Apparatus and method to correct for position errors in diagnostic imaging

[0002] This application is a continuation-in-part of patent application Ser. No. 09/191,094, filed Nov. 12, 1998, U.S. Pat. No. 6,092,928.

[0003] The present invention relates to the art of medical imaging. It finds particular application in conjunction with canceling or correcting for undesired movement especially mechanical movement in C-arm supports for generating three-dimensional computed tomography imaging data, more particularly fluoroscopic x-ray systems, and will be described with particular reference thereto. It is to be appreciated, however, that the invention is also applicable to other diagnostic imaging systems.

[0004] In some operating rooms, such as operating rooms for vascular catheter procedures, a projection x-ray imaging device is provided in association with the operating table. More specifically, an x-ray tube or generator and an x-ray detector are mounted on a C-arm which is positioned such that the area of interest or patient lies between the x-ray source and detector. The x-ray source and detector are rotatable and longitudinally displaceable as a unit to select a region and angle for projection imaging. Once the surgeon has positioned the x-ray source and detector in the proper position, the surgeon actuates the x-ray tube sending x-rays through the patient and onto the x-ray detector for a preselected exposure time. The x-rays received by the detector are converted into electronic, video image data representing a projection or shadow-graphic image. The projection or shadow-graphic image is displayed on a video monitor which is viewable by the physician.

[0005] In cardiac catheterization procedures, for example, images are generated to show the vasculature system and monitor the advance of the catheter through the blood vessels. More specifically, the surgeon advances the catheter into the patient, stops the surgical procedure, and initiates an x-ray imaging procedure. The x-rays are converted into electronic data and a projection image is displayed.

[0006] One of the drawbacks of these x-ray systems is that the resultant image is a projection or shadow-graphic image. That is, the 3-D vasculature system of the patient is projected into a single plane.

[0007] If 3-D diagnostic images are required, such images are often taken with a CT scanner or a magnetic resonance imaging device which is typically located in another part of the facility. Thus, any three-dimensional diagnostic images are commonly generated sometime before the surgical procedure starts. Even if a CT scanner is present in the surgical suite, the patient is still moved into the scanner. The transportation of the patient to the CT or MRI machine for further imaging often renders three-dimensional images impractical during many surgical procedures.

[0008] However, three-dimensional images obtained are valuable during surgical procedures. After generating a three-dimensional diagnostic image, a surgical procedure is commenced, such as a biopsy. From time to time during the procedure, additional projection diagnostic images are generated to monitor the advancement of the biopsy needle into the patient. The location of the needle can be mathematically predicted from the projection images and monitoring of the physical position of the needle or other instrument can be superimposed on the 3-D diagnostic images. As the needle moves, the superimposed images can be altered electronically to display the needle in the proper position. Various trajectory planning packages have been proposed which would enable the operator to plan the biopsy procedure in advance and electronically try various surgical paths through the three-dimensional electronic data.

[0009] Recently, there has been some interest in using relatively low power fluoroscopic systems to generate real time three-dimensional CT reconstructions. Such a technique, disclosed in U.S. Pat. No. 5,841,830 to Barni, et al. is assigned to the assignee of this invention. Barni suggests operating the x-ray tube of a CT scanner in a fluoroscopic mode. Unfortunately, the complete, encircling CT gantry can obstruct access to the surgical site or make that access inconvenient or uncomfortable for the physician.

[0010] Another solution disclosed by R. Fahrig, et al. in SPIE Volume 2708 entitled “Characterization Of A C-Arm Mounted XRII For 3-D Image Reconstruction During Interventional Neuro Radiology” recognizes that a C-arm would provide improved access to the surgical site. The Fahrig article also observes that the C-arm lacks sufficient rigidity to prevent the x-ray source and the detector plates from moving relative to each other, especially during a volume scan where the source and the detector are rotated about an area of interest. Relative motion misaligns the apparatus and causes image degradations. The Fahrig article describes a method wherein the motions and the deflections of the C-arm are premeasured or estimated in pilot scans. The deflections are assumed to remain the same for subsequent scans performed from the same starting point and within all other parameters. System calibration is performed by inserting a three-dimensional phantom containing metal beads or the like with known locations into the imaging field and performing a representative scan. Subsequent image analysis is used to determine positional errors, due to C-arm distortion and deflection. By comparing the detected position of the beads in each image with calculated ideal positions that would occur in the absence of any C-arm distortion, errors for each angular position of the C-arm are calculated. These errors, for each image scan, are stored in a long-term memory and applied to the data collected at corresponding positions of the C-arm, correcting for the calibration errors. Unfortunately, the Fahrig method requires that all volume imaging scans begin in exactly the same location and travel through the same arc. Moreover, any changes in the mechanical characteristics of the C-arm, such as bearing wear, changes in the source to image distance, drive speed, etc., will cause a deterioration in image quality due to the application of improper positional corrections.

