JP2001118058A - Image processor and radiation medical treatment planning system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain an image processor which can facilitate manual operation and reduce the work load, and extract an object fast with adequate precision. SOLUTION: This invented image processor generates a result image 117 which has an object extracted nearly automatically from a slice image by a ROI segmentation part 106, a threshold process part 108, a connection/ disconnection part 110, an area narrowing-down part 112, and a shape feature quantity calculation part 114. Then, a wrong extracted area decision part 118 decides whether or not a wrong object is extracted as to plural result images 117 and fills a result image 117 having a wrong object extracted with a correct object by interpolating the result images before and after the image to facilitate manual operation, so that the object can be extracted fast with adequate precision.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus, and more particularly, to a slice image such as a CT image, a feature relating to a pixel intensity of an object on a single image plane, and a feature relating to a shape of the object. The object is extracted based on the above, and an error extraction determination process is performed from the context between the images, and the movement and rotation of the extraction region between the images are performed for slice images in which the object cannot be extracted. The present invention relates to an image processing apparatus capable of extracting a contour of an object at high speed and with appropriate accuracy by performing an interpolation process in consideration of the interpolation process. Further, the present invention relates to a radiation treatment planning system using the image processing device.

[0002]

2. Description of the Related Art Image processing for extracting an object such as a bone or a muscle from a slice image of a human body or an object, for example, a CT (Computed Tomography) image, and grasping the contour or shape of the object. The device is generally known (see, for example, JP-A-06-070923 and JP-A-07-299060). As a method for extracting an object in this type of image processing apparatus, an automatic method of automatically performing image processing to extract the object and tracing the outline of the object using an input device such as a mouse are used. And a manual method of extracting an object by using a manual method. Further, a semi-automatic method is also used, which has the advantages of the automatic method such that the work is reduced and the advantages of the manual method such as high accuracy.

[0003] Specifically, the above-mentioned Japanese Patent Application Laid-Open No. 06-0709 is disclosed.
Although the image processing apparatus (image diagnostic apparatus) disclosed in Japanese Patent Publication No. 23 is of a semi-automatic type, this image processing apparatus creates a smaller number of cross-sectional images from a large number of CT images, and By tracing the outline of the object with the mouse with respect to the image, the amount of tracing work is reduced.

The image processing apparatus disclosed in the above-mentioned Japanese Patent Application Laid-Open No. 07-299060 is of an automatic type. In this image processing apparatus, a region of interest is specified in a CT image object, The average of the CT values of the region of interest and C
An extraction condition is set based on the upper limit value and the lower limit value of the T value, and among the pixels adjacent to the region of interest, one that satisfies the above extraction condition is determined to be a part (same tissue) of the object. By performing such processing in a three-dimensional manner, an object is extracted.

[0005]

However, among such conventional object extraction methods, the automatic method has a problem that it requires setting of a large number of parameters and requires a large amount of calculation time. is there. In addition, when the parameter setting is good, the output result is generally good, but there is also a problem that a large processing time is required and the result is not always satisfactory. Note that, in the automatic image processing apparatus disclosed in Japanese Patent Application Laid-Open No. 07-299060, if parameters are set poorly, an excessively large object is extracted, and its contour or shape is accurately extracted. There is a problem that can not be.

On the other hand, in the manual method, when a large number of slice images such as CT images are processed, contour tracing must be performed many times, which requires a complicated (heavy labor) and a long operation. There's a problem. In this case, since the operator traces each slice image while checking it, there is an advantage that the object can be extracted with high accuracy.

In the semi-automatic image processing apparatus disclosed in Japanese Patent Application Laid-Open No. H06-070923, it is difficult to determine whether the number of trace cross-sectional images generated from slice images is appropriate. Further, when the number of cross-sectional images is set to be large, there is a problem that the amount of work for contour tracing increases.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned conventional problems, and can reduce the amount of work by facilitating manual work while automating the processing process as much as possible. Or an object to solve the problem of obtaining an image processing device capable of extracting the image with high accuracy at a high speed. Another object of the present invention is to provide a radiation treatment planning system capable of extracting a contour of a body organ at high speed and optimizing radiation treatment beam conditions using the image processing apparatus. And

[0009]

