JPH1172568A - Method for semiautomatically setting concerning region of kidney in kidney scintigraphy - Google Patents

Abstract

PROBLEM TO BE SOLVED: To draw the image of even a kidney with low function with good reproducibility between operators. SOLUTION: A pixel near the center of each kidney on an image is clicked. Maximum pixel value included in the first region of a plurality of pixels when the center is the clicked pixel is obtained and made the maximum pixel value. A group of pixels having pixel values over the first threshold value determined with reference to the maximum pixel value in the second region of a plurality of pixels larger than the first region of a plurality of pixels when the center is the clicked pixel is regarded as the kidney center region. When there is a location where two regions in the kidney center region are continuous with only one pixel, the pixel is deleted, the pixel group consisting only of pixels having continuity with the clicked pixel inside the pixels in the kidney center region is newly made the kidney center region and the pixel having higher pixel value than the second threshold value determined with reference to the maximum value in the pixel which wholly or partly comes in the edge region with a width of 5 to 20 mm outside the kidney center region is added to the kidney center region to be a kidney concerning region.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

TECHNICAL FIELD The present invention relates to the ratio of renal accumulation to the dose of radiopharmaceutical (% injected) in renal function determination using renal scintigraphy, which is one of nuclear medicine tests.
The present invention relates to a method for setting a region where a kidney is present (a region of interest in the kidney) by a scintigram displayed on a computer monitor when obtaining a dose (% ID).

[0002]

2. Description of the Related Art Renal scintigraphy, which is one of the diagnostic imaging methods, is classified into static renal scintigraphy and dynamic renal scintigraphy. The former is a test mainly taking a morphological diagnosis of the kidney, such as taking a still image after a certain period of intravenous drug injection, taking the position, size, shape and radioactivity distribution of the renal parenchyma of the kidney,
Kinetic scintigraphy is a method aimed at diagnosing the filtering function and excretion function in addition to the morphology of the kidney by taking continuous images every few seconds immediately after the intravenous injection of the radiopharmaceutical.

The renal function quantification method by dynamic renal scintigraphy (gamma camera method) is an extremely simple and excellent test method which does not require blood sampling and urine collection and can measure renal function, and is currently used in many nuclear medicine facilities. Used.

Already in the above dynamic renal scintigraphy, technetium diethylenetriaminepentaacetate (abbreviation: 99mTc-DTPA), sodium iodide hippurate (abbreviation: 131I-OIH) and mercaptoacetyltriglycine technetium (abbreviation: 99mTc-MAG3)
Methods for measuring the glomerular filtration rate, effective renal plasma flow rate, and 99mTc-MAG3 clearance using the renal uptake rate of radiopharmaceuticals and the like have been developed.

[0005] It is said that many factors affect the above-mentioned renal function quantification, and many studies have been made on the improvement of test methods. For example, influencing factors include omission of injection of a drug, attenuation of radiation by a table, deviation of the start time between injection and imaging, background correction, renal depth correction (absorption correction), and the like. The optimal test method for 99mTc-MAG3 is described in the publication J Nu by Taylor et al.
cl Med 1995; 36: 1689-1695.

In a nuclear medicine image, a region where a target organ to be subjected to a quantitative or qualitative analysis exists is called a region of interest (ROI). However, most of renal function quantification by dynamic renal scintigraphy is performed. The commercially available analysis software is 99mTc-DTPA, 131I-OI
H, 99mTc-MAG3, etc., the ratio of renal accumulation at 1 to 2 minutes, 2 to 3 minutes, etc. after administration (percen
(injected dose;% ID), and the reproducibility between operators in the setting of the region of interest for the right and left kidneys becomes a problem.

In the end, the conventional method for setting a renal area of interest displays the collected image data on a computer monitor,
This was done by an operator manually using a trackball, light pen or mouse to outline the outline of the right and left kidneys on the monitor screen. Renal accumulation is defined as the total count within the renal area of interest. Therefore, a significant difference may occur in% ID depending on how the operator sets the regions of interest in the right and left kidneys.

In this regard, the publication J Nucl Med 1
996; 37 (suppl): page 293, Itoh
For example, it has been reported that if a standard for setting a region of interest is set in advance between operators, no significant difference occurs in% ID. On the other hand, Halker et al.
Point out that a potential error in calculating relative and absolute renal uptake of mTc-MAG3 lies in inter-operator variability in renal region of interest setting, resulting in significant differences between% IDs by the two technicians. It is reported that reproducibility between operators could be improved by using a semi-automatic renal region of interest setting method using a one-step threshold method.

