comprend un algorithme de mouvement axial de mise en forme d'entrée prédictive ou un algorithme de mouvement axial à courbe en s filtrée passe-bande compensée dans le temps ou un algorithme de mouvement de perfusion à courbe en s filtré passe-bas ;

dans lequel l'algorithme de mouvement axial de mise en forme d'entrée prédictive est mis en oeuvre en convoluant une séquence d'impulsions avec un profil de mouvement souhaité pour produire une entrée de profil de mouvement mise en forme ;

dans lequel l'algorithme de mouvement axial à courbe en s filtré par passe-bande et compensé dans le temps est mis en oeuvre en utilisant un filtre coupe-bande ;

dans lequel pour l'un des algorithmes de mouvement axial, l'algorithme de mouvement est basé sur une fréquence de résonance et un facteur d'amortissement du support de sujet (412), qui sont déterminés en fonction d'un poids supporté par le support sujet (412) ; et

dans lequel l'algorithme de mouvement de perfusion à courbe en S filtré passe-bas est mis en oeuvre en générant un profil de mouvement de l'algorithme de mouvement en échantillonnant et en filtrant passe-bas un profil de mouvement de perfusion prédéterminé sur la base d'une fréquence de coupure prédéterminée.


 
5. Procédé selon la revendication 4, dans lequel pour l'un desdits algorithmes de mouvement axial, le procédé comprend en outre les étapes consistant à :

recevoir (1702) un signal indiquant un poids du sujet ;

déterminer (1708) une fréquence de résonance et un facteur d'amortissement du support de sujet (412) sur la base du poids reçu ; et

générer l'algorithme de mouvement (424) sur la base de la fréquence de résonance et de l'amortissement, dans lequel le balayage est un balayage axial et l'algorithme de mouvement (424) est un algorithme de mouvement axial de mise en forme d'entrée prédictive ou un algorithme de mouvement axial à courbe en s filtré passe-bande et compensé dans le temps.


 
6. Procédé selon la revendication 5, dans lequel l'algorithme de mouvement (424) comprend un retard temporel qui compense un retard temporel introduit par la mise en forme ou le filtrage.
 




Drawing












































Cited references

REFERENCES CITED IN THE DESCRIPTION



This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

Patent documents cited in the description