JPH0654833A - Three dimensional blood vessel display processing method - Google Patents

Abstract

PURPOSE:To extract only a blood vessel area by discriminating exactly a blood vessel and a background by executing once a projection processing onto a two-dimensional projection surface from a certain direction, and thereafter, executing an extraction processing to its projection data, and executing a processing for extracting the blood vessel. CONSTITUTION:First of all, with respect to three-dimensional data, a voxel on each sight line looked at from a certain direction is projected to each projection point on the projection surface (203). Subsequently, a processing for dividing the inside on the projection surface into every area is executed, and a processing for discriminating whether its area is a blood vessel or a background and extracting the blood vessel is executed (204). This series of processings are executed several times from a different direction, and a processing for unifying blood vessel areas extracted from each direction by O R on two dimensions or three dimensions is executed (206). In order to contain the voxel of the blood vessel in this three-dimensional blood vessel data, an extension processing is executed (209). By executing this processing, the extracted blood vessel is provided with a volume, and also, an edge is varied smoothly, and the extracted blood vessel three-dimensional data can be obtained more exactly.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a tomographic imaging apparatus utilizing a nuclear magnetic resonance phenomenon, and particularly when displaying a blood vessel image by three-dimensional data, or displaying a blood vessel and other organs in combination. The present invention relates to an effective three-dimensional display method.

[0002]

2. Description of the Related Art The following method is known as a typical known method for displaying a three-dimensional blood vessel image by projecting three-dimensional data obtained by MRI.

MIP (Maximum Intensity Projection) method Blood vessel anterior-posterior relationship display method In order to obtain a three-dimensional blood vessel image, it is necessary to project three-dimensional data in two dimensions.

In the above-mentioned method, the blood vessel is considered to have a higher density value than the background portion on the line of sight from the viewpoint to the projection surface, and therefore the three-dimensional blood vessel image is displayed by taking the maximum value on the line. Is the method. It is generally called the maximum value method and is a method that is often used. For example, "three-dimensional phase contrast angiography""ThreeDim
ensional Phase Contrast Angiography '(Magnetic Re
sonance in Medicine Vol.9, No.1 January 1989).

According to the above-mentioned method, on the line of sight from the viewpoint to the projection surface, a change curve in which peak values are sequentially held from the viewpoint is obtained, and an object existing at a position where the difference between the change curves is non-zero is regarded as a blood vessel. Further, a process of considering an object that is continuously non-zero as one blood vessel is performed, and a projection process is performed on a plurality of blood vessels up to the maximum value from these viewpoints to display a three-dimensional blood vessel image. This method is' 3-Dimens
ional MR Angiography using a New Rendering Techni
que 'K. Sano et al. (SMRM IX p.500 1990).

In any of the above methods, the background portion other than the blood vessel is similarly processed, and since the blood vessel is not extracted, there is no distinction between the blood vessel and the background.

[0007] On the other hand, the following methods are known as typical publicly known examples of general extraction methods.

Density Threshold Background Removal Method Area Extraction Method Based on Density Information In the above method, since the extraction target area and the background area have different density distributions, the extraction target area and the background area are separated by a preset threshold value. Then, the background area is removed and the extraction target area is extracted.

In the method described above, first, one point in the area to be extracted is selected, then points connected to it based on the density information are searched for from adjacent pixels, and the connected points are successively captured in the area. To extract. It is generally called Region Growing. This method is based on "Digital Picture Processing" by Azriel Rosenfeld (Translated by Nagao).
pp.335) and “Japanese Patent Application No. 1-278383”.

[0010]

In the above-mentioned conventional projection technique, since there is no distinction between the blood vessel and the background, it is impossible to perform some processing on only the blood vessel. For example,
Conventionally, when displaying a blood vessel image, the background portion is not distinguished even if it is combined with other data and displayed, so that there is a problem that the background portion is displayed overlaid on the other data. Therefore, when attempting to extract blood vessels using conventional extraction techniques, it is difficult to distinguish between blood vessel regions with low density and background regions.Therefore, the threshold background removal method removes blood vessels even if the threshold value is set high. If it is set low, the background will be extracted, which is not good. Also, in region glowing, it is extracted by connecting similar density points around it, but once connecting backgrounds close to the density of blood vessels, there is a problem that background areas are connected one after another, which does not work. .

Therefore, an object of the present invention is to extract only the blood vessel region by accurately distinguishing the blood vessel from the background.