According to a first aspect of the present invention, there is provided an image processing apparatus comprising: (a) a region image including a target object from an original image (slice image); ROI image),
(B) threshold processing means for performing threshold processing on a region image cut out by the cutout means with a pixel intensity unique to the object, and (c) a target object in the region image subjected to the threshold processing. Cutting means for cutting off other areas (that is, unnecessary areas); (d) narrowing-down means for narrowing down candidate areas of the object from a plurality of areas cut by the cutting means; and (e) shape characteristics of the object. A shape feature amount calculating unit that calculates an amount, and extracts an object based on the shape feature amount from the region image in which the candidate regions of the object are narrowed down;
(F) error extraction area determination means for determining whether or not an erroneous target has been extracted for each area image based on a result obtained by performing processing by each of the above means on each of a plurality of original images. And (g) an erroneous object was extracted (that is, the object could not be extracted)
For the region image, the interpolation process is performed on the region images before and after the region image (after deleting an erroneous object).
Interpolating means for supplementing the object.

More specifically, this image processing apparatus extracts an object from slice image sequences (slice sequences) such as a CT image input to an image input unit for each slice image, and performs error extraction determination between slice images. Performing, for the slice image determined to be erroneous, by calculating the contour of the object by interpolation processing using the preceding and following slice images,
The object is three-dimensionally extracted at high speed. Here, the threshold processing means is configured to extract the target object by reducing noise from the region image and extracting pixels within a range of pixel intensities in the target object.

[0011] The cutting means cuts the connectivity of unnecessary areas and maintains the size of the object by performing contraction processing and expansion processing on the extracted area image. The narrowing-down means narrows down the candidates of the target object from the size of the target object from the extracted region image and the constraint condition that the target object does not exist near the frame of the region image. In each of the small regions in the extracted image, the shape feature amount calculating means obtains a shape feature amount defined by the object, and extracts the region having the maximum value as the object in the slice image. is there.

The error extraction area determination means determines an error in the extraction area of a certain slice image based on the continuity with the extraction areas of the preceding and following slice images in the extracted image sequence. The interpolation processing means interpolates from the slice image from which the target object has been extracted in the extracted region image sequence for the slice image from which the target object has not been extracted.

Thus, according to the image processing apparatus of the first aspect, it is possible to facilitate manual work while reducing the amount of work by automating the processing process as much as possible.
The object can be extracted at a high speed with a reasonable accuracy. In short, high-speed processing by simple processing and accurate object extraction in consideration of continuity between images of the object can be performed.

The image processing apparatus according to a second aspect of the present invention is the image processing apparatus according to the first aspect, wherein the extracting means converts a plurality of original images into a time-series moving image format and converts the original images into a moving image. The clipping position is set while displaying. According to this image processing apparatus, an area image can be cut out while confirming it on a two-dimensional image.

The image processing apparatus according to a third aspect of the present invention is the image processing apparatus according to the first aspect, wherein the extracting means processes the plurality of original image data as three-dimensional data and displays the three-dimensional data, A cutout position is set on the three-dimensional display.
In this image processing apparatus, a region image can be cut out while viewing the entire image three-dimensionally.

In an image processing apparatus according to a fourth aspect of the present invention, in the image processing apparatus according to the first aspect, the narrowing means includes information including a size of the object and a position of the object in the image. It is characterized in that the candidate area of the object is narrowed down using (knowledge). According to this image processing apparatus, the unnecessary area can be reduced, and the extraction of the target object becomes easy.

The image processing apparatus according to a fifth aspect of the present invention is the image processing apparatus according to the first aspect, wherein the shape feature quantity calculating means uses the degree of circularity for a substantially circular object area to obtain a shape. The feature amount is calculated. According to this image processing device, it is possible to easily extract an object that can approximate a circle.

The image processing apparatus according to a sixth aspect of the present invention is the image processing apparatus according to the first aspect, wherein the cutting means contracts the object using a structural element obtained by modeling the shape of the object. The object is cut off from the other region by being expanded. According to this image processing device, the object to be extracted can be subjected to contraction / expansion processing so as not to lose its shape.

An image processing apparatus according to a seventh aspect of the present invention is the image processing apparatus according to the first aspect, wherein the error extraction area determining means is configured to extract an arbitrary first area image in the area image sequence. A straight line that passes through the center of gravity of the region (object) and the center of gravity of the extraction region (object) in the second region image that is continuous with the region, and the extraction region in the third region image that is continuous with the second region image If the distance from the intersection with the (object) to the center of gravity of the extraction area (object) in the third area image exceeds a predetermined reference value, an incorrect object is extracted for the third area image. It is characterized in that it is determined that there is According to this image processing apparatus, an error extraction determination can be accurately performed in consideration of the continuity of the object in the image sequence direction.

