Collimator with variable focusing and direction of view for nuclear medicine imaging

[0001] 1. Field of the Invention

[0002] The present invention generally relates to nuclear medicine, and systems for obtaining nuclear medical images of a patient's body organs of interest. In particular, the present invention relates to a novel collimator with variable focusing and direction of view for nuclear medicine imaging, particularly for single photon imaging including single photon emission computed tomography (SPECT).

[0003] 2. Description of the Background Art

[0004] Nuclear medicine is a unique medical specialty wherein radiation is used to acquire images that show the function and anatomy of organs, bones or tissues of the body. Radio pharmaceuticals are introduced into the body, either by injection or ingestion, and are attracted to specific organs, bones or tissues of interest. Such radio pharmaceuticals produce gamma photon emissions that emanate from the body. One or more detectors are used to detect the emitted gamma photons, and the information collected from the detector(s) is processed to calculate the position of origin of the emitted photon from the source (i.e., the body organ or tissue under study). The accumulation of a large number of emitted gamma positions allows an image of the organ or tissue under study to be displayed.

[0005] Single photon imaging, either planar or SPECT, relies on the use of a collimator placed between the source and a scintillation crystal or solid state detector, to allow only gamma rays aligned with the holes of the collimator to pass through to the detector, thus inferring the line on which the gamma emission is assumed to have occurred. Single photon imaging techniques require gamma ray detectors that calculate and store both the position of the detected gamma ray and its energy.

[0006] Two principal types of collimators have been used in nuclear medical imaging. The predominant type of collimation is the parallel-hole collimator. This type of collimator contains hundreds of parallel holes, which can be formed by casting, drilling, or etching of a very dense material such as lead. Parallel-hole collimators are most commonly attached near the detector (scintillator) with holes arranged perpendicular to its surface. Consequently, the camera detects only photons traveling nearly perpendicular to the scintillator surface, and produces a planar image of the same size as the source object. In general, the resolution of the parallel-hole collimator increases as the holes are made smaller in diameter and longer in length. The parallel-hole collimator offers greater sensitivity than a pinhole collimator, and its sensitivity does not depend on how closely centered the object is to the detector.

[0007] The conventional pinhole collimator typically is cone-shaped and has a single small hole drilled in the center of the collimator material. The pinhole collimator generates a magnified image of an object in accordance with its acceptance angle, and is primarily used in studying small organs such as the thyroid or localized objects such as a joint. The pinhole collimator must be placed at a very small distance from the object being imaged in order to achieve acceptable image quality. The pinhole collimator offers the benefit of high magnification of a single object, but loses resolution and sensitivity as the field of view (FOV) gets wider and the object is farther away from the pinhole.

[0008] Other known types of collimators include converging and diverging collimators. The converging collimator has holes that are not parallel; rather, the holes are focused toward the organ with the focal point being located in the center of the FOV. The image appears larger at the face of the scintillator using a converging collimator. For equivalent spatial resolution the converging collimator has higher sensitivity than the parallel-hole collimator. The gain in point sensitivity is obtained at the price of a reduced FOV. The diverging collimator results by reversing the direction of the converging collimator. The diverging collimator is typically used to enlarge the FOV, such as would be necessary with a portable camera having a small scintillator. The diverging collimator has a lower sensitivity than the parallel-hole collimator, especially with thick objects.

[0009] Another type of collimator is slat collimator that has been used with a rotating laminar emission camera, also known as the rotating laminar radionuclide camera. This camera has linear collimators usually formed by mounting parallel collimating plates or slats between a line of individual detectors. Alternately, individual detector areas of a large-area detector are defined and isolated through the placement of slats. The slat collimator isolates planar spatial projections; whereas, the grid collimator of traditional scintillation detectors isolates essentially linear spatial projections. The detector-collimator assembly of a slat camera is typically rotated about an axis perpendicular to the detector face in order to resolve data for accurate two-dimensional image projection. The projection data collected at angular orientations around the subject are reconstructed into a three-dimensional volume image representation.

[0010] While maintaining certain advantages, such as a better sensitivity-resolution compromise, over, e.g., traditional Anger cameras, slat detectors are burdened by some other undesirable limitations. For example, the one dimensional collimation or slat geometry used by slat detectors complicates the image reconstruction process. The slat geometry results in a plane integral reconstruction as opposed to the line integral reconstruction that is generally encountered in traditional Anger camera applications. Moreover, the geometry produces a plane integral only in a first approximation.