The outline of the one-step threshold method of Halker et al. Is as follows. A large ellipse including the right and left kidneys is set on a scintigram displayed on a computer monitor (only this point is manually set) and is included in the ellipse area. In this method, a group of pixels having a pixel value equal to or more than a predetermined threshold value (for example, 20%) based on the maximum pixel value of the pixel is set as a region of interest of the kidney. Note that the pixel value on the scintigram indicates a numerical value proportional to the radiation dose at that position.

[0010]

However, in the conventional method of setting a region of interest by manually designating a kidney contour as in the conventional method, no matter how much a standard for setting a region of interest is previously determined among operators. Can occur. In particular, the renal function is remarkably deteriorated, the accumulation is poor, and it is difficult to set a region of interest when the organ is close to other organs, and a significant difference between operators tends to increase.

In addition, inexperienced operators who have not been trained sufficiently tend to actually evaluate the kidneys small. This is because the conventional region of interest setting software calculates the total pixel value of only the pixels completely included in the handwritten region. If this is not fully understood, the count of the margins of the kidney will be reduced, the% ID will be lower than it actually is, and the renal function will be underestimated.

In a semi-automatic renal region of interest setting method using a one-step threshold method, such as Halker, which has solved the above-mentioned problem of variation between operators, a pixel larger than the renal maximum pixel value is set within a large elliptical region that has been set. When there is a pixel having a value, there is a problem that the renal region of interest is set small. Further, a low threshold value must be set so as to include the marginal region of the kidney. However, in many cases, the adjacent organ cannot be separated from the kidney, and there is still a problem that the rendering ability is low.

The present invention has been made in view of the above circumstances, and makes it possible to improve reproducibility among operators and to render a poorly functioning kidney in setting a renal region of interest in renal scintigraphy. And a novel semi-automatic method for setting a renal region of interest.

[0014]

According to the present invention, there is provided: (1) radioactivity released from a radiopharmaceutical for renal scintigraphy administered to a living body is collected by a gamma camera;
The obtained image data (scintigram) is transferred to a computer, the digital image is displayed on a computer monitor, and the region of interest in the right and left kidneys is semi-automatically set based on the pixel values. In the region setting method, first, a scintigram is displayed on a computer monitor, and pixels near the center of each of the right and left kidneys on the displayed image are each clicked. Second, the center is set at the clicked pixel. The maximum pixel value of the pixels included in the first plurality of pixel regions in the predetermined range is obtained, the maximum pixel value of the kidney is determined, and thirdly, from the first plurality of pixel regions centered on the clicked pixel. Within a second plurality of pixel regions having a predetermined size that is also greater than or equal to a first threshold value based on the maximum pixel value of the kidney. Fourth, a group of pixels having a pixel value of is defined as a renal central region. Fourth, when there is a place where two regions are continuous via only one pixel in the renal central region, the one pixel is defined as a renal central region. Fifthly, fifthly, a set of pixels consisting of only the pixels having continuity with the clicked pixel among the pixels in the renal center region is newly set as the renal center region, and sixth, the renal center region A pixel having a pixel value higher than a second threshold value based on the maximum pixel value among the pixels entirely or partially included in the outer 5-20 mm width peripheral region is defined as a kidney peripheral region, A semi-automatic renal region of interest setting method in renal scintigraphy, wherein the first to sixth processes are sequentially performed in addition to the renal central region as the region of interest of the kidney. To solve the above-mentioned problems.

(2) In the semi-automatic region-of-interest setting method in renal scintigraphy according to (1), the first plurality of pixel regions are set to have a horizontal size of 30 to 70 mm and a vertical size of 60 mm.
A semi-automatic kidney region of interest setting method in renal scintigraphy, characterized in that the second plurality of pixel regions have a width of 100 to 250 mm and a vertical size of 150 to 300 mm. Solve the problem.

(3) In the semi-automatic renal region of interest setting method for renal scintigraphy according to the above (1) or (2), the outside of the renal central region may be replaced by the fifth process instead of the sixth process.
Renal scintigraphy, wherein all pixels partially or entirely in a peripheral region having a width of 〜20 mm are defined as a renal peripheral region, and the renal scintigraphy is performed by adding the renal peripheral region to the renal central region to obtain a renal region of interest. The above problem is solved by providing a semi-automatic renal region of interest setting method.

