CN108670409B - Three-dimensional lung tissue reconstruction and visualization device for surgical planning - Google Patents

Abstract

The present invention relates to a device for three-dimensional reconstruction and visualization of lung tissue for surgical planning, comprising: a computer programmed to perform the following steps: Step 1: receiving an imported CT image in Dicom format, performing segmentation and three-dimensional reconstruction for required The corresponding target tissues are respectively established with corresponding marker maps, wherein the marker maps are based on the read-in image space coordinates, and the marker maps are indexed by integers ranging from 0 to 9, representing different tissues respectively; step 2: the CT image Median filtering is performed on all layers in the CT image; Step 3: In the CT image, the spatial region selection is performed for the part to be observed. The above-mentioned 3D reconstruction and visualization device for pulmonary tissue for surgical planning is specially designed for 3D reconstruction and visualization of pulmonary nodule surgery. The system works stably and reliably, and has strong clinical applicability; the segmentation of trachea, blood vessels and nodules is simple and efficient. The segmentation, the workflow is simple, the operation is simple, and the efficiency is high.

Description

A device for 3D reconstruction and visualization of lung tissue for surgical planning

technical field

The present invention relates to lung tissue, in particular to a three-dimensional reconstruction and visualization device for lung tissue for surgical planning.

Background technique

Due to the high incidence of lung tumors, it is the most common form of multiple tumors, and surgical removal of tumor sites is the main method. Three-dimensional reconstruction and visualization of pulmonary nodules and lung tissue is an important method for preoperative planning. The current 3D reconstruction and visualization methods for surgical planning of pulmonary nodules are mainly based on the developed segmentation algorithms based on CT images, with less clinical validation and development.

Features of the prior art: The three-dimensional reconstruction and visualization of pulmonary nodules and tissues in the prior art mainly focus on the segmentation and reconstruction of the trachea. Segmentation and visualization algorithms and software for whole lung tissue and nodules are mainly achieved by integrating existing segmentation methods. Existing software includes Mimics and OsiriX, which segment the trachea using the region growing method. Slicer, users need to independently select and combine segmentation methods to segment different lung tissues and nodules. And DeepInSight, which uses the method of regional growth to segment and reconstruct each tissue.

References are as follows:

Olav Leira,Frank Lindseth,Toril Anita Nagelhus Hernes,Tore Amundsen,Hanne Sorger,and Thomas

.2015.“Airway Segmentation and CenterlineExtraction from Thoracic CT-Comparison of a New Method to State of the ArtCommercialized Methods.”PLoS ONE 10(12):1–20.doi:10.1371/journal.pone.0144282.Reynisson,Pall Jens,Marta Scali,Erik Smistad,Erlend Fagertun Hofstad,

Olav Leira, Frank Lindseth, Toril Anita Nagelhus Hernes, Tore Amundsen, Hanne Sorger, and Thomas

.2015. “Airway Segmentation and Centerline Extraction from Thoracic CT-Comparison of a New Method to State of the ArtCommercialized Methods.” PLoS ONE 10(12):1–20.doi:10.1371/journal.pone.0144282.

The traditional technology has the following technical problems:

General-purpose software such as mimics has a variety of segmentation algorithms to choose and use, but there is no special segmentation for lung tissues, organs and nodules. The segmentation and reconstruction operations are complicated, and clinical application is difficult. Large amount and low efficiency, it is difficult to achieve rapid segmentation, reconstruction and visualization of lesions and pulmonary organ tissues before pulmonary nodule surgery; the stability of the existing special software is not high, the operation is cumbersome, and the reconstruction of local areas cannot be achieved. with visualization.

SUMMARY OF THE INVENTION

Based on this, it is necessary to provide a three-dimensional reconstruction and visualization device for pulmonary tissue for surgical planning, specifically for the three-dimensional reconstruction and visualization of pulmonary nodule surgery, in view of the above technical problems, the system works stably and reliably, and has strong clinical applicability; Simple and efficient segmentation is adopted for the segmentation of trachea, blood vessels and nodules, with simple workflow, simple operation and high efficiency; it can segment and visualize local areas.

