CN108550147B - Automatic tracking method and system for whole brain nerve bundle projection path - Google Patents
Automatic tracking method and system for whole brain nerve bundle projection path Download PDFInfo
- Publication number
- CN108550147B CN108550147B CN201810329472.XA CN201810329472A CN108550147B CN 108550147 B CN108550147 B CN 108550147B CN 201810329472 A CN201810329472 A CN 201810329472A CN 108550147 B CN108550147 B CN 108550147B
- Authority
- CN
- China
- Prior art keywords
- point
- path
- brain
- calculating
- signal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 210000004556 brain Anatomy 0.000 title claims abstract description 66
- 238000000034 method Methods 0.000 title claims abstract description 51
- 210000005036 nerve Anatomy 0.000 title claims abstract description 27
- 238000004364 calculation method Methods 0.000 claims abstract description 20
- 230000010354 integration Effects 0.000 claims abstract description 8
- 238000012800 visualization Methods 0.000 claims abstract description 8
- 238000007781 pre-processing Methods 0.000 claims abstract description 7
- 238000000638 solvent extraction Methods 0.000 claims abstract description 7
- 210000003792 cranial nerve Anatomy 0.000 claims abstract 2
- 239000011159 matrix material Substances 0.000 claims description 32
- 238000002610 neuroimaging Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 8
- 238000001914 filtration Methods 0.000 claims description 7
- 239000003550 marker Substances 0.000 claims description 7
- 230000011218 segmentation Effects 0.000 claims description 7
- 238000000605 extraction Methods 0.000 claims description 6
- 230000000903 blocking effect Effects 0.000 claims description 3
- 241000700605 Viruses Species 0.000 description 5
- 238000003384 imaging method Methods 0.000 description 5
- 238000002347 injection Methods 0.000 description 5
- 239000007924 injection Substances 0.000 description 5
- 238000011160 research Methods 0.000 description 5
- 238000000399 optical microscopy Methods 0.000 description 4
- 238000013507 mapping Methods 0.000 description 3
- 210000002569 neuron Anatomy 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000037361 pathway Effects 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 208000014644 Brain disease Diseases 0.000 description 1
- 108700019146 Transgenes Proteins 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000003925 brain function Effects 0.000 description 1
- 238000001493 electron microscopy Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000007850 fluorescent dye Substances 0.000 description 1
- 238000001215 fluorescent labelling Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 238000002595 magnetic resonance imaging Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 210000000337 motor cortex Anatomy 0.000 description 1
- 210000004126 nerve fiber Anatomy 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 230000014511 neuron projection development Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012634 optical imaging Methods 0.000 description 1
- 210000002442 prefrontal cortex Anatomy 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000010845 search algorithm Methods 0.000 description 1
- 230000008054 signal transmission Effects 0.000 description 1
- 210000003863 superior colliculi Anatomy 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G06T5/70—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20021—Dividing image into blocks, subimages or windows
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30016—Brain
Abstract
The invention provides a method and a system for automatically tracking a projection path of a global cranial nerve bundle, which comprises the following steps: preprocessing data; extracting a signal; partitioning a data space; calculating the signal density; calculating a projection path; path integration and visualization. The invention can automatically track the projection path of the nerve bundle for the whole brain data set, and the calculation result can be highly superposed with the signal, thereby having high accuracy. And calculating the vectorized nerve bundle path, so that the brain region connection relation can be quantitatively described.
Description
Technical Field
The invention relates to a biomedical image processing technology, in particular to an automatic tracking method and system for a whole brain nerve bundle projection path of brain imaging data.
Background
To understand the brain's working mechanisms, it is necessary to understand the connectivity (training) between different brain regions, nuclei and even neurons in the brain. These connective pathways, which are considered pathways for signal transmission in the brain, are also the structural basis for understanding brain function and brain disease. Although it is not to be suspected, the brain junction mapping research is a hotspot of the current brain science research and is also a difficult point, and the acquisition of the whole brain junction mapping is an important part in the brain junction mapping research.
In recent years, magnetic resonance imaging, optical microscopy and electron microscopy have been widely used for brain connectivity studies and represent the dominant techniques for macroscopic, mesoscopic and microscopic level brain connectivity studies, respectively. Among these, optical microscopy imaging with single neuron resolution and high throughput properties is considered a powerful tool for studying brain connectivity. In combination with specific labeling techniques (such as transgene fluorescence labeling or virus injection labeling), optical microscopy imaging has acquired a vast amount of data of whole brain junction structures. However, the display and analysis of the brain junction result often depends on maximum projection or neuron manual tracking, and an effective quantitative description of the whole brain junction is lacking.