(4) In the semi-automatic renal area setting method for renal scintigraphy according to the above (1) or (2), 60 mg after intravenous injection of mercaptoacetyltriglycine technetium (99mTc-MAG3) as a radiopharmaceutical.
The first threshold value in the added image (1-2 minute image) from seconds to 120 seconds is set to 50% to 70% of the maximum pixel value, and the second threshold value is set to 1% to 40% of the maximum pixel value. The object is achieved by providing a semi-automatic renal region of interest setting method in renal scintigraphy.

(5) In the semi-automatic renal region of interest setting method in renal scintigraphy according to the above (1) or (2), 12 hours after intravenous injection of mercaptoacetyltriglycine technetium (99mTc-MAG3) as a radiopharmaceutical.
The first threshold in the added image (2-3 minutes image) from 0 seconds to 180 seconds is a pixel value of 40% to 70% of the maximum pixel value, and the second threshold is 1% to 40% of the maximum pixel value.
The above object is achieved by providing a semi-automatic renal region of interest setting method in renal scintigraphy characterized by using pixel values of

(6) In the method of setting a region of interest in renal renography according to the above (1) or (2) or (3) or (4) or (5), except for the fourth process, The above object is achieved by providing a semi-automatic renal region of interest setting method in renal scintigraphy characterized by sequentially performing a fifth process and a sixth process following the third process.

In the first process in the above (1), the operator clicks on each of the positions considered to be the centers of the right and left kidneys on the scintigram displayed on the monitor (only this point is manual). Processing
The renal region of interest is set in a shorter time than the conventional method of setting the region of interest by manually designating the renal contour. Then, the region of interest of the left and right kidney is separated from the region of other organs and almost specified by the second to fifth series of processes, and the marginal region to be added is added by the sixth process, thereby avoiding underestimation. You.

[0021]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a semi-automatic renal region-of-interest setting method in renal scintigraphy according to the present invention will be described below in detail with reference to drawings showing clinical experiments and the results thereof.

FIG. 1 is a flow chart showing a processing procedure in a renal scintigram according to the present invention.

FIGS. 2 to 6 are views showing setting areas in the process of the semi-automatic kidney region of interest setting method according to the present invention in the 1-2 minute image of one subject displayed on the monitor.

FIG. 7 is a diagram for explaining a method of cutting continuity between two regions through one pixel in the central kidney region.

FIGS. 8 to 19 are diagrams showing the correlation of% ID under each condition.

In renal function quantification using renal scintigraphy, radioactivity released from a living body is collected by a gamma camera, and the obtained image data (scintigram) is transferred to a computer, and the obtained data is displayed on a computer monitor. Conventionally, a digital image is displayed, and the outline of the left and right kidneys is manually specified using a trackball, a mouse, a light pen, etc., and a set of pixels included in the outline is defined as a renal area of interest. The semi-automatic kidney region of interest setting method according to the present invention controls the use of a computer (hardware resources) including a processing device such as a CPU and a memory device, and generates a digital image displayed on a computer monitor. By simply clicking near the center of each kidney, the existence range of each kidney is determined based on the pixel value of the digital image, and it is set as the renal area of interest. And performs the computer processing based on the regulations of thought. The processing procedure is described below, and FIG. 1 shows an outline of the flow.

First processing: A scintigram is displayed on a computer monitor, and pixels near the center of each of the right and left kidneys on the displayed image are clicked.

Second processing: The maximum pixel value of the pixels included in the first plurality of pixel regions within a predetermined range centered on the clicked pixel is obtained and set as the maximum pixel value of the kidney.

Third processing: a second plurality having a predetermined size (for example, 12 times the first plurality of pixel regions) larger than the first plurality of pixel regions centered on the clicked pixel In the pixel region, a group of pixels having a pixel value equal to or more than a first threshold value based on the maximum pixel value of the kidney is defined as a kidney central region.

Fourth processing: when there is a portion where two regions are continuous only through one pixel in the central kidney region,
This one pixel is deleted from the central kidney region.

Fifth processing: Among the pixels in the renal central region, a set of pixels consisting only of pixels having continuity with the clicked pixel is newly set as a renal central region.

Sixth treatment: 5 to 2 outside the central kidney area
A pixel having a pixel value higher than a second threshold value based on the maximum pixel value among the pixels that are entirely or partially included in the 0 mm width peripheral region is defined as a kidney peripheral region, and this is defined as the kidney central region. In addition, the region of interest of the kidney is used.