A device for three-dimensional reconstruction and visualization of lung tissue for surgical planning, comprising: a computer programmed to perform the steps of:

Step 1: Receive the imported CT image in Dicom format, and establish a corresponding marker map for the target tissue of the desired segmentation and 3D reconstruction, wherein the marker map is based on the read-in image space coordinates, and the value is an integer from 0 to 9. Index the marker graphs to represent different organizations;

Step 2: perform median filtering on all layers of the CT image;

Step 3: in the CT image, perform spatial region selection for the part to be observed;

Step 4: Within the optional threshold range of -1024 pixel value to 1023 pixel value, select a typical area in the target area to be segmented and reconstructed, analyze the pixel distribution, and calculate the mean and variance; based on the mean of the pixel distribution, set Determine the segmentation threshold, and generate a marker map corresponding to the target area; when the variance is greater than 50, adjust the specific pixel values in the neighborhood of the mean plus or minus 50 pixel values, observe and obtain the region segmentation result; otherwise, Adjust the specific pixel value with the actual pixel value, observe and obtain the regional segmentation result; perform the above operation on all layers of the CT image;

Step 5: In the divided area, select one or several marker point sets Λ0 as the starting value, select the connected area in the three-dimensional space based on the marker points, and use 6, 18 or 26 connected areas to search, and the output is: In the processed marked map, in the search process, based on the 6, 18 or 26 connected areas, find the marked pixels around the pixels, and combine the found marked points with the initial marked point set Λ0 to obtain a new set of pixel points Λi, (i=1, .

Step 6: Perform 3D reconstruction on the segmentation result Λn, if it does not meet the requirements, repeat steps 3, 4 and 5 to perform automatic segmentation again.

In another embodiment, 0 is background, 1 is trachea, 2 is vein, 3 is artery, and 4-10 is tumor.

In another embodiment, the calculated radius of the median filter ranges from 1 pixel to 3 pixels.

In another embodiment, the step "in the CT image, select the spatial region for the part to be observed;" specifically includes: constructing two planes P2 and P3 perpendicular to P1 based on the plane P1 of the CT image , select the spatial region Ω in the three orthogonal planes P1-P3.

A computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, the following steps are implemented:

Step 1: Receive the imported CT image in Dicom format, and establish a corresponding marker map for the target tissue of the desired segmentation and 3D reconstruction, wherein the marker map is based on the read-in image space coordinates, and the value is an integer from 0 to 9. Index the marker graphs to represent different organizations;

Step 2: perform median filtering on all layers of the CT image;

Step 3: in the CT image, perform spatial region selection for the part to be observed;

Step 4: Within the optional threshold range of -1024 pixel value to 1023 pixel value, select a typical area in the target area to be segmented and reconstructed, analyze the pixel distribution, and calculate the mean and variance; based on the mean of the pixel distribution, set Determine the segmentation threshold, and generate a marker map corresponding to the target area; when the variance is greater than 50, adjust the specific pixel values in the neighborhood of the mean plus or minus 50 pixel values, observe and obtain the region segmentation result; otherwise, Adjust the specific pixel value with the actual pixel value, observe and obtain the regional segmentation result; perform the above operation on all layers of the CT image;

Step 5: In the divided area, select one or several marker point sets Λ0 as the starting value, select the connected area in the three-dimensional space based on the marker points, and use 6, 18 or 26 connected areas to search, and the output is: In the processed marked map, in the search process, based on the 6, 18 or 26 connected areas, find the marked pixels around the pixels, and combine the found marked points with the initial marked point set Λ0 to obtain a new set of pixel points Λi, (i=1, .

Step 6: Perform 3D reconstruction on the segmentation result Λn, if it does not meet the requirements, repeat steps 3, 4 and 5 to perform automatic segmentation again.

In another embodiment, 0 is background, 1 is trachea, 2 is vein, 3 is artery, and 4-10 is tumor.

In another embodiment, the calculated radius of the median filter ranges from 1 pixel to 3 pixels.

In another embodiment, the step "in the CT image, select the spatial region for the part to be observed;" specifically includes: constructing two planes P2 and P3 perpendicular to P1 based on the plane P1 of the CT image , select the spatial region Ω in the three orthogonal planes P1-P3.