An image processing apparatus according to an eighth aspect of the present invention is the image processing apparatus according to the first aspect, wherein the interpolation processing means determines the main axis direction of the object in the area image used for the interpolation processing. By performing interpolation after normalizing the position or orientation of the center of gravity of the object, the interpolation is performed in consideration of the movement and rotation of the object on the image plane between the region images. It is characterized by the following. According to this image processing apparatus, it is possible to accurately perform shape interpolation in consideration of the movement / rotation of an object between image sequences.

A radiation treatment planning system according to a ninth aspect of the present invention uses the image processing apparatus according to any one of the first to eighth aspects to irradiate an irradiation target and a target organ from a treatment site image of a patient. High-speed extraction, suppresses the dose to the organ of interest, accurately calculates the optimal irradiation beam conditions that can apply a sufficiently high dose to the irradiation target, and performs radiation therapy on the patient based on the irradiation beam conditions. It is characterized in that it is performed. According to this radiation treatment planning system, effective radiation treatment can be easily performed on a patient.

[0022]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be specifically described below. Embodiment 1 FIG. First, a first embodiment of the present invention will be described. FIG. 1 is a block diagram illustrating a configuration of the image processing apparatus according to the first embodiment. Then, FIGS. 2 (a) to 2 (g)
Is a photograph replacing a drawing of a halftone image in which various real images created by the image processing apparatus shown in FIG. 1 are displayed on a display.

In FIG. 1, reference numeral 102 denotes an image input unit to which a series of a plurality of slice images (for example, CT images), that is, a series of slice images is input, and the slice image series is stored. Reference numeral 104 denotes a two-dimensional object extraction processing unit that captures a slice image from the image input unit 102 and extracts a two-dimensional object (object candidate) from the slice image.

As shown in FIG. 1 and FIG. 2, the two-dimensional object extraction processing unit 104 includes a region (hereinafter, referred to as a ROI (Region of
Interest). ), A ROI extraction unit 106 that generates an ROI image 107, a threshold processing unit 108 that applies a threshold processing to the ROI image 107 with a pixel intensity unique to the target object to generate a threshold processing image 109, and a threshold processing image 109. Connecting section 110 for performing a connection cutting process on the object to cut the object and the unnecessary area, and an image 11 after the connection cutting processing
An area narrowing unit 112 for narrowing down a candidate area of an object from a plurality of areas in 1; calculating a shape characteristic amount of the object; and extracting an object in the image 113 after the area narrowing down based on the shape characteristic amount Then, a shape feature amount calculation unit 114 that creates a region image 115 from which the target object is extracted is provided.

Reference numeral 116 denotes a three-dimensional object extraction processing unit. The three-dimensional object extraction processing unit 116 extracts the two-dimensional object extracted for each slice image by the two-dimensional object extraction processing unit 104. From the (object candidate), a three-dimensional object is extracted in consideration of the continuity of the object between the slice images. The three-dimensional object extraction processing unit 116 includes an error extraction area for determining whether or not an erroneous object (area) has been extracted for each result image 117 created by the two-dimensional object extraction processing unit 104. A determination unit 118 is provided. Further, the target image is supplemented (given) by performing an interpolation process on the result image 117 in which the target object could not be extracted based on images before and after the image, and the result of the interpolation process is given. An interpolation processing unit 120 that creates an image 121 is provided.

Next, the operation of the image processing apparatus will be described. In this image processing apparatus, first, the image input unit 10
2, a slice image sequence is input. Next, a slice image sequence is captured from the image input unit 102 to the two-dimensional object extraction processing unit 104. Two-dimensional object extraction processing unit 104
First, in order to reduce unnecessary processing, RO
The I extraction unit 106 performs an extraction process of extracting a region image including an object, that is, an ROI image 107, from the slice image sequence, and performs a smoothing filter process on the ROI image 107 to remove noise. Is performed.

Here, the extraction processing of the ROI image 107 can be performed by a method of converting a slice image sequence into a time-series moving image, and setting a region image completely encompassing the object while displaying a moving image. . Alternatively, the slice image sequence may be processed as three-dimensional data to perform three-dimensional display, and the extraction processing may be performed by a method of setting an extraction area on the three-dimensional display with a mouse or the like.