Taking whole brain imaging after injecting the neurite outgrowth marker virus into a specific brain region of a mouse as an example, the optical microscopic imaging can obtain the connection projection relation from an injection point to different regions in the whole brain range. In 2014, a fast marching algorithm (fast marching) adopted by an Allen brain research institute in the U.S. has calculated projection paths of the whole mouse brain from a virus injection point to other brain regions, and some important connection topological relations are obtained at a voxel resolution of 100 microns. In 2016, the Deisseroth team in the United states shows the tracking result of the projection path from the medial prefrontal cortex of a mouse to other brain areas based on a structure tensor method, and nerve connection relation is analyzed in local brain areas with optical imaging resolution. However, both methods fail to simultaneously and effectively exert the advantages of optical microscopy imaging high resolution and whole brain imaging for researching neural connection relationship.
Therefore, the invention aims to provide a method and a system for automatically tracking the whole brain nerve projection path so as to obtain the whole brain connection relationship and provide a brain connection map research service.
Disclosure of Invention
The invention aims to provide a method and a system for automatically tracking a projection path of a global brain nerve bundle, which can acquire a global brain connection relation.
A method for automatically tracking the projection path of the whole brain nerve bundle comprises the following steps:
s1, a data preprocessing step, namely, separating a brain contour on a brain imaging projection image, applying contour information to a corresponding original image, and setting the gray level of an area outside the contour as 0;
s2, a signal extraction step, namely obtaining and subtracting background noise from the image from which the noise outside the brain contour is removed through convolution calculation, filtering after removing the influence of the background noise, and then performing binarization by using threshold segmentation to obtain a foreground signal;
s3, a data space blocking step, namely dividing a three-dimensional data space into a plurality of data blocks with equal size;
s4, a signal density calculation step, namely calculating the signal density of each data block, wherein the signal density is the ratio of the number of pixels occupied by the signals in the data block to the total number of pixels of the data block, and a signal density matrix is obtained;
s5, calculating a projection path, namely calculating a time cost matrix from the starting point to other positions according to the signal density matrix, obtaining a vector field of a search direction of tracing the point in the space to the starting point according to the time cost matrix, and calculating the projection path by using an iterative search method;
and S6, path integration and visualization, namely converting all the calculated connection paths of the starting point and other positions into a standard tree structure to be stored as a marker file, wherein the starting point serves as a root node, and the paths are gradually added as branches.
Further, in step S2, the image and an approximately estimated grayscale threshold value for distinguishing the foreground from the background are subjected to a small operation, and then are subjected to multiple convolutions with the mean template to obtain a noise background image, and the noise background image is subtracted from the original image.
Further, in step S3, the size of the data block is set according to the resolution requirement.
Further, in step S3, the size of the data block is in the order of 10 microns square.
Further, in step S5, a time cost matrix is calculated by using a multi-modal fast marching method, and 26 neighborhood iterations are used.
Further, the step S5 includes:
when a multi-mode fast marching method is used, information provided by points on a diagonal line needs to be considered, 6 modes exist, time cost under the 6 modes needs to be calculated for a certain point in a space respectively, and the minimum value is taken as the time cost of the point;
and calculating the unit vector of the point pointing to the point with the minimum time cost in the neighborhood of 26 according to the cost matrix, thereby obtaining a vector field of the searching direction traced back to the starting point.
Further, for any point in the space where there is a signal and the signal density is higher than a certain threshold, an iterative search is performed along the search direction to obtain a path from the point to the starting point.
Further, the threshold value is set according to the requirement of distinguishing whether the signal exists at the point.
Further, in the step S6, the bifurcation point is accurately located by using a bisection method.
Further, the step S6 includes:
when calculating the path, the position of the next point to be searched is determined by the position of the current point and the searching direction, and the path adding process is as follows: after a bifurcation point between the current path and the existing path is found, taking the bifurcation point as a father node, and adding the path between the bifurcation point and the terminal point;
the idea of binary search is adopted to locate the intersection, and the process is as follows:
step S61, calculating the distance between the middle point of the current path and the existing path;
step S62, if the distance between the point and the existing path is less than a certain threshold, the point is considered to be intersected with the existing path, and the distance between the middle point of the starting point and the existing path is continuously calculated;
step S63, if the distance between the point and the existing path is larger than the certain threshold, the point is considered not intersected with the existing path, and the distance between the point and the middle point of the end point and the existing path is continuously calculated;
step S64, repeating step S62 and step S63 until a bifurcation point is found.