In the first process, the operator simply clicks on a point (α point, β point) that is considered to be the center of the right and left kidneys in FIG. In addition, in order to store the correct maximum pixel value of the kidney, for example, if the size of each pixel is a square having a side of 9.3 mm, the first plurality of pixel regions P1 set in the second processing are horizontal and vertical. It is considered that a region of 5 × 9 pixels is appropriate. The size is preferably large enough to include the pixel having the maximum pixel value of the kidney and small enough not to include the pixel having the large pixel value of the peripheral organ. The reason is that a range of about 70 mm and a vertical dimension of 60 to 120 mm is appropriate.

In the third process, as shown in FIG. 3, a second plurality of pixel regions P2 (regions surrounded by broken lines) having a size sufficient to include the kidneys are set. Therefore, the second multiple-pixel region P2 may be a whole-body region, but in practice, when setting a renal region of interest, a size that is larger than the first multiple-pixel region P1 and within 15 times will be sufficient. . Setting P2 larger than this has no meaning in setting the renal region of interest. Also,
Since the second plurality of pixel regions P2 are separately set for the left and right kidneys, there is no problem even if they overlap (the center portion is overlapped in FIG. 3).

A group of pixels having a pixel value equal to or more than a first threshold value (for example, 60% of the maximum pixel value) based on the maximum pixel value in the second plurality of pixel regions P2 is defined as a central kidney region M1 ( (A region surrounded by a solid line in FIG. 3). In other words, the region of interest is roughly set, ignoring the discrimination from other organs.

In the fourth process, as can be seen from a comparison between FIG. 3 and FIG. 4, when two regions have continuity via only one pixel in the central kidney region M1, the one pixel (crosslinked pixel) K) is removed from the central kidney region M1 to break the continuity between the two regions via the bridging pixel K. As a specific method of the fourth processing, for example, as shown in FIGS. 7A and 7B, an arbitrary three rows 3
In a column block, if “1” is continuously arranged in the center horizontal row or center vertical column, and there is a bit pattern in which “0” is located above, below, or right and left of the center “1”, A pixel corresponding to the center “1” is determined as a bridge pixel K, and is set to “0” so that the pixel is deleted from the renal center region and the continuity between the two regions is cut. Here, the bit of “0” in FIG. 7 is a pixel outside the central kidney region, the “1” bit is a pixel included in the central kidney region, and the empty bits are “0” and “1” regardless of inside and outside the central kidney region. A pixel which may be any one is shown. By this fourth process, the peripheral organ and the true renal center region are separated.

It should be noted that the case where the above-mentioned fourth processing is performed is rare, and only one case (the case shown in FIGS. 2 to 6) is used in the experiment of setting a renal region of interest for 59 kidneys described later. there were. Therefore, even if the fourth processing is omitted, the renal area of interest can be set normally without any problem except for the rare case (corresponding to claim 6).

In the fifth process, as shown in FIG. 5, the continuity between the first clicked pixel and the continuity (only those whose pixels touch each other at the side is determined to be continuity). A set of pixels consisting only of pixels having the same renal center region is newly defined as the renal central region M2, and the other renal central regions M1 are deleted. By this fifth process, the other organ is identified.

In the sixth process, since the first threshold is set high to separate the kidney from the peripheral organ, the kidney is underestimated as it is, so that, for example, the kidney contacts the periphery of the central kidney region M2. Pixel (size 9.3mm square)
Among the pixels, a pixel having a pixel value higher than a second threshold value based on the renal maximum pixel value (for example, a threshold value lower than the first threshold value as 20% of the renal maximum pixel value) is defined as a renal margin region R This is added to the renal central region M2 to be a region of interest S of the kidney. In the cases shown in FIGS. 2 to 6, as can be seen from FIG. 6, all the pixels in contact with the periphery of the renal center region M2 in both the right and left kidneys are higher than the second threshold value, and It is fleshed out to say M2.

As expected from the above case, since the pixels around the peripheral region around the renal central region M2 almost have a pixel value of 10 to 20% or more of the maximum pixel value of the kidney, these pixels are unconditionally rejected. The region of interest S may be included as the peripheral region R (corresponding to claim 3). That is, instead of the sixth processing, all the pixels that entirely or partially enter the peripheral region having a width of 5 to 20 mm outside the central kidney region are defined as the peripheral kidney region R, and this is defined as the central kidney region M2. In addition, a process for setting the region of interest S of the kidney is performed.

Next, an experiment comparing the reproducibility of the case of using the above-mentioned semi-automatic renal region of interest setting method in the renal scintigraphy according to the present invention with the conventional manual region of interest setting using% ID will be described.

(Experiment Details) Preconditions: 22 renal disease patients and 11 normal subjects (18 males and 15 females; 32 kidneys on the left and 30 kidneys on the right) were used as subjects. Age 18-83 years (average age 46 years), serum creatinine 0.5-3.5 mg / dl (average 1.0
mg / dl).