Next, a single ROI is
The image 107 is subjected to threshold processing using a previously determined pixel intensity unique to the target as a threshold, and a threshold processing image 109 in which a region including the target is extracted is created. In the threshold processing image 109, the target object and other areas (unnecessary areas) are extracted in a connected state. For this reason, in the connection / disconnection unit 110, the threshold processing image 10
9 is subjected to a connection cutting process, that is, a region contraction / expansion process. Specifically, first, each region in a connected state is cut into a plurality of regions including an object (region) and other regions by a contraction process. Since the size of each of the cut regions has been reduced by the contraction process, the size of the region is restored by performing the expansion process next. The contraction / expansion processing is performed on, for example, eight neighboring pixels.

Next, in order to accurately extract the object from the cut area, the area narrowing unit 112 deletes an unnecessary area by a general rule. In addition, as a deletion method based on a general rule, for example, a method of removing an area smaller than a certain size, or in view of the fact that an object is generally located near the center of the ROI image 107, in the vicinity of the frame of the ROI image 107 A technique such as erasing the area can be used.

Then, the shape feature amount calculating section 114 calculates the shape feature amount of the object candidate, and extracts a region having the extreme value (for example, the maximum value) as the object in the ROI image. For example, when the target is circular, the circularity of the region can be used as the shape feature amount. The circularity is represented by the following equation 1.

(Equation 1)

The threshold processing section 108, the connection cutting section 110, the area narrowing section 112, and the shape feature quantity calculation section 1
By performing a series of processes performed by each of the processes 14 on each ROI image 107, an object in the ROI image sequence can be extracted. However, since the extracted area is not always the correct object, the three-dimensional object extraction processing unit 116 removes an erroneous object in consideration of the relationship (relevance) between the ROI images. Then, the extraction (compensation) of the correct object is performed.

More specifically, the ROI image (hereinafter referred to as the ROI image) to be determined by the error extraction region determining unit 118
The logical product area of the object between the target ROI image and the ROI image (extraction result) before and after the target ROI image is obtained. If this logical product area is an empty set, the target R
The OI image is determined as extracting an erroneous target, and the erroneous target is deleted from the target ROI image. Such processing is performed on all ROI images (entire ROI image sequence).

Then, in the ROI image sequence, an erroneous object is deleted, that is, an ROI image from which a correct object cannot be extracted (hereinafter referred to as an "interpolation target ROI image"). In the unit 120, a correct target is obtained from the ROI images before and after the ROI image to be interpolated by interpolation processing, and the target is supplemented. For example, an object (region) at the same position in the ROI image to be interpolated is determined by performing a primary interpolation process using a pixel value at an arbitrary same pixel position on the ROI image used for the interpolation process. 2 (a) to 2 (g) show, respectively,
Processed images (actual images) 107, 10 obtained in each step
Specific examples of 9, 111, 113, 115, 117 and 121 will be shown.

As described above, in the image processing apparatus according to the first embodiment, the amount of work can be reduced by simplifying (simplifying) the manual operation while automating the processing as much as possible. In addition, by taking into account the continuity (relevance) between the slice images, erroneously extracted objects are eliminated, and objects that cannot be extracted are supplemented (supplemented) by interpolation processing. The object can be extracted at a high speed with a reasonable accuracy.

Embodiment 2 Hereinafter, a second embodiment of the present invention will be described. FIGS. 3A and 3B show the second embodiment.
Connection disconnection unit 1 in FIG.
It is a conceptual diagram which shows the Example (specific example) of the connection cutting part of a structure different from 10. In FIG. 3A, reference numeral 202 denotes an object to be extracted, and reference numeral 204 denotes a structural element used for expansion / contraction processing for controlling connectivity between regions.

Here, the expansion process is a process of calculating a logical sum set of all the regions obtained by moving the object 202 up, down, left, and right by the respective coordinate values of the structural element 204 in parallel. In other words, when the center point of the structural element 204 is slid within the object 202, the process can be interpreted as a process of obtaining a set 206 of the entire area covered by the structural element 204.
The contraction process is a process of calculating a logical product set of the entire region obtained by moving the object 202 up, down, left, and right by the respective coordinate values of the structural element 204. In other words, when the structural element 204 is slid within the object 202, it can be interpreted as a process of obtaining a set 208 of central points of the structural element 204 included in the object 202.