[0012]

In order to achieve the above object, the present method does not directly extract from three-dimensional original data, but after performing projection processing from a certain direction on a two-dimensional projection surface. , The projection data is subjected to an extraction process to extract blood vessels. By the process of once projecting onto two dimensions, the background region that occupies most of the three dimensions is removed, and there is an effect that the target of extraction can be narrowed down to only those that are highly likely to be blood vessels, and more accurate extraction is possible. .
Also, the extraction processing time can be shortened. Then, the blood vessel region is expanded in the three-dimensional voxel data based on the two-dimensional extracted blood vessel. By performing this expansion process, three-dimensional extracted blood vessel data can be obtained from the extracted blood vessels in two dimensions.

The details of each of the above processes will be described below. Regarding the projection process, the conventional method projected only the density information of voxels, whereas this method performs processing to store not only the density but also the position information of that voxel. The projection data is the data in which the density and the position are stored as a pair.

Regarding the extraction processing, the conventional method directly extracts from the original data, whereas the present method divides the projected data using the above-mentioned projection method into regions and distinguishes between blood vessels and background. The blood vessel is extracted by performing the processing described above. The process of splitting is to track all the projected voxels in the projection plane by tracking the connected projected voxels from the position information between the adjacent projected voxels of the projection data and adding the connected projected voxels to the region. A process of dividing each connected region is performed. Next, in the determining process, a process of determining whether each connected region is a blood vessel or a background is performed on the divided region based on the number of connected voxels in the region and the change in the density value.

By performing the above processing, even a blood vessel having a low density that is difficult to distinguish from the background, which has been difficult to extract in the past, can be recognized as a continuous blood vessel from thin blood vessels to dark blood vessels because the continuity of the position is tracked. Can be extracted. In addition, the above process may not be able to extract blood vessels from only one direction, so the same process is performed from another direction and the results are integrated into one, so that it cannot be extracted in one direction. Extraction is possible even if there are free blood vessels.

[0016]

In this method, three-dimensional data is projected into two dimensions, and a series of processes for extracting blood vessels from the two-dimensional projection data is performed not only from one direction but also from multiple directions, and the two-dimensional extracted data is extracted. Three-dimensional extracted data is obtained by integrating the data in three dimensions and expanding the data in three dimensions. The details of each process will be described below.

First, a method of projecting three-dimensional data will be described.

It is considered that voxels on each line of sight viewed from one direction are projected on each projection point on the projection plane for three-dimensional data. Here, considering the case where there are multiple blood vessels on the line of sight, instead of a single vessel, by projecting multiple voxels with a high probability of blood vessels, even if there are overlapping blood vessels on the line of sight, the information of the blood vessels is dropped. It is possible to project without. The maximum number of voxels to be projected is the number of voxels on the line, but since it is not generally necessary to project all voxels on the line, this projected voxel number n (n is usually about 2 to 5) It is set in advance as a parameter of the projection process. In addition, by storing the position of the projected voxel on the line of sight, it is possible to determine the connectivity between adjacent projected voxels necessary for extracting a blood vessel, which will be described later. The above projection process will be described in detail below. First, the voxel having the maximum density value on the line of sight is searched for and projected on the projection point, and then the position of the voxel is also saved. This maximum value is defined as a first maximum value, and the projected projection surface is defined as a first projection surface. This projection plane is
A pair of density information and position information is stored. Next, on the line of sight except for the first maximum value, the maximum value in voxels exceeding the density threshold (projection threshold) of the preset projection processing parameter is set to the second maximum value ( The second maximum value) is projected on the second projection plane, and the voxel position is also stored. Here, when removing a projected voxel from the line of sight, not only one voxel having the maximum value but also a voxel adjacent to it and having a projection threshold value or more are excluded so that the next maximum value does not fall within the same blood vessel. Try to remove the group of voxels (connected voxels). The above processing is performed until there is no value exceeding the projection threshold value, or until the n-th maximum value (n-th maximum value) of the maximum number of projection values and the n-th projection surface.

Next, the blood vessel extraction processing will be described. In the blood vessel extraction process, first, a process of dividing the projection plane into regions is performed, and then the region is identified as a blood vessel or a background and a blood vessel is extracted.