[0011] It is well known in the art that nuclear medicine imaging of small organs, such as brain, heart, kidneys, thyroid, and the like present special problems in collecting radiation emission and creating images from the collected data. Different systems including the use of the above described collimators have been used for nuclear imaging of small organs. Although images of such organs are routinely made, there remains a need for a system and methodology for improving imaging of small organs and for overcoming the shortcomings of the prior art, such as a novel collimator for a nuclear imaging camera and a method of forming the same.

[0012] The present invention solves the existing need by providing a new collimator geometry that enhances the imaging of small organs with high resolution or in an efficient manner. According to the present invention, a novel slat collimator for use in nuclear medicine imaging is provided. The slat collimator comprises a first layer comprising a plurality of spaced apart elongated slats and a second layer comprising a plurality of spaced apart elongated slats. The slats of the second layer are positioned orthogonally with respect to the slats of the first layer. The slats are constructed of a radiation attenuation material, such as tantalum, tungsten, lead and the like.

[0013] In one embodiment, the collimator is a static, i.e., the spaces between the slats are not variable. The spaces can be fixed by several means, including foam, grooves in the slats and guide plates.

[0014] In a second embodiment, the collimator is variable, i.e., the spaces between the slat can be varied. The spaces can be varied through several means, including springs, air bubbles and magnetic force. Pressure can be differentially applied to one end of a slat layer to control the pointing direction of the slats.

[0024] The present invention is directed to a slat collimator that comprises two layers of slats. The present invention also describes a method of collimator fabrication using two stacks of slats. Also described are various techniques by which the angles of the slats can be varied to create non-parallel beam collimators. Such collimators may be advantageous in SPECT studies of small organs, such as brain, heart, kidney, thyroid, etc. The convergence of the collimator can be changed to adapt for each study. Also, the convergence can be changed in a SPECT study as the distance from the camera to the organ changes during the scan. For a given spatial resolution, the gain of sensitivity with 2D-convergence will dominate the small sensitivity loss due to the extra collimator thickness relative to a conventional hole collimator. The slat collimator of the present invention is used on a scintillation camera of the type which is used to carry out SPECT studies, i.e., is used with a nuclear imaging acquisition system for SPECT studies. The nuclear imaging acquisition system comprises the slat collimator described herein and a detector having a side which detects radiation emanating from an object after passing through said collimator.

[0025] As shown in FIG. 1, a collimator in accordance with the present invention comprises two stacks of slats. The collimator comprises a first layer (100) of a plurality of elongated spaced apart slats (101a, 101b, . . . 101n) and a second layer (200) of a plurality of elongated spaced apart slats (201a, 201b, . . . 201n). The second layer (200) is positioned orthogonally with respect to said first layer (100). The slat material should be a suitable gamma ray attenuator, e.g., tantalum, tungsten, lead, etc. The slats (101a, 101b, . . . 101n) of the first layer (100) may be perpendicular to the surface of detector (not shown) or they may be at an angle greater than zero. All of the slats (101a, 101b, . . . 101n) in the first layer (100) are angled in the same direction. Similarly, the slats (201a, 201b, . . . 201n) of the second layer (200) may be perpendicular to the surface of the first layer (100) or they may be at an angle greater than zero. All of the slats (201a, 201b, . . . 201n) in the second layer (200) are angled in the same direction. By constructing collimators from two orthogonal layers of slats, similar to “Venetian blinds”, a very general collimation viewing configuration is realized.

[0026] In one embodiment, the spaces (102a, 102b, . . . 102n) between the slats (101a, 101b, . . . 101n) in the first layer (100) are non-variable, i.e., fixed to produce static (non-variable) collimation. Similarly, the spaces (202a, 202b, . . . 202n) between the slats (201a, 201b, . . . 201n) in the second layer (200) are non-variable, i.e., fixed.

[0027] In one aspect of this embodiment, the spaces (102a, 102b, . . . 102n; 202a, 202b, . . . 202n) between the slats (101a, 101b, . . . 101n; 201a, 201b, . . . 201n) can be filled with a low density foam materials, such as ROHACELL® rigid plastic foam material.