Thus, the contour of the figure obtained by performing the expansion processing or the compression processing on the object 202 is affected by the shape of the structural element 204. For this reason, by making the shape of the structural element 204 a general shape corresponding to the target object 202, it is possible to perform expansion / contraction processing without damaging the shape of the target object. As is clear from FIG. 3A, the expansion processing result 206 when the rectangular structural element 204 is used for the rectangular object 202 is a rectangle.

On the other hand, as shown in FIG.
The expansion processing result 212 when the circular structural element 210 is used for the rectangular object 202 is affected by the shape (circular) of the structural element 210, and has a shape in which four corners of the rectangle are rounded. As a general shape of the structural element corresponding to the object, a shape used in the shape feature amount calculation unit can be used.

Embodiment 3 Hereinafter, a third embodiment of the present invention will be described. FIG. 4 shows an example (specific example) of an error extraction area determination process by an error extraction area determination unit having a configuration different from that of the error extraction area determination unit 118 in FIG. 1 in the image processing apparatus according to the third embodiment. It is a conceptual diagram. FIG.
, 302, 304, and 306 are ROI images (region images) in which three consecutive objects are extracted. Further, 308 and 310 respectively represent two consecutive
Extraction area (object) in ROI images 302 and 304
It is.

In this error extraction area determination process, the center of gravity 314 of the extraction area 308 and the center of gravity 3 of the extraction area 310
16 and the vector 32 connecting the centroids 314 and 316
0 is required. Then, the intersection 32 between the ROI image 306 and the half line 322 extending in the direction extending the vector 320 from the center of gravity 316 of the extraction region 310 of the ROI image 304
If the distance 326 between the No. 4 and the center of gravity 318 of the extraction region 312 (object) of the ROI image 306 is larger than a predetermined reference value, the extraction region 312 of the ROI image 306 Object).

Such an error extraction area determination process is performed according to RO
This is performed on the I image sequence. At this time, by selecting the initial ROI images 302 and 304 from the ROI images having substantially the same center of gravity, it is possible to avoid setting errors of the initial vector 320. By such processing, the directionality of the object in the ROI image sequence is considered,
It becomes possible to delete the extraction area in the ROI image that deviates from that direction.

Further, the directionality in the error extraction area determination processing is obtained by interpolation or extrapolation using three or more ROI images, instead of the vectors between the centers of gravity of two consecutive ROI images. It may be. This makes it possible to more accurately determine the directionality between images, and to more accurately consider the direction of the target object.

Embodiment 4 FIG. Hereinafter, a fourth embodiment of the present invention will be described. FIG. 5 is a concept showing an example (specific example) of interpolation processing between ROI image sequences by an interpolation processing unit having a configuration different from that of the interpolation processing unit 120 in FIG. 1 in the image processing apparatus according to the fourth embodiment. FIG. In FIG. 5, reference numeral 402 denotes an arbitrary ROI image from which an object is extracted, 404 denotes an object region in the ROI image 402, 406 denotes a main axis of the object region 404, and 408 denotes an object region 404. Is the center of gravity G.

In this interpolation processing, first, the spindle 406
Angle of rotation θ (rotation direction) from
The translation amount (−g _x , −g _y ) from 08 is obtained.
Then, a linear conversion is performed on the target object region 404 by the following Expression 2.

(Equation 2)

The effect of the translation and rotation of the object can be eliminated by performing the linear transformation according to Equation 2, and as a result, the position or orientation of the object can be normalized. Then, an intersection or a division point R _{i, φ} (X) of a straight line that divides the ROI image 410 (the converted image 410) after the linear transformation at an arbitrary division angle φ and a contour line of the object area 404.
_{i, φ} , Y _{i, φ} ) are obtained. Here, i is RO
I is an I image number, and φ is a division angle.

[0046] In addition, the ROI image 412 and a ROI image 414 linear conversion respectively ROI image numbers p and q, intersection or dividing point R _p in the case where the division angle phi was _{φ 0,} φ ₀ (X _{p , φ 0} , Y _{p, φ 0} ) and R _q,
Corresponding division points such as _{φ 0} (X _{q, φ 0} , Y _{q, φ 0} ) are obtained. An inverse transformation is performed on these division points by an inverse matrix of each transformation matrix, a coordinate value in the original image is obtained, and a linear interpolation process of the coordinate value is performed using the value. As a result, the intermediate RO having the ROI image number r
The corresponding point R _{r, φ 0} (X _{r, φ 0} , Y _{r, φ
0} ) is required.