30 minutes before the test, drink 300 ml of water,
The subject was placed in a supine position, a bolus of 99mTc-MAG3 was administered from the cubital fossa, and a gamma camera equipped with a low-energy general-purpose collimator was used from the back side for 120 seconds at 10 seconds per frame.
Continuous data collection of frames was performed. The image matrix is 64 × 64, and the size of each pixel is 9.3 m in length and width.
Data was collected at m. The front syringe and the rear syringe were photographed for 10 seconds each, and the matrix size was 12 seconds.
8 × 128. The photo peak was 140 keV and the energy window was 10%. The collected image data was transferred to a workstation for image analysis.

A kidney region of interest is set in the added image (1-2 minute image) from 60 seconds to 120 seconds after intravenous injection and the added image (2-3 minutes image) from 120 seconds to 180 seconds after intravenous injection. The value obtained by dividing the total pixel value in the region of interest by the difference between the syringe counts before and after the time correction is multiplied by 100 to obtain% ID.

The first plural pixel region P1 in the renal region of interest setting method is a rectangular region of 5 × 9 pixels (46.5 mm × 83.7 mm), and the second plural pixel region P2 is a large region. A rectangular area of 21 × 27 pixels (195.3 mm in width × 251.1 mm in height) was set.

(Experiment 1) To evaluate reproducibility among operators, three operators with different nuclear medicine histories (A: doctor with 3 years experience, B: doctor with 20 years experience, C: engineer with 1 year experience) Manually encircled the kidney outline in a 1-2 minute image to set a region of interest. At this time, using a monochrome color map of 256 gradations, the upper level of the display window is set to the maximum pixel value of each image, and the lower level is set to the upper level.
The image was displayed at a setting of 10%. This display condition is called monochrome display.

(Experiment 2) The operator A manually set a region of interest in the 1-2 minute image again in monochrome display, and evaluated reproducibility within the same operator as compared with the first time.

(Experiment 3) To evaluate whether the difference in display conditions affects% ID, the operator A converts the 1-2 minute image into 256 gradation colors (black, blue, red, yellow, and white gradations). ) Colormap, and set the upper level of the display window to the maximum pixel value of each kidney, lower lev
el was set to 10% of the upper level and the region of interest was manually set again. This display condition is called color display.

(Experiment 4) The operator A uses the semi-automatic renal region of interest setting method according to the present invention, and
Various threshold combinations were tested on the -3 minute image and compared to the manual% ID in color display.

(Experiment 5) In order to evaluate the reproducibility between operators of the semi-automatic renal region-of-interest setting method according to the present invention, three operators A, B, and C used this method in a 1-2 minute image. % ID was compared using parameters considered to be optimal in the evaluation of (5) above.

(Experimental results) In the following results,% ID
Is expressed as mean ± standard deviation, and reproducibility among the three operators was evaluated by one-way repeated measures analysis of variance. And the comparison of the two in this group is a post hoc test (Fisher's PLSD
Method, Scheffe method, Bonferroni / Dunn method).
In the same operator, a comparison between the manual method and the semi-automatic renal region-of-interest setting method according to the present invention (hereinafter, also simply referred to as the semi-automatic method) was performed by Student's t-test. p <0.05 was considered significant. In addition, a linear regression analysis by the least squares method was added to all of the above.

(Experiment 1)% ID is 4.14 ± 1.2 for A
7, 4.32 ± 1.28 for B, 3.28 ± 1.07 for C
And a significant difference was recognized in% ID (p <0.000
1). FIGS. 8, 9, and 1 show the relationship between A, B, and C of% ID.
0 is shown. Although they showed good correlation with each other, AC
And BC, a significant difference was observed (r is a correlation coefficient in the figure).

(Experiment 2) The% ID of the operator A in the second trial was 4.12 ± 1.26. This was not significantly different from the first time (p = 0.3
7). The relationship of% ID is shown in FIG.

(Experiment 3)% ID obtained by the manual method of color display by operator A was 4.63 ± 1.33.
A significant difference in% ID was observed in comparison with the first monochrome display (p <0.0001). FIG. 12 shows the relationship between% ID in color display and monochrome display.