First, the process of dividing the projection plane into regions will be described. Check the connectivity between adjacent projection points based on the position information in the projection plane created by the above projection processing,
The inside of the projection plane is divided for each area by performing a process of collecting the connected projection points as an area. The process of collecting the connected projection points as a region is a process of sequentially tracking those having the connectivity based on the positional information between the adjacent projection points and adding them to the region. The point at which the tracking is started (tracking start point) is selected from the projection points on the first projection plane on which all blood vessels other than the overlapping portion are projected. For each of the projection point in the projection plane adjacent to the tracking start point and the projection point in the second projection plane that is also adjacent, the connectivity with the tracking start point is checked. Perform the process of capturing as the same area. Here, when the adjacent points in the second projection plane are not connected, the adjacent projection points from the third to the n-th projection plane are sequentially examined until the connected projection points exist, and the connection points are connected. When there is a projection point that has a property, the process of adding it to the area is performed. By the above processing, if there is a projection point (new area point) newly added to the area, among the projection points adjacent to the new area point, those that do not belong to any area yet The connectivity with the area is checked, and if there is connectivity, processing to add to the area is performed. By repeating the series of processes described above until all the points on the first projection plane belong to the region, the inside of the projection plane is divided for each connected region.

The conditional expression for determining the connectivity between projection points is shown below.

| P1-p2 | ≦ α (1) where p1 is position information of projection points belonging to the region p2 is position information α of projection points not belonging to and adjacent to the projection point p1 Position difference parameter of a certain connection condition (In general, it is considered that the connection is α = 1 because the difference is 1 voxel.) The above connection condition formula calculates the difference value of the position information between the projection points and α voxel. If it is within the range, it is judged that they are connected. However, in a blood vessel with a high density and a wide width, the positional difference between the adjacent maximum values is not necessarily one voxel. On the other hand, it is necessary to widen the position difference parameter α, which is usually set to 1.

An expression for expanding the positional difference parameter α is shown below.

Α = α + β (β> 0) (2) where β is a positional difference relaxation parameter of the connection condition (β is usually in the range of 1 to 3, if it is too wide, noise may be picked up. The density information (v1 and v2) for both the projection points (p1 and p2) that are the objects of connectivity determination is based on the density threshold of the preset connection condition. If the position exceeds the value parameter, it is determined that the width of the blood vessel is wide, and the process of expanding the position difference parameter of formula (1) is performed according to formula (2).

Next, the process of extracting the blood vessels by distinguishing and dividing the respective connected areas into blood vessels and backgrounds based on the number of connected points and the change in density of each divided connected area on the projection plane will be described.

This method is based on the idea that the blood vessel portion is a continuous area area and the background portion is a noise area, so that there is little continuity in terms of position. It is determined whether or not the blood vessel has sufficient continuity by the following equation by comparing with a certain connection number parameter.

M> γ (3) Here, m is the number from the region start point (v1, p1) to the last region point (vm, pm) γ is the concatenation of preset region discrimination conditions Number parameter (γ may not be a fixed value, but may be variable according to changes in concentration in the region, etc.) More accurately extract blood vessels from a region that satisfies the above-mentioned expression (3) and is determined as a blood vessel candidate. Therefore, it may be determined again whether the background. Generally, since the background portion has little change in density, the minimum value vmin and the maximum value v in the area are used.
max is calculated, and the difference is compared with a preset area concentration difference parameter, and the following equation is used to determine whether the blood vessel candidate area is the background and classify it.

Vmax-vmin> δ (4) Here, δ is a region density difference parameter of a preset region discrimination condition (δ is usually 5% to 2% of the maximum value in the three-dimensional voxel data).
It may be about 0%) In order to more accurately extract the blood vessel in the area determined as the blood vessel candidate as in the above formula (4), it may be determined again whether it is the background. Generally, using the fact that the background part does not change much in density, calculate the average variance in the region, compare the variance with the preset blood vessel region dispersion parameter, and determine whether the blood vessel candidate region is the background by the following formula. You may classify it.

Variance> ε (5) Here, variance is the variance within the region ε is the blood vessel region dispersion parameter of the region discrimination condition that has been set in advance. It becomes possible to extract blood vessels by the method above. However, when projection processing is performed in two dimensions from three-dimensional data, blood vessels running parallel to the projection line are not always projected with continuity maintained. Therefore, for example, the series of processes described above is performed several times from different directions such as front-back, left-right, and up-down directions, and as a result, a process of integrating the blood vessel regions extracted from each direction by two-dimensional or three-dimensional logical sum is performed. By doing so, even if there is a blood vessel that cannot be extracted in one direction due to lack of continuity, it is possible to reliably extract the blood vessel.