The above-mentioned 3D reconstruction and visualization device for lung tissue used for surgical planning is specially designed for 3D reconstruction and visualization of pulmonary nodule surgery. The system works stably and reliably, and has strong clinical applicability; the segmentation of trachea, blood vessels and nodules is simple and efficient. The segmentation is simple, the workflow is simple, the operation is simple, and the efficiency is high; it can segment and visualize the local area.

Description of drawings

FIG. 1 is a flow chart executed by a computer in an apparatus for three-dimensional reconstruction and visualization of lung tissue for surgical planning according to an embodiment of the present application.

FIG. 2 is a schematic diagram of searching for connected regions (6, 18 or 26 connected regions) in a device for three-dimensional reconstruction and visualization of lung tissue for surgical planning according to an embodiment of the present application.

Detailed ways

In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, but not to limit the present invention.

A device for three-dimensional reconstruction and visualization of lung tissue for surgical planning, comprising: a computer programmed to perform the steps of:

Step 1: Receive the imported CT image in Dicom format, and establish a corresponding marker map for the target tissue of the desired segmentation and 3D reconstruction, wherein the marker map is based on the read-in image space coordinates, and the value is an integer from 0 to 9. Index the marker graphs to represent different organizations;

Step 2: perform median filtering on all layers of the CT image;

Step 3: in the CT image, perform spatial region selection for the part to be observed;

Step 4: Within the optional threshold range of -1024 pixel value to 1023 pixel value, select a typical area in the target area to be segmented and reconstructed, analyze the pixel distribution, and calculate the mean and variance; based on the mean of the pixel distribution, set Determine the segmentation threshold, and generate a marker map corresponding to the target area; when the variance is greater than 50, adjust the specific pixel values in the neighborhood of the mean plus or minus 50 pixel values, observe and obtain the region segmentation result; otherwise, Adjust the specific pixel value with the actual pixel value, observe and obtain the regional segmentation result; perform the above operation on all layers of the CT image;

Step 5: In the divided area, select one or several marker point sets Λ0 as the starting value, select the connected area in the three-dimensional space based on the marker points, and use 6, 18 or 26 connected areas to search, and the output is: In the processed marked map, in the search process, based on the 6, 18 or 26 connected areas, find the marked pixels around the pixels, and combine the found marked points with the initial marked point set Λ0 to obtain a new set of pixel points Λi, (i=1, .

Step 6: Perform 3D reconstruction on the segmentation result Λn, if it does not meet the requirements, repeat steps 3, 4 and 5 to perform automatic segmentation again.

In another embodiment, 0 is background, 1 is trachea, 2 is vein, 3 is artery, and 4-10 is tumor.

In another embodiment, the calculated radius of the median filter ranges from 1 pixel to 3 pixels.

In another embodiment, the step "in the CT image, select the spatial region for the part to be observed;" specifically includes: constructing two planes P2 and P3 perpendicular to P1 based on the plane P1 of the CT image , select the spatial region Ω in the three orthogonal planes P1-P3. The spatial region Ω is the range region for subsequent calculation processing.

A computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, the following steps are implemented:

Step 1: Receive the imported CT image in Dicom format, and establish a corresponding marker map for the target tissue of the desired segmentation and 3D reconstruction, wherein the marker map is based on the read-in image space coordinates, and the value is an integer from 0 to 9. Index the marker graphs to represent different organizations;

Step 2: perform median filtering on all layers of the CT image;

Step 3: in the CT image, perform spatial region selection for the part to be observed;

Step 4: Within the optional threshold range of -1024 pixel value to 1023 pixel value, select a typical area in the target area to be segmented and reconstructed, analyze the pixel distribution, and calculate the mean and variance; based on the mean of the pixel distribution, set Determine the segmentation threshold, and generate a marker map corresponding to the target area; when the variance is greater than 50, adjust the specific pixel values in the neighborhood of the mean plus or minus 50 pixel values, observe and obtain the region segmentation result; otherwise, Adjust the specific pixel value with the actual pixel value, observe and obtain the regional segmentation result; perform the above operation on all layers of the CT image;

Step 5: In the divided area, select one or several marker point sets Λ0 as the starting value, select the connected area in the three-dimensional space based on the marker points, and use 6, 18 or 26 connected areas to search, and the output is: In the processed marked map, in the search process, based on the 6, 18 or 26 connected areas, find the marked pixels around the pixels, and combine the found marked points with the initial marked point set Λ0 to obtain a new set of pixel points Λi, (i=1, .