Corresponding points in the intermediate ROI image 416 are obtained for a plurality of division angles φ, and an area having a contour formed by a curve connecting these corresponding points smoothly is an approximate area of the object in the intermediate ROI image 416. It is said. The approximate region obtained by this interpolation processing is obtained by calculating the amount of translation and rotation of the object and then normalizing the position and orientation thereof. It becomes possible to obtain the shape of an object.

Embodiment 5 FIG. Hereinafter, a fifth embodiment of the present invention will be described. The fifth embodiment relates to a radiation treatment planning system including the image processing device according to the first to fourth embodiments as a component. FIG. 6 is a schematic diagram (block diagram) schematically showing the radiation treatment planning system and the radiation treatment method. In FIG. 6, 50
Reference numeral 2 denotes a patient who receives radiation treatment, and reference numeral 504 denotes a radiation treatment planning system. Reference numeral 506 denotes a patient image (slice image sequence) such as a CT image of the patient 502. 508
Is an image segmentation unit for extracting a radiation irradiation target in radiation therapy or a region of an attention organ such as the spinal cord using the patient image 506 as input information.

Reference numeral 510 denotes a manual editing unit for editing the result image using a mouse or the like, using the patient image 506 and the result image created by the image segmentation unit 508 as input information. Reference numeral 512 denotes an optimum radiation beam condition calculation unit, which receives an output image of the manual editing unit 510 as input information. Note that the patient image 506, the image segmentation unit 508, the manual editing unit 510, and the radiation beam condition calculation unit 512
It is a component of the radiation treatment planning system 504.

Next, the operation of the radiation treatment planning system or the radiation treatment method will be described. First, patient 5
Patient image 5 such as a CT image obtained by imaging a target site 02
06 is input to the image segmentation unit 508.
Then, the image segmentation unit 508 performs high-speed segmentation of an area related to radiation treatment, such as an irradiation target or an attention organ, from the input patient image 506 based on the image intensity characteristics and shape characteristic amounts.

In the manual editing section 510, the segmentation result obtained in the image segmentation section 508 is edited by manual operation using a mouse or the like, and the contour of the target area is accurately extracted. Here, since the processing process in the image segmentation unit 508 is a simple combination of processes, a segmentation result can be obtained at high speed. Further, the obtained segmentation image is obtained by interpolation processing in consideration of parallel movement and rotation between the images and connectivity of the object, and therefore, it is possible to extract a shape with appropriate accuracy. .
In this radiation treatment planning system, the accurate shape is obtained by manually editing the segmentation image with reasonable accuracy. The burden on the user can be greatly reduced.

In this radiation treatment planning system, a conventional semi-automatic method for extracting a target object with high accuracy by manually editing the segmentation processing results obtained over a long processing time by adjusting complicated parameters. Processing time is shorter than the extraction method of
In addition, since the shape can be extracted with reasonable accuracy without substantially changing the manual operation amount, the processing time and the operation amount can be comprehensively determined to improve the efficiency of the segmentation processing. In addition, by applying the information of the irradiation target and the attention organ subjected to the segmentation to the radiation beam condition calculation unit, the beam condition can be calculated with high accuracy.

[0053]

According to the image processing apparatus according to the first aspect of the present invention, the cut-out means, the threshold processing means, the cutting means, the narrowing-down means, the shape characteristic amount calculating means, and the extraction area determining means determine the original image. A result image (region image) is created almost automatically from the extracted object. Next, the error extraction area determination means determines whether or not an erroneous target has been extracted for a plurality of result images. Since the correct object is compensated by the interpolation process,
High-speed processing by simple processing and accurate object extraction are possible. In other words, the amount of work can be reduced by facilitating manual work while automating the processing process as much as possible, and the object can be extracted at a high speed with reasonable accuracy.

According to the image processing apparatus according to the second aspect of the present invention, the cutout means converts a plurality of images into a time-series moving image format and sets the cutout position while displaying the moving image. Therefore, the region image can be cut out while checking it on the two-dimensional image. As a result, the extraction accuracy of the target object can be increased, and the extraction operation can be speeded up.