Finally, the process of expanding the blood vessel region extracted by the above-mentioned extraction process and integration process in three dimensions will be described.

The extracted blood vessel (three-dimensional blood vessel data) is located in the original blood vessel corresponding to the same position in the three-dimensional voxel data before extraction, but does not include all blood vessel voxels. In order to include voxels of blood vessels in the three-dimensional blood vessel data, the difference in the density value with the corresponding peripheral region in the three-dimensional voxel data before extraction is calculated, and the difference is set in advance in the three-dimensional connection condition. After performing the process of capturing the peripheral region that exceeds the threshold of
Further, the peripheral region of the blood vessel region is weighted according to the distance and re-acquired to perform expansion processing. By performing this process, it is possible to give the extracted blood vessel a volume and smooth the edge changes, and obtain more accurate three-dimensional data of the extracted blood vessel.

[0032]

Embodiments of the present invention will now be described in detail with reference to the drawings.

First, an embodiment of the present invention will be described with reference to FIGS.

This embodiment particularly describes a method relating to a process of extracting a blood vessel region from three-dimensional voxel data in the body imaged by MRI.

FIG. 1 shows an example of the configuration of a system to which the present invention is applied. This device performs a process of extracting a blood vessel by using three-dimensional voxel data obtained by MRI. In FIG. 1, a straight line with an arrow indicates a data transmission path, and an arrow indicates the direction of data flowing between each device.

The three-dimensional voxel data obtained by MRI is taken into the system from the image input / output device 101 and stored in the memory 103. The processing device 102 is
By referring to the data in the memory 103, the projection process of density, position, width information, etc., the extraction process of the blood vessel region from the projection result, the expansion process of the blood vessel region from the extraction result, the extraction expanded blood vessel region and the image input / output device 101 The composite display processing with other data read in is performed. The projection result, the extraction progress, the extraction result, the integration result, the extension progress, the extension result, the synthesis result, etc. are displayed on the CRT 104. The operation input device 105 is used to input parameters such as parameter setting of various processes and cancellation of execution change of processes.

The outline of the blood vessel extraction processing of the present invention will be described below with reference to the system configuration diagram of FIG. 1 and the flowchart of FIG. It should be noted that the projection blood vessel threshold value, the maximum projection value number n, the number of projections and the direction thereof (hereinafter collectively referred to as “projection condition”) in the projection process are set in advance in the program.

(Step 201) The three-dimensional voxel data photographed by MRI is loaded into the memory 103 through the image input / output device 101.

(Step 202) From the input device 105,
Input parameters (determination (1), (2) of connectivity (connection condition), determination of regions (3), (4), (5) (region determination condition)) of extraction process) in the extraction processing.

(Step 203) With reference to the input data in the memory 103, projection processing is performed under preset projection conditions. The projection data created as a result is C
It is displayed on the RT 104. For a detailed explanation of this step,
This is performed in steps 301 to 304 described later.

(Step 204) The projection data created in Step 203 is read, the connectivity is judged based on the positional information of the projection data under the connection condition input in Step 202, and the blood vessel region extraction processing is performed. A detailed description of this step will be given later in steps 401 to 410.

(Step 205) Steps 203 to 204 to obtain projection extraction results from multiple directions.
Is repeatedly performed according to the projection direction and the number thereof input in step 202.

(Step 206) Based on the position information of the extracted blood vessels (two-dimensional blood vessel data) created in Step 203 to Step 205, the extracted blood vessels from each direction are integrated and written in the three-dimensional space on the memory by logical sum. Processing is performed to create three-dimensional blood vessel data.

(Step 207) The result of the projection, extraction and integration process displayed on the CRT 104 is confirmed, and if necessary, the process returns to step 202, and if there is no problem, the process proceeds to the next step 208.

(Step 208) A judgment condition (hereinafter referred to as "expansion condition") for expanding the peripheral region in the blood vessel region expansion process is input from the input device 105.

(Step 209) Referring to the three-dimensional blood vessel data created in the memory 103 in step 206, the blood vessel region is expanded in the three-dimensional voxel data based on the expansion condition input in step 208. The expansion result is projected and displayed on the CRT 104. A detailed description of this step will be given in steps 501 to 502 described later.

(Step 210) The projection result displayed on the CRT 104 is compared with the expanded projection result, and the process returns to Step 208 if necessary. The expansion process from step 208 to step 209 is repeated until the expansion is completed without any problem.