(Experiment 4) When the first threshold was set to 50% in the 1-2 minute image, 8 kidneys (4 on the left and 4 on the right) could not be separated from the liver and spleen. In 60%, three kidneys (left 2 and right 1) were inseparable. Even at 70%, two kidneys are still inseparable,
In 70%, the central renal region did not reflect the morphology of the kidney. Therefore, 60% was adopted as the first threshold. Next, the second threshold value was set to 30%, 20%, and 10%, and compared with the value of the manual method in color display. The relationship between these% IDs is shown in FIG. 13, FIG. 14, and FIG. A significant difference was found in% ID between the manual method and the semi-automatic method with the second threshold of 30% (p <
0.0001). No significant difference was observed with the second threshold value of 20% (p = 0.10). Although no significant difference was observed with the second threshold of 10% (p = 0.77), 20% was considered to be optimal as the second threshold from the shape of the region of interest.

3 that cannot be separated from surrounding organs at the first threshold of 60%
The% ID of the kidney was calculated as 1.87
%, Right kidney 1.83%, left kidney 2.95%. These could not be visualized because of their decreased renal function and proximity to the surrounding organs.

In the 2-3 minute image, a large defect was observed in the renal central region of one kidney at the first threshold value of 60%, and the renal central region did not reflect the morphology of the kidney. In 50%, all 62 kidneys could be visualized. In 40%, three kidneys (left 2 and right 1) were inseparable from liver and spleen. So the first threshold is 50%
It was adopted.

Next, the second threshold value is set to 30%, 20%, 10%
And compared with the manual method in color display. 30% and 1
At 0%, there was a significant difference in% ID (p <0.0
001). No significant difference was observed at 20% (p =
0.06). The relationship between these% IDs is shown in FIGS.
7, shown in FIG.

(Experiment 5) First threshold value 6 with 1-2 minute image
The% ID in the semi-automatic method of 0% and the second threshold value of 20% was 4.60 ± 1.31 for operators A, B and C. No significant difference was found in% ID among the three operators (p> 0.9999). FIG. 19 shows% ID between A, B and C.
Shows the relationship. Complete reproducibility was observed among the three operators in 58 kidneys out of 59 kidneys that could be visualized. The remaining one
The% ID of the kidney was 4.87% for A and 4.76% for B and C, the latter being only underestimated within 2.3% of the former.

(Examination) As can be seen from the above experimental results, when mercaptoacetyltriglycine technetium (99mTc-MAG3) was used as the radiopharmaceutical, the first threshold value was determined in the 1-2 minute image in the semi-automatic method of the present invention. Preferably, the maximum threshold value is 60% of the maximum pixel value, and the second threshold is 20% of the maximum kidney pixel value. In the 2-3 minute image, the first threshold is 50% of the maximum kidney pixel value, and the second threshold is the maximum kidney pixel value. It has been found that the most desirable value is 20%.

As for the reproducibility between operators, it was proved that almost perfect reproducibility could be obtained for the 1-2 minute image as apparent from FIG. 19 when the above threshold value was adopted. In other words, a region of interest similar to that of a skilled nuclear medicine physician can be set even by an inexperienced operator.

Regarding the renal imaging ability, out of all 62 kidneys, only 3 kidneys could not be imaged in the 1-2 minute image, and all kidneys could be imaged in the 2-3 minute image. In addition, the minimum% I of the kidney that can be drawn with a 1-2 minute image
D was only 0.72. On the other hand, in the one-step threshold method,
Inability to depict in 1-2 minute images: 51 kidneys out of 62 kidneys 2-3
The minute image was 20 kidneys. Compared with the one-step threshold method, it can be seen that the semi-automatic renal region-of-interest setting method of the present invention has significantly better delineation ability.

Incidentally, in this experiment, mercaptoacetyltriglycine technetium (99mTc-MAG3) was used as a radiopharmaceutical for renal function diagnosis. It goes without saying that the present invention can be applied to 99mTc-DTPA and 131I-OIH scintigraphy.

Further, if the semi-automatic renal region of interest setting method in the renal scintigraphy described above is extended, it is easily conceived that it can be an effective region of interest setting method not only for the kidney but also for scintigrams of other organs. Will be done. The procedure is summarized as follows.