[0011] The present invention provides a new and improved method and apparatus which overcomes the above-referenced problems and others.

[0012] In accordance with the present invention, a diagnostic imaging apparatus includes an x-ray source for transmitting a beam of x-rays through an examination region. An accelerometer is associated with the x-ray source such that a change in linear velocity of the source corresponds to an acceleration reading being registered by the accelerometer.

[0013] In accordance with a more limited aspect of the present invention, the diagnostic imaging apparatus includes a second accelerometer for measuring acceleration of the detector.

[0014] In accordance with a more limited aspect of the present invention, a position calculator mathematically calculates a position of both the source and the detector from data including signals provided by the accelerometers.

[0015] In accordance with a more limited aspect of the present invention, an image reconstruction processor is included to receive a plurality of image data views and for processing the views into a three-dimensional image representation using the calculated position data.

[0016] In accordance with a more limited aspect of the present invention, the diagnostic imaging apparatus also includes a collimator movably mounted to the x-ray source for restricting the cone beam of x-rays onto the detector. A misalignment processor receives the position data and controls a drive system mechanically linked to the collimator.

[0017] In accordance with another embodiment of the present invention, a radiographic imaging apparatus includes a penetrating radiation source and a radiation receiver. The source and receiver are held in position by a mechanical structure on opposite sides of an examination region. Moreover, the apparatus includes a sensor which detects motion in a selected portion of the mechanical structure and generates a signal in response to that motion. A processor receives the signal from the sensor and calculates a correction required to compensate for at least a portion of the motion detected. The processor then directs the correction calculated to be applied either as a physical correction to the structure or as imaging processing on the received radiation information.

[0018] In accordance with another aspect of the present invention, a process for diagnostic imaging includes positioning a radiation source and a receiver relative to an examination region. The process also includes sensing a motion affecting either the source, the receiver, or both and calculating a correction to compensate for the motion. Once the correction has been calculated it is applied either to physically cancel oscillations in a mechanical support before the source transmits a radiation beam; or as imaging processing on the received radiation information after the source transmits a radiation beam.

[0019] One advantage of the present invention resides in decreased bulk, or rigidity in the mechanical support needed to resist vibration or motion.

[0020] Another advantage of the present invention resides in the acquisition and display of more accurate volumetric images.

[0021] Another advantage of the present invention resides in computationally simpler and more efficient image signal manipulation.

[0022] Another advantage of the present invention resides in the ability to obtain volume scans from any starting and stopping position.

[0023] Another advantage of the present invention resides in the ability to provide dynamic corrections without relying on periodic calibration with three-dimensional phantoms.

[0024] Yet other benefits and advantages of the present invention will become apparent to those skilled in the art upon reading and understanding the preferred embodiments.

[0025] 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 a preferred embodiment and are not to be construed as limiting the invention.

[0026] FIG. 1 is a diagrammatic illustration of an imaging system in accordance with the present invention;

[0027] FIG. 2 is a block diagram of a signal processor in accordance with the present invention;

[0028] FIG. 3 is a depiction of error induced by deflection;

[0029] FIG. 4 is a diagrammatic illustration of collimator adjustment to correct for induced errors;

[0030] FIG. 5 shows a diagrammatic illustration of an alternate embodiment of the present invention;

[0031] FIG. 6 is a diagrammatic illustration of an imaging system suitable to practice an alternate embodiment of the present invention;

[0032] FIG. 7 is an oscillation curve showing motion of an isocenter over time; and,

[0033] FIG. 8 is a block diagram of signal processing system for practicing the present invention.

[0034] With reference to FIG. 1, a mechanical structure M₁configured as a C-arm C is supported by a rotational support assembly A, for rotation around an examination region E. The examination region E is describable by orthogonal axes X, Y, and Z. An x-ray transparent couch 10 is positioned such that a region of interest of a subject 12 is positioned in the examination region E. Vertical and horizontal drives (not shown) move the couch to facilitate positioning the region of interest at the center or other appropriate location in the examination region E. Alternatively, the rotational support assembly arm A moves in the X or Z direction.