[0028] In a second aspect, air spaces between slats could be used if slats are sufficiently rigid. The spacing of the slats (101a, 101b, . . . 101n; 201a, 201b, . . . 201n) can be fixed by mounting the slats into grooves (not shown) on the top edge of the slats (101a, 101b, . . . 101n) of the first layer (100) and on the bottom edge of the slats (201a, 201b, . . . 201n) of the second layer (200).

[0029] In a third aspect, the spacing of the slats (101a, 101b, . . . 101n; 201a, 201b, . . . 201n) can be fixed by mounting the slats into grooves of slide guide plates. FIG. 2 is a cross-section view of the construction of one layer, e.g., first layer (100) using slide guide plates. As shown in FIG. 2, two slide guide plates (103a, 103b) are provided. The slide guide plates (103a, 103b) have grooves (104a, 104b . . . 104n; 105a, 105b, . . . 105n) on their inside edges into which the slats (101a, 101b, . . . 101n) are positioned. The slide guide plates (103a, 103b) are constructed out of low radiation attenuation material, such as aluminum or plastic. It can be appreciated that the spaces between slats (201a, 201b, . . . 201n) of the second layer (200) can be fixed in the same manner.

[0030] In a second embodiment the spaces (102a, 102b, . . . 102n) between the slats in the first layer (100) can be varied. Similarly, the spaces (202a, 202b, . . . 202n) between the slats in the second layer (200) can be varied. As used herein, variable spaces is intended to mean that the distance between the slats (101a, 101b, . . . 101n; 201a, 201b, . . . 201n) at one end of said slates is less than the distance between slats at the other end of said slats. By varying the spaces between the slats in this manner, non-parallel beam collimators are created. In addition, the direction of view can be changed using such a variable collimator. In order for direction of view to be changed in a general manner, a means creating repulsive forces between the slats needs to be created.

[0031] In one aspect of this embodiment, the slats (101a, 101b, . . . 101n; 201a, 201b, . . . 201n) can be held apart by springs. As shown in FIG. 3A, the slats (101a, 101b, . . . 101n) are held apart by springs (106a, 106b, 106c) at one end of the first layer (100) and springs (107a, 107b, 107c) at the other end of the first layer (100). The pointing direction of the slats (101a, 101b, . . . 101n) may be controlled by setting the orientation of slats at either end of the layer, such as by using springs of different sizes or by applying a force at either end of the layer. For example, springs (107a, 107b, 107c) may be larger than springs (106a, 106b, 106c) such that a direction of collimation of radiation is achieved. Alternatively, as shown in FIG. 3B, single springs (108a, 108b, . . . 108n) can be used between the slats (101a, 101b, . . . 101n). By applying a force to one end of the slats (101a, 101b, . . . 101n) in the first layer (100), direct directional control of the source slats between the ends of the array can be provided. As shown in FIG. 3B, such force can be applied by motors (109a, 109b) that push plates (110a, 110b) into one end of the slats (101a, 101b, . . . 101n) to provide directional control. In addition, by tilting the end slats the direction of view of the array can be deflected or focused. It can be appreciated that springs and similar direct directional control can be performed for the slats (201a, 201b, . . . 201n) of the second layer (200). It can further be appreciated that the directional control can be applied to only one or both of the layers of the slats.

[0032] In a second aspect, slats could be held apart by a bubble-wrap between the slats. As shown in FIG. 4, slats (101a, 101b, . . . 101n) of the first layer (100) are separated by plastic (111a, 111b, . . . 111n) that contains bubbles (112a, 112b, . . . 112n) of air. By tilting the end slats, such as described above and as shown by the arrows in FIG. 4, the direction of view of the array can be deflected or focused. It can be appreciated that bubble-wrap and similar direct directional control can be performed for the slats (201a, 201b, . . . 201n) of the second layer (200). It can further be appreciated that the directional control can be applied to only one or both of the layers of the slats.

[0033] In a third aspect, slats (101a, 101b, . . . 101n) of the first layer (100) could be held apart magnetically. As shown in FIG. 5, each slat (101a, 101b . . . 101c) is encompassed by a current loop. The loop wires (not shown) are attached to the slats (101a, 101b, . . . 101n). Alternate slats (e.g., 101a and 101b) have the current flowing in the opposite sense, as shown by the + and − in FIG. 5. The opposite current flow sets up a repulsive magnetic force between the slats. By tilting the end slats, such as described above, the direction of view of the array can be deflected or focused. It can be appreciated that magnetic repulsion and similar direct directional control can be performed for the slats (201a, 201b, . . . 201n) of the second layer (200). It can further be appreciated that the directional control can be applied to only one or both of the layers of the slats.