According to the image processing apparatus of the third aspect of the present invention, the extracting means performs three-dimensional display using a plurality of image data as three-dimensional data, and sets the extracting position on the three-dimensional display. Therefore, the region image can be cut out while viewing the entire image three-dimensionally, and the accuracy of extracting the object can be increased, and the speed of the extraction operation can be increased.

According to the image processing apparatus of the fourth aspect of the present invention, the narrowing-down means uses the information including the size of the object and the position of the object in the image to determine the candidate area of the object. Since the narrowing is performed, unnecessary areas can be reduced, and the operation of extracting an object can be speeded up.

According to the image processing apparatus of the fifth aspect of the present invention, the shape feature amount calculating means calculates the shape feature amount of the substantially circular object using the circularity. Therefore, it is possible to easily extract an object that can be approximated to a circle, and to speed up the operation of extracting the object.

According to the image processing apparatus of the sixth aspect of the present invention, the cutting means contracts and expands the object by using a structural element obtained by modeling the shape of the object, and Since it is designed to cut off other areas, the object to be extracted can be subjected to contraction / expansion processing so as not to lose its shape, and thus the extraction accuracy of the object can be improved. it can.

According to the image processing apparatus of the seventh aspect of the present invention, the error extraction area determining means determines the center of gravity of the extraction area in the first area image and the center of gravity of the extraction area in the second area image. If the distance from the intersection of the straight line passing through the extraction region and the extraction region in the third region image to the center of gravity of the extraction region in the third region image exceeds a predetermined reference value, an error is detected for the third region image. Since it is determined that the target object is extracted, it is possible to perform an error extraction determination in consideration of the continuity of the target object in the image sequence direction, and thereby to improve the extraction accuracy of the target object. it can.

According to the image processing apparatus of the eighth aspect of the present invention, the interpolation means determines whether the main axis direction of the object in the area image used for the interpolation processing and the position or orientation of the center of gravity of the object are normal. By performing the interpolation processing after the conversion, the interpolation processing is performed in consideration of the movement and rotation of the object on the image plane between the region images,
Shape interpolation can be performed in consideration of the movement and rotation of the object between the image arrays, and the accuracy of extracting the object can be improved.

According to the radiation treatment planning system according to the ninth aspect of the present invention, the irradiation target and the organ of interest can be quickly detected from the treatment site image of the patient by using the image processing apparatus according to the first to eighth aspects. Extract and accurately calculate the optimal irradiation beam conditions that can suppress the dose to the organ of interest and apply a sufficiently high dose to the irradiation target, and perform radiation therapy on the patient based on the irradiation beam conditions. As a result, the contours of the internal organs of the patient can be extracted at high speed to optimize the radiation treatment beam conditions, and effective radiation treatment can be easily performed on the patient.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of an image processing apparatus according to a first embodiment of the present invention;

FIGS. 2 (a) to 2 (g) are photographs in place of drawings, each of which is a halftone image in which various actual images created by the image processing apparatus shown in FIG. 1 are displayed on a display.

FIG. 3A is a conceptual diagram illustrating a method of cutting an object and an unnecessary area in the image processing apparatus according to the second embodiment of the present invention; FIG. FIG. 11 is a conceptual diagram showing another method of cutting the image.

FIG. 4 is a conceptual diagram illustrating a method of error extraction determination in an image processing device according to a third embodiment of the present invention;

FIG. 5 is a conceptual diagram showing a method of ROI image sequence interpolation processing in an image processing apparatus according to a fourth embodiment of the present invention;

FIG. 6 is a system configuration diagram schematically illustrating a radiation treatment planning system according to a fifth embodiment of the present invention;

[Explanation of symbols]