(Step 211) The extracted blood vessel data is displayed on the CRT 104. In this step, detailed description of a case where the extracted organ data and the extracted blood vessel data previously extracted from the three-dimensional voxel data are combined and displayed is given in steps 601 to 609 described later.

The projection blood vessel threshold value, the number of projections and their directions (projection conditions) in the preset projection processing are as follows:
You may input from the input device 105.

In the processing of step 203, if the created projection result is insufficient, the changed projection condition may be input from the input device 105 and the projection processing of step 203 may be repeated.

Instead of writing the extraction result combination of step 206 on the memory, the extraction results from the other directions are combined and displayed on the extraction results in the respective two-dimensional projection planes, and the extraction results are displayed. If not enough, step 2
The extraction process from 03 to step 205 may be repeated.

The three-dimensional blood vessel data created in the synthesis of the extraction result in step 206 is displayed on the CRT 104, and the undesired synthesis area is the input device 1
The deletion process may be performed by giving an instruction from 05.

In the composite display of step 211,
It can be displayed not only by combining with one organ data that has been extracted in advance, but also by combining with many organ data, artificially generated graphics data, body skin, separately captured data, etc. good.

Next, step 203 in the flowchart of FIG.
The projection processing method for creating projection data on the two-dimensional projection plane from the three-dimensional data will be described with reference to FIGS. 3 and 4.

(Step 301) Each pixel (voxel) on the line from the viewpoint to the projection point on the projection surface is regarded as one-dimensional data (FIG. 3A), the maximum value on that line is searched, and i = Set to 1.

(Step 302) The density and position of the maximum value are stored in the i-th projection plane as projection information of the i-th maximum value (1≤i≤n).

(Step 303) Voxels (connecting voxels) adjacent to the maximum value and exceeding the threshold value of the preset projection condition are removed from the line. (FIG. 3 (b)) (Step 304) If i = n or there is no value exceeding the threshold value on the line, the process ends. Otherwise, search for the maximum value on the line and set i = i + 1. The processing is moved to 302.

Next, step 204 in the flowchart of FIG.
A method of extracting a blood vessel region based on the position information of projection data will be described with reference to the flowchart of FIG. Here, in the process of extracting a blood vessel region, a process is performed in which the information in which the position information of adjacent projection points in the projection plane are connected is connected as a region, but the point at which the connection is started (the connection start point) Is performed from the point which is not yet connected to the region on the first projection plane. Here, the process of scanning and extracting the connection start points one by one from the first projection plane is performed, but it is necessary to skip the points already connected to the region. Therefore, a scan table corresponding to the first to n-th projection planes is set in the memory, and a scanning flag is set at a point on the scan table corresponding to a point scanned (scanned) in the area. To do.

(Step 401) First to nth scanning tables corresponding to the first to nth projection planes projected in step 203 are set and initialized as unscanned for all points.

(Step 402) Referring to the first scanning table, an unscanned projection point on the first projection plane is taken out as a region connection start point, and the point in the first scanning table corresponding to that point is determined to have been scanned. To do.

(Step 403) The difference between the above-mentioned projection point (region point) and the unscanned projection points (surrounding points) on the first to n-th projection planes is calculated.

(Step 404) Based on the difference obtained in step 403, peripheral points which can be judged to have high connectivity are fetched as connected regions. A detailed description of this step will be given later in steps 411 to 415.

(Step 405) The point in the scanning table corresponding to the point newly formed as the connected region in step 404 is set as scanned.

(Step 406) If there is a point that becomes a new connected area, step 40 is executed for the new connected point.
The process from 3 to step 405 for fetching the surrounding connecting points is repeated. The process continues to step 407 when no new connection point is generated.

(Step 407) If the connected region created in steps 403 to 406 is the background region, the process proceeds to step 402, and if it is the blood vessel region, the process continues to step 408. A detailed description of this step will be given later in steps 421 to 425.

(Step 408) The created connected area is written in the memory 103 as a blood vessel area.

(Step 409) If all the points in the first scanning table have been scanned, the processing is continued to step 410, and if there is an unscanned point, the processing is moved to step 402.

(Step 410) The extraction area created by the extraction processing of steps 402 to 409 is C
The process of displaying the image on the RT 104 is performed, and a region that is obviously unnecessary in the extraction result is deleted from the extraction table according to an instruction from the input device 105.

In the process of step 402, a projection point taken out from another projection plane instead of the first projection plane may be used as the connection start point.