That is, radioactivity released from a radiopharmaceutical administered into a living body is collected by a gamma camera, the obtained image data (scintigram) is transferred to a computer, and the digital image is displayed on a computer monitor. Then, in a semi-automatic renal region of interest setting method in scintigraphy to semi-automatically set the region of interest of the target organ based on the pixel value, first, display a scintigram on a computer monitor, and on the displayed image Click on a pixel near the center of the target organ, and secondly, determine the maximum pixel value of the pixels included in the first plurality of pixel regions in the predetermined range centered on the clicked pixel, and Third, the pixel value is larger than the first plurality of pixel areas centered on the clicked pixel. In a second plurality of pixel regions having a fixed size, a group of pixels having a pixel value equal to or greater than a first threshold value based on the maximum pixel value of the target organ is defined as a central region of the target organ, and fourthly. If there is a portion where two regions are continuous via only one pixel in the central region of the target organ, this one pixel is deleted from the central region of the target organ, and fifthly, the central region of the target organ is removed. A set of pixels consisting of only the pixels having continuity with the clicked pixel among the pixels within is newly set as the center region of the target organ, and sixthly, a peripheral edge having a predetermined width outside the center region of the target organ. A pixel having a pixel value higher than a second threshold value based on the maximum pixel value among all or part of the pixels included in the region is defined as a target organ peripheral region, and this is defined as the target organ. In addition to the heart region is a region of interest of the target organ is carried out by using a sequential hardware resources the first to sixth processing.

Of course, the parameters such as the values of the first and second thresholds in the above processing, the sizes of the first and second plural pixel regions and the width of the peripheral region are the target organ, the objective, It should be determined appropriately based on clinical results depending on the radiopharmaceutical used and the like.

As described in detail above, the semi-automatic renal region of interest setting method in renal scintigraphy according to the present invention realizes almost completely the reproducibility between operators, and is an extremely excellent setting method having good renal rendering ability. There is no doubt that this is an extremely useful region of interest setting method applicable to scintigraphy of other organs.

[0068]

The semi-automatic kidney region of interest setting method in renal scintigraphy according to the present invention has the following excellent effects because of the above-described processing procedure.

(1) It is possible to set a renal region of interest with good reproducibility by different operators.

(2) The region of interest in the kidney can be set in a short time.

(3) It is possible to set a region of interest similar to the setting of a skilled nuclear medicine physician.

(4) It is possible to visualize even a kidney with low function.

(5) Applicable to setting a region of interest in scintigraphy of other organs.

[Brief description of the drawings]

FIG. 1 is a flowchart showing a processing procedure of a semi-automatic renal region of interest setting method in renal scintigraphy according to the present invention.

FIG. 2 shows setting regions in the first and second processes of the semi-automatic renal region of interest setting method according to the present invention, in an added image of one subject displayed on a computer monitor.

FIG. 3 shows a renal center region in a third process of the semi-automatic renal region of interest setting method according to the present invention in an added image of one subject displayed on a computer monitor.

FIG. 4 shows a renal center region in a fourth process of the semi-automatic renal region of interest setting method according to the present invention, in an added image of one subject displayed on a computer monitor.

FIG. 5 shows a renal center region in a fifth process of the semi-automatic renal region of interest setting method according to the present invention in an added image of one subject displayed on a computer monitor.

FIG. 6 shows a renal area of interest in a sixth process of the semi-automatic renal area of interest setting method according to the present invention, in an added image of one subject displayed on a computer monitor.

FIG. 7 is a diagram for explaining a method of cutting continuity between two regions through one pixel in the central kidney region.

FIG. 8 shows a correlation between operators A and B of% ID by a manual method in a 1-2 minute image and monochrome display (n
= 59).

FIG. 9 shows the correlation between% ID operators A and C by the manual method in a 1-2 minute image, monochrome display (n
= 59).

FIG. 10 shows the correlation between% ID operators B and C by the manual method in a 1-2 minute image and monochrome display (n = 59).

FIG. 11 shows the first and second correlations by the operator A of% ID by the manual method in a 1-2 minute image and monochrome display (n = 59).

FIG. 12 shows a correlation between% ID in a monochrome image and a color display by a manual method of the operator A for a 1-2 minute image (n = 59).

FIG. 13 shows the correlation between% ID by a manual method in a color display and% ID by a semi-automatic method (first threshold 60%, second threshold 30%) by a 1-2 minute image, operator A (n = 59).

FIG. 14 shows a correlation between the% ID by the manual method in the color display and the% ID by the semi-automatic method (first threshold 60%, second threshold 20%) by the operator A in the 1-2 minute image, operator A (n = 59).

FIG. 15 shows a correlation between% ID by a manual method in a color display and a% ID by a semi-automatic method (first threshold: 60%, second threshold: 10%) by a 1-2 minute image, operator A (n) = 59).

FIG. 16 shows a correlation between% ID by a manual method in a color display and% ID by a semi-automatic method (first threshold value 50%, second threshold value 30%) by a 2-3 minute image, operator A (n = 62).