[0034] In a third embodiment, the spaces in one layer, e.g., spaces (102a, 102b, . . . 102n) between the slats in the first layer (100) are non-variable, i.e., fixed, such as described above. The spaces in a second layer, e.g. spaces (202a, 202b, . . . 202n) between the slats in the second layer (200) can be varied and under direct directional control, such as described above. Alternatively, the spaces (102a, 102b, . . . 102n) between the slats in the first layer (100) can be varied and under direct directional control. The spaces in a second layer, e.g. spaces (202a, 202b, . . . 202n) between the slats in the second layer (200) are non-variable, i.e., fixed.

[0035] By varying the spaces between the slats in this manner, non-parallel beam collimators are created. Such collimators may be advantageous in SPECT studies of small organs, such as brain, heart, kidney, thyroid, etc. The convergence of the collimator can be changed to adapt for each study. Also, the convergence can be changed in a SPECT study as the distance from the camera to the organ changes during the scan. To image a small organ (or region-of-interest), it is desireable to spend a greater share of the available scan time and a greater share of the available detector area detecting photons mainly from this area. An initial fast SPECT scan (or the use of two orthogonal views) would give enough information to allow the position of the organ-of-interest (ROI) to be determined. Using this position information, the collimator can be dynamically focused on the ROI during the scan for a large fraction of the total study time.

[0036] Although the slat collimation system of the present invention has some drawbacks, particularly in static configuration, in comparison to foil or cast collimator, it has several advantages. The drawbacks include:

• • [0037] The static system will be thicker (at least double) than a conventional collimator. Thus, for a given special resolution, the sensitivity will be somewhat reduced, approximately by the square of the ratio of distances from source to detector (scintillation crystal).
  • [0038] It is made out of more costly materials.
  • [0039] It has more complex control and calibration.

[0040] The advantages include:

• • [0041] For SPECT imaging of organs or regions significantly smaller than the typical camera field of view (FOV), the variable slat system can yield overall improvements in imaging speed (higher sensitivity). As shown in FIGS. 6A and 6B, the SPECT acquisition change have two phases of differing durations, T₁ and T₂. For imaging a small organ it will be advantageous to dynamically focus on the organ-of-interest for time T₂ and image the entire object (no truncation) for another time period T₁. Generally, T₂ is much greater than T₁, since the untruncated data is only needed to form the image of the organ surround at lower resolution. A SPECT acquisition commonly consists of a multiplicity of different views. Each view is defined by a specification camera position of orientation. The i-th view may have focused and unfocused temporal phases T_2i and T_1i.
  • [0042] The system can focus collimators so that more (most) of the time is spent acquiring counts from the organ or region of interest and less time spent acquiring counts from the overall background.
  • [0043] Focus can controlled to provide tight focusing on organ of interest without truncation. The focus could also be offset with respect to the center of the collimator. The use of quick prescan SPECT study, perhaps only two orthogonal planar views can suffice in many cases, allows the organ of interest to be located. Position encoders on the camera system give the position and angular orientation of each camera head (detector). Using this information together of the prescan data permits determination of the organ position for tight, dynamic focusing on the organ for the remainder of the scan. The focus of the slat collimator does not necessary have to be centered, but can be offset. This can be achieved by means of non-symmetric orientation and drive of the push plates (110a, 110b), see FIG. 3B.

[0044] As disclosed above, the variable slat system can yield overall improvements in sensitivity. There are several sensitivity considerations that can be envisioned. The sensitivity (solid angle) of a convention 2D-hole collimator is given by the equation

[0045] $\begin{array}{cc}\Omega ={\left[\frac{{\mathrm{kD}}^2}{L\left(D+S\right)}\right]}^2& \left(\mathrm{Eq}.1\right)\end{array}$
where k is a form factor depending on hold shape, D is the size of the hole (˜across “flats” dimension), S is septal thickness and L is light. (Anger, H. O. (1964), “Scintillation Camera with Multichannel Collimators.” J Nucl Med 5:515-531.) FIG. 7A shows D and k for a square hole. FIG. 7B shows D and k for a hexagonal hole.