In the processing of the above step 403, the number and position of the surrounding points for which the difference is calculated may be input from the input device 105, and the projection plane of the surrounding points to be calculated up to the projection surface input from the input device 105. You may limit it.

In the processing of steps 407 to 408 described above, not only the blood vessel region is determined and written in the memory, but the blood vessel region and the background region may be distinguished and written in the memory. Alternatively, the background region may be directly removed from the first to nth projection planes.

In the erasing process of the unnecessary area in step 410, the corresponding area in the extraction result from another direction may be erased based on the position information of the erasing area.

According to the above-described embodiment, by separating and discriminating each connected region, it becomes possible to distinguish the blood vessel region and the background region, which cannot be distinguished by the conventional projection method and extraction method, and perform the processing.

Next, step 404 of the flowchart in FIG.
The connectivity determination processing in FIG. 6 will be described with reference to the flowchart of FIG.

(Step 411) If the density values of the projection information of the area points and the surrounding points which are the objects of connectivity determination are higher than the preset connection threshold value of the connection condition, the process proceeds to step 412. And if low, step 413
Transfer processing to.

(Step 412) If the difference between the positional information of the area point and the surrounding point is within the preset number of pixels for relaxing the difference of the connection condition, the process is moved to step 414, and is not within the number of pixels. Moves the process to step 415.

(Step 413) If the difference between the positional information of the area point and the surrounding point is within 1 pixel, step 4
The process proceeds to step 14, and if it is not within one pixel, the process proceeds to step 415.

(Step 414) It is judged that the peripheral points are connected, and the points are taken into the area.

(Step 415) It is judged that the surrounding points are not connected.

Incidentally, the above steps 412 and 41
The pixel value in the allowable range of the positional difference used as the determination reference in 3 may be any value input from the input device 105. Further, it may be variable based on the density information of the area point to be judged and the surrounding points.

Next, step 407 of the flowchart in FIG.
The process of separating and determining the connected region into the blood vessel region and the background region in step S1 will be described with reference to the flowchart of FIG.

(Step 421) If the number of connected points in the connected area is larger than the number of connected area determination conditions set in advance, the processing is continued to step 422, and if it is smaller, the processing is moved to step 425.

(Step 422) The density MIN and MAX values in the area are calculated from the projected density information in the connected area, and if the difference is higher than the density difference value of the area discrimination condition set in advance, the processing is continued to step 424. If it is lower, the process proceeds to step 425.

(Step 423) The average variance in the region is calculated from the projected density information in the connected region, and if the variance is higher than the preset blood vessel region variance value of the region discriminating condition, the process is continued to step 424 and is low. If so, the process moves to step 425.

(Step 424) The connected area is discriminated as a blood vessel area.

(Step 425) The connected area is discriminated as the background area.

Next, step 209 of the flowchart in FIG.
The expansion processing method of the blood vessel region in the above will be described with reference to the flowchart of FIG.

(Step 501) Steps 202 to 2
The three-dimensional blood vessel data obtained by the projection extraction processing of 07 and the three-dimensional voxel data before the extraction are associated with each other to obtain the peripheral area of the three-dimensional blood vessel data in the three-dimensional voxel data, and the blood vessel data area and the peripheral area. The expansion processing is performed by taking in a peripheral area in which the density difference between and is within the density difference value of the preset expansion condition as a blood vessel area.

(Step 502) The three-dimensional blood vessel data expanded in the above step 501 is expanded by adding the area around the blood vessel area to the blood vessel area by weighting it according to the position.

Next, step 211 in the flowchart in FIG.
The combined display method of displaying the extracted blood vessel data in step (3) and the combined extraction with the previously extracted organ data will be described with reference to the flowchart of FIG.

(Step 601) Extracted organ data extracted in advance is taken into the memory through the image input / output device 101.

(Step 602) 3 before extraction processing
The surface position (skin) of the body is obtained from the three-dimensional voxel data. The surface position of the organ is obtained from the extracted organ data acquired in step 601.

(Step 603) The light amount attenuation parameter in the rendering process is input from the operation input device 105 for each of the blood vessel and the organ, and the initial value of the light amount is also set.

(Step 604) If blood vessel data exists between the skin position determined in step 602 and the surface position of the organ, the process proceeds to step 606, and if not, the process continues to step 605.

(Step 605) The extracted organ data is rendered while the light quantity is attenuated by the organ light quantity attenuation rendering parameter input in step 603 until the remaining light quantity is exhausted or the data ends.