[0035] The rotational support assembly A of the mechanical structure M₁includes an overhead rotational mount or bearing 14 mounted to a ceiling or other overhead fixture for rotation about a vertical axis Y. In one preferred embodiment, the overhead rotational mount 14 is movably fixed to a track or other mechanism also allowing movement in the X and Z directions. An arm 16 extends away from the overhead mount 14 in the -Y direction through an elbow 18 to a lower rotational mount or bearing 20 with a horizontal axis of rotation.

[0036] A midpoint 22 of the C-arm C is rotatably attached to the lower bearing 20. The C-arm C defines two opposing parallel ends 24, 26, on either side of the examination region E. A receiver 28, such as a flat panel detector or the like for detecting x-ray radiation, is attached to the first end 24. The receiver 28 is preferably a solid state device, such as a grid of amorphous silicon detector elements, that generate x-ray intensity signals for each element of the grid. An x-ray source 30 is attached on the second end 26 of the C-arm. The C-arm C has sufficient strength to maintain the receiver 28 and x-ray source 30 in a substantially fixed spatial relationship. However, due to the mass of the x-ray tube and the arms, the ends 24, 26 deflect or move during positioning. On stopping the ends oscillate, but dampen back to the fixed position.

[0037] In order to obtain three-dimensional images, the source 30 and the receiver 28 are rotated in a plane perpendicular to the axis of the lower bearing 20 i.e. in the X-Y plane. Elongated volume scans can also be achieved by adding relative motion along the Z axis. Accurate three-dimensional reconstruction requires an accurate knowledge of the x-ray path through the examination region E and a distance 32 between the receiver 28 and the source 30, more particularly a distance 34 between the source 30 and the region of interest and a distance 36 between the region of interest and the receiver 28.

[0038] In a preferred embodiment, the receiver 28 and the x-ray source 30 both include three-dimensional Micro-Electro-Mechanical (MEMS) accelerometers 40, 42. MEMS accelerometers are known and commercially manufactured. The accelerometers 40, 42 are positioned ideally along a central axis 44 passing through a geometric center of the receiver 28 and a center of a focal spot on the source 30 for mathematical convenience. However, it is appreciated that the accelerometers could also be located at other known positions within the housings of the receiver and the source offset from the central axis. Moreover, in addition to MEMS devices, use of other accelerometers or accurate position sensing devices is also envisioned by the present invention such as gyroscopes, inertial sensors, and the like.

[0039] Now referring to FIG. 2, the accelerometers 40, 42 each generate electronic acceleration data indicative of acceleration along volumetric coordinates. Integrators 50, 52 perform a double integral on the acceleration data along each of the coordinates to determine cumulative displacement relative to each coordinate from an initial starting point. At each data sampling point, the integrators 50, 52 are sampled to provide the deflection error corresponding to each view of image data. It is to be appreciated that, the accelerometers and integrators can provide the displacement data in rectangular coordinates, polar coordinates, or the like. Further, depending on the motion direction and the construction of the C-arm, one can choose to determine the displacement along only one or two coordinates.

[0040] A position encoder 58 determines the initial starting position preferably in the same coordinates. A coordinate processor 60 calculates the orientation of the central axis 44, the source and receiver positions for each sampling position from the starting position and the cumulative displacements. If the integrators are zeroed at the starting position, actual position can be calculated by simply adding the cumulative displacement to the starting position. Optionally, the coordinate processor 60 uses the output of the encoder 58 to determine an offset error, skew error or the like at each view sampling position. Either way, the output indicates the actual trajectory of each ray of each view through the examination region.

[0041] An image reconstruction processor 62 applies known reconstruction algorithms for cone beam data to each sampled view. Each view is convolved or otherwise processed and then back projected into a volume image memory 64. Conventional reconstruction algorithms assume that each view is collected at the selected sampling position and represents radiation attenuation along rays (shown in solid) between the source 30 and the receiver 28 as shown in FIG. 3. Rather than projecting the data along the expected rays, the reconstruction processor 62 uses the actual position of the source 30′ and the receiver 28′ to project the data along the actual rays (shown in phantom). Depending on the exact algorithm chosen, the ray trajectories are corrected by rotational offsets, angular offsets, magnification correction, redefining the ray trajectories, and the like.

[0042] The volume image is reconstructed, built, and stored as voxels in the volume memory 64. An operator control panel 66 enables the operator to select various image representations, such as slice images, slab images, volume images, and the like to be displayed. A video processor 68 samples the appropriate voxels in the volume memory and converts the data to approximate form for display on a monitor 70.