Step 6: Perform 3D reconstruction on the segmentation result Λn, if it does not meet the requirements, repeat steps 3, 4 and 5 to perform automatic segmentation again.

The above steps are used in combination to achieve segmentation reconstruction of the trachea and arteries, venous vessels, and ribs.

In another embodiment, 0 is background, 1 is trachea, 2 is vein, 3 is artery, and 4-10 is tumor.

In another embodiment, the calculated radius of the median filter ranges from 1 pixel to 3 pixels.

In another embodiment, the step "in the CT image, select the spatial region for the part to be observed;" specifically includes: constructing two planes P2 and P3 perpendicular to P1 based on the plane P1 of the CT image , select the spatial region Ω in the three orthogonal planes P1-P3.

The above-mentioned 3D reconstruction and visualization device for lung tissue used for surgical planning is specially designed for 3D reconstruction and visualization of pulmonary nodule surgery. The system works stably and reliably, and has strong clinical applicability; the segmentation of trachea, blood vessels and nodules is simple and efficient. The segmentation is simple, the workflow is simple, the operation is simple, and the efficiency is high; it can segment and visualize the local area.

The technical features of the above-described embodiments can be combined arbitrarily. For the sake of brevity, all possible combinations of the technical features in the above-described embodiments are not described. However, as long as there is no contradiction between the combinations of these technical features, All should be regarded as the scope described in this specification.

The above-mentioned embodiments only represent several embodiments of the present invention, and the descriptions thereof are specific and detailed, but should not be construed as a limitation on the scope of the invention patent. It should be pointed out that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can also be made, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention should be subject to the appended claims.

Claims (1)

1. A device for three-dimensional reconstruction and visualization of lung tissue for surgical planning, comprising: a computer programmed to perform the following steps: Step 1: Receive the imported CT image in Dicom format, and establish a corresponding marker map for the target tissue required for segmentation and 3D reconstruction. Index the marker graphs to represent different organizations; Step 2: perform median filtering on all layers of the CT image; Step 3: in the CT image, perform spatial region selection for the part to be observed; Step 4: Within the optional threshold range of -1024 pixel value to 1023 pixel value, select a typical area in the target area to be segmented and reconstructed, analyze the pixel distribution, and calculate the mean and variance; based on the mean of the pixel distribution, set Determine the segmentation threshold, and generate a marker map corresponding to the target area; wherein, when the variance is greater than 50, adjust the segmentation threshold in the neighborhood of plus or minus 50 pixel values of the mean, observe and obtain the region segmentation result; otherwise , perform regional segmentation according to the set segmentation threshold, observe and obtain the regional segmentation result; perform the above operations on all layers of the CT image; Step 5: In the divided area, a number of marked point sets Λ0 are used as the starting value, and the connected areas in the three-dimensional space are selected based on the marked points, and 6, 18 or 26 connected areas are used to search, and the output is processed. Marking graph, in the search process, based on 6, 18 or 26 connected areas, find the marked pixels around the pixels, and combine the found marked points with the initial marked point set Λ0 to obtain a new pixel point set Λi, where , i=1,...,n, where i is the number of iterative calculation steps; the search is repeated until all marked points are found, and the pixel point set Λn of the segmentation result is obtained; Step 6: Perform three-dimensional reconstruction on the pixel point set Λn of the segmentation result, if it does not meet the requirements, repeat steps 3, 4 and 5 for automatic segmentation again; The calculated radius of the median filter is 1 pixel to 3 pixels; The step "in the CT image, perform spatial region selection for the part to be observed" specifically includes: based on the plane P1 of the CT image, constructing two planes P2 and P3 that are perpendicular to P1, and three orthogonal planes P1-P3 are constructed. Select the spatial region Ω within the plane.

Images