(Step 606) Extracted blood vessel data existing from the position of the skin to the surface position of the extracted organ is rendered by the organ light amount attenuation rendering parameter input in step 603 to attenuate the light amount.

(Step 607) If the remaining amount of light attenuated in step 606 is not present, the process proceeds to step 609. If there is still the remaining amount, the process continues to step 608.

(Step 608) The extracted organ data is rendered while the light quantity is attenuated by the organ light quantity attenuation rendering parameter input in step 603 until the remaining amount runs out or the data ends.

(Step 609) The rendering result is C
Display on RT104.

The above steps 605 and 60
The extracted organ rendering process of 8 may be skipped if the surface of the organ is not detected.

A modified example in which blood vessel width information is added to the series of extraction processing of the blood vessel extraction processing shown in the above embodiment and the processing is performed will be described. As a result, the connection condition relaxation parameter, which is fixed once set, is set to a value based on the width of the blood vessel at that location, so that the blood vessel connectivity can be determined easily and accurately. In addition, when integrating blood vessel extraction results in three dimensions, width information, which is the original number of voxels of the blood vessel, is stored. Therefore, integration based on that information simplifies the expansion processing of the peripheral region. It becomes possible to do.

Step 203 of the above-mentioned flowchart in FIG.
A modified example in which blood vessel width information is added to the projection processing method in (3) will be described with reference to the flowchart in FIG.

(Step 311) Each pixel (voxel) on the line from the viewpoint to the projection point on the projection plane is regarded as one-dimensional data, the maximum value on that line is searched, and i =
Set to 1.

(Step 312) The density and position of the maximum value are stored in the i-th projection plane as projection information of the i-th maximum value (1≤i≤n).

(Step 313) After storing the maximum value and the number of connected voxels (connected voxels) which are adjacent to the maximum value and which exceed a preset projection condition threshold value as the blood vessel width of the projection information, the connected voxels are lined. Remove from above.

(Step 314) If i = n or there is no value exceeding the threshold value on the line, the process is terminated,
Otherwise, the maximum value on the line is searched for, i = i + 1, and the process proceeds to step 312.

Step 404 of the above-mentioned flowchart in FIG.
A modified example in which blood vessel width information is added to the connectivity determination processing in step S1 will be described with reference to the flowchart in FIG.

(Step 431) If the difference between the position information of the area point which is the object of connectivity determination and the surrounding point is within the pixels of the width information of both points, step 432
The process shifts to step 433. If it is not within the pixel of the width information, the process shifts to step 433.

(Step 432) It is judged that the peripheral points are connected, and this is taken into the area.

(Step 433) It is judged that the surrounding points are not connected.

Step 209 of the above-mentioned flowchart in FIG.
A modified example in which blood vessel width information is added to the blood vessel region expansion processing method will be described with reference to the flowchart of FIG.

(Step 511) Steps 202 to 2
The three-dimensional blood vessel data obtained by the projection extraction processing of 07 and the three-dimensional voxel data before extraction are made to correspond to each other, and the peripheral region of the three-dimensional blood vessel data is based on the number of pixels of the blood vessel width information.
Expansion processing is performed by taking in the surrounding area as a blood vessel area.

(Step 512) The three-dimensional blood vessel data expanded in the above step 511 is expanded by adding a weight corresponding to the position to the area surrounding the blood vessel area and incorporating it into the blood vessel area.

The expansion processing of the peripheral area using the blood vessel width information described above may be carried out simultaneously when the data is written in the memory in step 206.

[0115]

According to the present invention, it becomes possible to easily and reliably extract only the blood vessel region from the three-dimensional data obtained by the in-vivo tomographic image pickup device, and the extracted blood vessel It is possible to combine and display other data.

[Brief description of drawings]

FIG. 1 is a diagram showing an example of a configuration of a system for implementing the present invention.

FIG. 2 is a main flowchart of an embodiment of the present invention.

FIG. 3 is an explanatory diagram of projection processing for three-dimensional voxel data.

FIG. 4 is a flowchart of projection processing for three-dimensional voxel data.

FIG. 5 is a flowchart of an extraction process for projection data.

FIG. 6 is a flowchart of processing for determining connectivity in the extraction processing.

FIG. 7 is a flowchart of processing for separating and determining connected areas in the extraction processing.

FIG. 8 is a flowchart of expansion processing for the extracted blood vessel data.

FIG. 9 is a flowchart of a process for displaying the extracted blood vessel data in a combined manner.