[0043] Referring to FIG. 4, the x-ray source 30 includes a collimator 72 which projects the x-ray beam through the examination region E onto the x-ray receiver 28 with little or no overscan. When the receiver and source deflect by different amounts, the source and receiver may become misaligned. To correct the misalignment, a misalignment processor 74 calculates a misalignment correction and controls a servomotor 76 to reposition the collimator 72 thus realigning the x-ray beam onto the receiver.

[0044] The use of three-axis accelerometers on the position sensitive imaging components can result in an overall reduction in system cost. Less massive structures can be used because precise positional information is readily available even when the structures deflect. The invention thus has the advantage that a highly versatile positioning structure, such as a fluoroscopic C-arm or other devices that are subject to mechanical distortion, can be used for a high precision image procedures, such as volume image reconstructions. Moreover, accelerometer use can also result in increased system availability. Time consuming calibration and re-calibration runs to correct for wear and aging are no longer required since position correcting data is developed on-the-fly during the diagnostic imaging procedure.

[0045] Now referring to FIG. 5, an independent movable radiation source 80 and receiver 82 can be utilized. In other words, the source and the receiver are not connected to a common frame. In a preferred embodiment, the source 80 is mounted to the floor by a mechanical support structure 84 capable of inducing motion. The receiver is attached to a similar three degrees of movement structure 86 suspended from the ceiling. MEMS accelerometers 88, 90, as discussed above with reference to FIG. 1, are mounted to the source and the receiver. Preferably, a fixed repeatable docking position is defined as a reference position for zeroing the accelerometers.

[0046] Referring now to FIG. 6, an imaging apparatus includes a mechanical structure M₂supporting a radiation source 100 and a radiation receiver 102 surrounding an isocenter 104. A mechanical actuator 106 is preferably provided to vary the height of the isocenter 104. Those skilled in the art will recognize that activating the actuator 106 will induce an oscillatory motion in the mechanical structure M₂, particularly in the cantilevered arms and pivots that support the mechanical actuator from the ceiling. Motion of the isocenter 104 due to the oscillation is best appreciated by reference to FIG. 7.

[0047] Referring back to FIG. 6, the apparatus also includes a sensor 108 for detecting motion, especially vibration, in the mechanical structure M₂. Sensor 108 additionally generates a signal indicative of the motion sensed for processing by processor 110 (FIG. 8). In one preferred embodiment the sensor 108 is configured as an accelerometer which detects motion or vibration in at least one axis, particularly the vertical movement. Alternately, the sensor includes a strain gauge which senses and signals deflection.

[0048] Referring now to FIG. 8, the processor 110 preferably compares the sensed or detected oscillation against a database 112 of previously calibrated and stored data. From the data, the processor 110 generates a cancellation signal, preferably an equal amplitude, out of phase oscillation. The processor 110 then drives the actuator 106 to impart the cancellation movement to the mechanical structure M₂. More specifically to the preferred embodiment, as the sensor senses that the C-arm assembly is starting to move downward, the cancellation signal causes the actuator 106 to contract pulling the C-arm assembly up. Conversely, when the C-arm assembly starts to move upward, the actuator 106 extends creating a downward motion. By matching the magnitudes of the extensions and contractions of the actuator to the magnitude of the vibrations, the C-arm assembly is held stationary as the upper arms oscillate.

[0049] Those skilled in the art will recognize that actuator 106 can comprise existing actuators used to position the imaging apparatus. Alternately, piezo-electric or other rapidly responding actuators can be combined with the mechanical actuator. For example, the rapidly responding actuator can be mounted between the mechanical actuator 106 and the C-arm assembly. Also, while the description above illustrates sensing and correcting oscillations in only one plane, the present invention envisions detection and cancellation of oscillatory motion in all three axes. It will be noted that as the ceiling arms oscillate, the C-arm assembly tends to move not only downward (or upward) but also tip or lean downward (or cant upward). The rapidly responding actuators are advantageously positioned fore and aft the interconnection between the actuator 106 and the C-arm assembly to counteract the cyclic tipping.

[0050] Alternately, an additional motion sensor(s) is mounted to the ceiling arms to monitor their motion directly. As yet another alternative, motion sensors are mounted to the detector 102 and the x-ray tube 100. Rapidly responding actuators are positioned at the detector and the x-ray tube to counteract any vibratory or other degrading motion that reaches the detector and the x-ray tube.

[0051] The invention has been described with reference to the preferred embodiments. Potential modifications and alterations will occur to others upon a reading and understanding of the specification. For example, accelerometers could also be employed in other medical imaging applications such as fluoroscopy, angiography, ultrasound, etc. It is our intention to include all such modifications and alterations insofar as they come within the scope of the appended claims, or the equivalents thereof.