FIG. 17 shows a correlation between% ID by a manual method in color display and% ID by a semi-automatic method (first threshold value 50%, second threshold value 20%) by a 2-3 minute image, operator A (n = 62).

FIG. 18 shows a correlation between% ID by a manual method in color display and% ID by a semi-automatic method (first threshold value 50%, second threshold value 10%) by a 2-3 minute image, operator A (n = 62).

FIG. 19: 1-2 minute image, semi-automatic method (first threshold 60)
%, The second threshold value 20%) indicates the correlation between% ID operators A, B, and C (data of B and C are completely equal) (n = 59).

[Explanation of symbols]

 P1 First multiple pixel area P2 Second multiple pixel area M1, M2 Kidney center area R Kidney margin area S Kidney area of interest K Crosslinked pixel α, β Click point

Claims (6)

[Claims] 1. A radioactivity released from a radiopharmaceutical for renal scintigraphy administered into a living body is collected by a gamma camera, the obtained image data (scintigram) is transferred to a computer, and the obtained data is transferred to a computer monitor. In a semi-automatic renal region of interest setting method in renal scintigraphy, which displays a digital image and semi-automatically sets the region of interest in the left and right kidney based on the pixel values, first, displays and displays a scintigram on a computer monitor Secondly, a pixel near the center of each of the left and right kidneys on the image is clicked, and secondly, a maximum pixel value of a pixel included in a first plurality of pixel regions within a predetermined range centered on the clicked pixel is obtained. Third, the maximum pixel value of the kidney, and thirdly, the first pixel area centered on the clicked pixel. In a second plurality of pixel regions having a predetermined size larger than the maximum, a group of pixels having a pixel value equal to or greater than a first threshold value based on the maximum pixel value of the kidney is defined as a kidney central region, and fourthly. If there is a place where two regions are continuous via only one pixel in the renal center region, this one pixel is deleted from the renal center region, and fifthly, among the pixels in the renal center region, A new set of pixels consisting of only the pixels having continuity with the clicked pixel is newly defined as a renal center region, and sixthly,
A pixel having a pixel value higher than a second threshold value based on the maximum pixel value among pixels entirely or partially included in a 5-20 mm width peripheral region outside the renal center region is defined as a renal peripheral region. Wherein the first to sixth processes described above are sequentially performed in addition to the renal central region and the region of interest of the kidney, wherein a renal region of interest setting method for renal scintigraphy is provided. 2. The semi-automatic region of interest setting method in renal scintigraphy according to claim 1, wherein the first plurality of pixel regions are 30 to 70 mm in horizontal dimension and 60 to 120 mm in vertical dimension.
, And the second plural pixel area has a horizontal dimension of 100 to
A semi-automatic renal region of interest setting method in renal scintigraphy, wherein the method has a range of 250 mm and a vertical dimension of 150 to 300 mm. 3. The semi-automatic renal area of interest setting method in renal scintigraphy according to claim 1 or 2, wherein the sixth processing is replaced with 5 to 20 mm outside the central renal area.
A semi-automatic renal scintigraphy characterized in that a process is performed in which all pixels partially or entirely included in a peripheral region of a width are defined as a peripheral region of the kidney and are added to the central region of the kidney to be a region of interest of the kidney. Method of setting renal area of interest. 4. The method for semi-automatic kidney region setting in renal scintigraphy according to claim 1 or 2, wherein 60 to 12 seconds after intravenous injection of mercaptoacetyltriglycine technetium (99mTc-MAG3) as a radiopharmaceutical.
A renal scintigraphy characterized in that the first threshold value in the added image of 0 seconds (1-2 minute image) is 50% to 70% of the maximum pixel value, and the second threshold value is 1% to 40% of the maximum pixel value. A semi-automatic method for setting renal regions of interest in graphography. 5. The semi-automatic renal area of interest setting method for renal scintigraphy according to claim 1 or 2, wherein from 120 seconds after intravenous injection of mercaptoacetyltriglycine technetium (99mTc-MAG3) as a radiopharmaceutical.
The first threshold value in the 80-second added image (2-3 minutes image) is a pixel value of 40% to 70% of the maximum pixel value, and the second threshold value is 1% to 40% of the maximum pixel value. Semi-automatic method for setting a renal area of interest in renal scintigraphy characterized by: 6. The method according to claim 1, wherein said first and second means are different from each other.
Alternatively, in the method for setting a region of interest in renal renography according to claim 4 or 5, the fifth processing and the sixth processing are sequentially performed after the first to third processing except for the fourth processing. A semi-automatic method for setting a renal area of interest in renal scintigraphy.