102 image input unit, 104 two-dimensional object extraction processing unit, 106 ROI extraction unit, 107 ROI image (real image), 108 threshold processing unit, 109 threshold processing image (real image), 110 connection cutting unit, 111 connection cutting processing After image (real image), 112 area narrowing unit, 113 image after area narrowing (real image),
114 shape feature amount calculation unit; 115 area image (real image) extracted based on the shape feature amount; 116 three-dimensional object extraction processing unit; 117 result image (real image);
118 error extraction area determination unit, 120 interpolation processing unit,
121 Result image (real image) subjected to interpolation processing,
202 Object, 204 Structural element, 206 Dilation processing result of the object 202 by the structural element 204, 20
8 Result of contraction processing by the structural element 204 of the object 202, 210 structural element, 212 Result of dilation processing by the structural element 210 of the object 202, 302 ROI image, 304 ROI image, 306 ROI image,
308 ROI image 302 extraction area, 310 RO
Extraction region of the I image 304, 312 extraction region of the ROI image 306, 314 centroid of the extraction region 308, 316
318 The center of gravity of the extraction region 310, 318 The center of gravity of the extraction region 312, 320 The vector formed by the center of gravity 314 and the center of gravity 316, 322 The half line drawn from the intersection 324 in the extension direction of the vector 320, 324 The intersection of the ROI image 306 and the half line 322 , 402 ROI image, 404
Object area, 406 the main axis of the object area 404,
408 the center of gravity of the object area 404; 410, a transformed image obtained by converting the coordinates of the object area 404; 412, the ROI image of the ROI image number p;
OI images, 416 ROI images with ROI image numbers between p and q (r), 502 patients, 504
Radiation treatment planning system 506 patient image of patient 502 508 image segmentation unit 510
Manual editing unit, 512 radiation beam condition calculation unit.

Continued on front page F term (reference) 4C082 AE01 AJ13 AN01 AR02 AT03 5B057 AA09 BA24 BA29 CF05 DA08 DB03 DC06 DC09 5L096 AA09 BA06 BA13 CA24 EA02 EA14 EA33 EA35 EA43 FA04 FA60 GA51

Claims (9)

[Claims] 1. An extracting device for extracting an area image including an object from an original image, and:
Threshold processing means for performing threshold processing at a pixel intensity unique to the object, cutting means for cutting the object and other areas in the area image subjected to the threshold processing, and cutting by the cutting means From a plurality of regions, a narrowing-down unit that narrows down a candidate region of the target object, a shape feature amount of the target object is calculated, and the target object is determined based on the shape feature amount from the region image in which the candidate region of the target object is narrowed down. Determining whether or not an erroneous target has been extracted for each region image, based on a shape feature amount calculating unit to be extracted and a result obtained by performing a process by each of the above units on each of the plurality of original images; Error extraction area determination means, and an interpolation processing means for supplementing the target image to the area image from which the erroneous target object has been extracted by performing interpolation processing on the area images before and after the target area image. The image processing apparatus according to claim Rukoto. 2. The method according to claim 1, wherein the extracting unit converts a plurality of original images into a time-series moving image format and sets an extracting position while displaying the original images as a moving image. The image processing apparatus according to any one of the preceding claims. 3. The method according to claim 1, wherein the cutout means processes the plurality of original image data as three-dimensional data, displays the processed three-dimensional data, and sets a cutout position on the three-dimensional display. Item 2. The image processing device according to Item 1. 4. The apparatus according to claim 1, wherein said narrowing means narrows down a candidate area of the object using information including a size of the object and a position of the object in the image. 2. The image processing device according to 1. 5. The image processing apparatus according to claim 1, wherein said shape feature value calculation means calculates a shape feature value of a substantially circular object using circularity. apparatus. 6. The cutting means cuts and expands the target object by contracting and expanding the target object using a structural element obtained by modeling the shape of the target object. The image processing apparatus according to claim 1, wherein: 7. The error extraction area determination means passes through a center of gravity of an extraction area in an arbitrary first area image in the area image sequence and a center of gravity of an extraction area in a second area image that is continuous with the extracted area. If the distance from the intersection of the straight line and the extraction region in the third region image that is continuous with the second region image to the center of gravity of the extraction region in the third region image exceeds a predetermined reference value, 3
The image processing apparatus according to claim 1, wherein it is determined that an erroneous target is extracted for the region image of (1). 8. The interpolation processing means performs an interpolation process after normalizing the main axis direction of the object in the region image used for the interpolation process and the position or orientation of the center of gravity of the object, 2. The image processing apparatus according to claim 1, wherein interpolation processing is performed in consideration of movement and rotation of an object on an image plane between region images. 9. An irradiation target and an attention organ are extracted at high speed from a treatment site image of a patient by using the image processing apparatus according to any one of claims 1 to 8, thereby suppressing a dose to the attention organ. A radiation treatment plan characterized by accurately calculating an optimum irradiation beam condition capable of applying a sufficiently high dose to an irradiation target, and performing radiation treatment on a patient based on the irradiation beam condition. system.