[0046] In a stack slat collimator system, the main factor degrading sensitivity for a fixed spatial resolution will be the increased distance of the object from the camera due to the increased thickness of the collimator. The sensitivity of the stacked slat (shown representatively in FIG. 7C in which stack 1 is layer (100) and stack 2 is layer (200) will be approximately

[0047] $\begin{array}{cc}\Omega ={\theta }_1{\theta }_2=\left[\frac{k_1{D}_1^2}{L_1\left(D_1+S_1\right)}\right]\left[\frac{k_2{D}_2^2}{L_2\left(D_2+S_2\right)}\right]& \left(\mathrm{Eq}.2\right)\end{array}$
where k₁≈k₂=√{square root over (k)}

[0048] The angular sorting in x and y direction is separable and for L₁=L₂=L, D₁=D₂=D, S₁=S₂=S, the net solid angle of the stack collimator is approximately the same for a conventional collimator given by Eq. 1.

[0049] For a fixed collimator spatial resolution, the collimator angle (shown in FIG. 8 as θ) is

[0050] $\begin{array}{cc}\theta =\frac{R_c}{\left(a+b+c\right)}& \left(\mathrm{Eq}.3\right)\end{array}$
where a is the distance from the mean detection plane (in the scintillation crystal) (800) to the collimator (801), b is thickness of the collimator (801), and c is the distance from the collimator (801) to the object. R_c is the geometric spatial resolution of the collimator.

[0051] If R_c, a and c are held fixed, then an increase of b implies a decrease in θ and a decrease in sensitivity.

• [0052] Ω₁≈θ² for conventional collimators
• [0053] Ω₂=θ_xθ_y for stacked (layers of slats) collimator

[0054] $\left(\frac{{\Omega }_2}{{\Omega }_1}\right)=\frac{{\theta }_x{\theta }_y}{{\theta }^2}={\left(\frac{a+b_1+c}{a+b_2+c}\right)}^2$
For typical values: a=0.5 cm, c=20 cm, b₁=2.5 cm, b₂>2b₁ or b₂=2b₁ (best case)

[0055] $\left(\frac{{\Omega }_2}{{\Omega }_1}\right)={\left(\frac{23}{25.5}\right)}^2=0.81$
For c=16 cm (reasonable for brain imaging)

[0056] $\left(\frac{{\Omega }_2}{{\Omega }_1}\right)={\left(\frac{19}{21.5}\right)}^2\approx 0.78$
Thus, with a 2D-converging slat system according to the present invention, a magnification gain >2 could easily be obtained. Hence, the small loss of sensitivity due to increased thickness of the collimator is more than offset by the gain in magnification.

[0057] For SPECT imaging of organs or regions significantly smaller than the typical camera field of view (FOV), the variable slat system can yield overall improvements in imaging speed (higher sensitivity). As shown in FIGS. 6A and 6B, the SPECT acquisition can have two phases of differing durations, T₁ and T₂. For imaging a small organ it will be advantageous to dynamically focus on the organ-of-interest for time T₂ and image the entire object (no truncation) for another time period T₁. Generally, T₂ is much greater than T₁, since the untruncated data is only needed to form the image of the organ surround at lower resolution. A SPECT acquisition commonly consists of a multiplicity of different views. Each view is defined by a specification camera position of orientation. The i-th view may have focused and unfocused temporal phases T_2i and T_1i.

[0058] While a preferred embodiment of the present invention has been described above, it should be understood that it has been presented by way of example only, and not limitation. Thus, the breadth and scope of the present invention should not be limited by the above described exemplary embodiment.

[0059] Obviously, numerous modifications and variations of the present invention are possible in light of the above teachings. It is therefore to be understood that the invention may be practiced otherwise than as specifically described herein.

[0060] As disclosed above, the variable slat system can yield overall improvements in sensitivity. There are several sensitivity considerations that can be envisioned. The sensitivity (solid angle) of a convention 2D-hole collimator is given by the equation

[0061] $\begin{array}{cc}\Omega ={\left[\frac{{\mathrm{kD}}^2}{L\left(D+S\right)}\right]}^2& \left(\mathrm{Eq}.1\right)\end{array}$
where k is a form factor depending on hole shape, D is the size of the hole (˜across “flats” dimension), S is septal thickness and L is light. (Anger, H. O. (1964), “Scintillation Camera with Multichannel Collimators.” J NucI Med 5:515-531.) FIG. 7A shows D and k for a square hole. FIG. 7B shows D and k for a hexagonal hole.