FIG. 10 is a flowchart of a modified example in which blood vessel width information is added to the projection process.

FIG. 11 is a flowchart of a modified example in which blood vessel width information is added to the process of determining connectivity.

FIG. 12 is a flowchart of a modified example in which blood vessel width information is added to expansion processing.

[Explanation of symbols]

101 ... Image input / output device, 102 ... Processing device, 103 ...
Memory, 104 ... CRT, 105 ... Operation input device.

Claims (11)

[Claims] 1. For three-dimensional voxel data obtained by using an in-vivo tomographic image pickup device, a density value of at least one voxel on a line viewed from one direction is projected for each pixel on a two-dimensional projection plane. And a process of storing the position information of the voxels used for the projection, and a process of connecting the projected voxels projected based on the position information,
A three-dimensional blood vessel display processing method, characterized by processing for separating a connected area in which connected projected voxels exist into a blood vessel candidate area and a background area based on in-area information, and displaying the separated and discriminated blood vessel candidate area. . 2. A process of projecting and storing from a three-dimensional to a two-dimensional projection plane, a process of projecting a voxel having a maximum value (first maximum value) on a line and storing the position thereof, and a process of storing the position on the line. The maximum value (second maximum value) on the line excluding the group of voxels adjacent to the first maximum value (concatenated voxel) is also projected for voxels above a preset threshold value in and the position is saved. The three-dimensional blood vessel display processing method according to claim 1, wherein the processing and the above processing are repeatedly performed up to the final voxel maximum value (n-th maximum value). 3. The process of connecting projection voxels is a process of relaxing the criterion for determining the connectivity of positions when the densities of both adjacent projection voxels are higher than a preset value. Dimensional blood vessel display processing method. 4. A process of separating a connected region into a blood vessel candidate region and a background region includes a process of classifying a region having a large number of connected voxels (connection points) into a blood vessel candidate and a region having a small number of connected voxels into a background region. The three-dimensional blood vessel display processing method according to claim 1. 5. A process of separating a connected region into a blood vessel candidate region and a background region is performed by obtaining an average variance of density information in the connected region,
2. The three-dimensional blood vessel display processing method according to claim 1, further comprising a process of classifying the one having a variance larger than a predetermined value as a blood vessel region and the one having a variance smaller than a predetermined value for classifying the background region. 6. The process of separating a connected region into a blood vessel candidate region and a background region includes a process of obtaining a maximum value and a minimum value in the connected region and determining a region having a small difference as a background region. 3D blood vessel display processing method. 7. The method according to claim 1, further comprising, after the blood vessel candidate region is separated, based on position information of the blood vessel candidate, density points of three-dimensional voxel data exceeding a threshold value around the blood vessel candidate are taken in as blood vessel candidates. A three-dimensional blood vessel display processing method. 8. The process of displaying the extracted blood vessel data,
2. The three-dimensional blood vessel display processing method according to claim 1, further comprising a processing of combining with another organ that has been extracted in advance and displaying it. 9. The process of synthesizing the extracted blood vessel three-dimensional voxel data and other organs stores the surface position of the skin seen from the line-of-sight direction based on a threshold value as preprocessing, and synthesizes it from the skin position. 9. The three-dimensional blood vessel display processing method according to claim 8, wherein the noise existing outside the skin is removed by starting the calculation, and the calculation range is limited to perform the synthesis processing. 10. The process of synthesizing the extracted blood vessel three-dimensional voxel data and other organs calculates the surface positions of the organs viewed from the line-of-sight direction of a plurality of synthesis target organs (blood vessels, brains, etc.) in advance, The three-dimensional blood vessel display processing method according to claim 8, wherein an organ to be displayed for each line of sight is determined based on the positional relationship, and display calculation is started from the surface position thereof. 11. For three-dimensional voxel data obtained by using an in-vivo tomographic image pickup device, a density value of at least one voxel on a line viewed from one direction is projected for each pixel on a two-dimensional projection plane. And the process of storing, the process of storing the position information of the voxels used for the projection, the process of connecting the projected voxels (projection voxels) based on the position information, and the connected projected voxels (connected region)
Is divided into a blood vessel candidate region and a background region based on the in-region information, and the above-described series of processes is performed on the 3D voxel data.
A three-dimensional blood vessel display processing method, characterized in that after performing from at least two directions within three directions parallel to the coordinate axes, the results are integrated and displayed by a logical sum.