The invention also discloses an automatic tracking system for the projection path of the whole brain nerve bundle, which comprises:
the data preprocessing unit is used for separating the brain contour on the brain imaging projection drawing, applying contour information to a corresponding original image and setting the gray level of an area outside the contour to be 0;
the signal extraction unit is used for obtaining and subtracting background noise from the image from which the noise outside the brain contour is removed through convolution calculation, filtering after removing the influence of the background noise, and then performing binarization by using threshold segmentation to obtain a foreground signal;
the data space partitioning unit is used for partitioning the three-dimensional data space into a plurality of data blocks with equal size;
the signal density calculating unit is used for calculating the signal density of each data block, and the signal density is the ratio of the number of pixels occupied by the signals in the data block to the total number of pixels of the data block to obtain a signal density matrix;
the projection path calculation unit is used for calculating a time cost matrix from the starting point to other positions according to the signal density matrix, obtaining a vector field of a search direction traced back to the starting point by a point in space according to the time cost matrix, and calculating a projection path by using an iterative search method;
and the path integration and visualization unit is used for converting all the connection paths of the starting point and other positions obtained by calculation into a standard tree structure and storing the standard tree structure as a marker file, wherein the starting point is used as a root node, and the paths are gradually added as branches.
The invention has the advantages that: the automatic tracking of the nerve beam projection path can be carried out on the whole brain data set, the calculation result can be highly coincided with the signal, and the accuracy is high. And calculating the vectorized nerve bundle path, so that the brain region connection relation can be quantitatively described.
Drawings
Fig. 1 is a schematic flow chart of the automatic tracking method of the whole brain nerve bundle projection path according to the present invention.
Fig. 2 is a schematic diagram of the multi-modal fast marching method, and fig. (a) - (f) show 6 modes, respectively.
FIG. 3 is a graph showing the calculation results of the projection path of the mouse whole brain nerve bundle obtained according to the present invention, in which FIG. 3(a) is a sagittal plane view and FIG. 3(b) is a horizontal plane view.
FIG. 4 is a block diagram of an automatic tracking system for a global brain nerve bundle projection path according to the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and embodiments, and fig. 1 is a schematic flow chart illustrating an automatic tracking method for a global brain nerve bundle projection path according to the present invention. The automatic tracking method of the whole brain nerve bundle projection path specifically comprises the following steps:
step S1, a data preprocessing step, namely, separating a brain contour on the brain imaging projection drawing, applying contour information to a corresponding original image, and setting the gray level of an area outside the contour as 0;
specifically, a brain contour is divided from a projection image of brain imaging, and is applied to a corresponding original image, the gray scale of a pixel in an area outside the contour is assigned to be 0, and the influence of noise outside the contour in the image is removed.
And step S2, a signal extraction step, namely, for the image from which the noise outside the brain contour is removed, obtaining and subtracting the background noise through convolution calculation, filtering after removing the influence of the background noise, and then performing binarization by using threshold segmentation to obtain a foreground signal.
Specifically, the image and an approximately estimated gray threshold value for distinguishing the foreground and the background are subjected to small operation, and are subjected to multiple convolution with the mean template to obtain a background noise image, and the original image is used for subtracting the background noise image to remove the influence of noise on the background. After gaussian filtering, binarization is performed using automatic threshold segmentation, thereby obtaining a foreground signal.
Step S3, data space blocking step, dividing the three-dimensional data space into several data blocks with equal size.
Specifically, the size of the data block may be set according to the resolution requirement. In this embodiment, the size of the data block is on the order of 10 microns square.
Step S4, a signal density calculating step, which calculates the signal density of each data block, where the signal density is the ratio of the number of pixels occupied by the signal in the data block to the total number of pixels in the data block, and obtains a signal density matrix.
And step S5, a projection path calculation step, namely calculating a time cost matrix from the starting point to other positions according to the signal density matrix, obtaining a vector field of a search direction of tracing the point in the space to the starting point according to the time cost matrix, and calculating the projection path by using an iterative search method.
Specifically, a multi-modal fast marching method is used for calculating a time cost matrix from a starting point to each other point in the whole brain according to the signal density matrix. The fast marching method tracks the signal propagation position by solving the equation of the function, which is as follows:
where T is the time cost of each point, Γ0Is the initial position, i.e., the starting point, and it is assumed here that the propagation velocity is related only to the signal density, and F is the signal density.
In the three-dimensional space, the device is provided with a plurality of channels,
the numerical calculation method comprises the following steps:
wherein the content of the first and second substances,
the standard backward and forward finite difference is adopted, and the second-order finite difference is selected, so that the following equation is solved:
wherein:
in the multi-modal fast marching method, considering the information provided by the points on the diagonal, there are 6 modes, as shown in fig. 2.
In isotropic space, Δ x ═ Δ y ═ Δ z ═ h, for all modes, we can simplify to:
wherein the content of the first and second substances,
gv(h) the values of (A) are shown in the following table:
| modality | g1(h) | g2(h) | g3(h) |
| S1 | 9/4h2 | 9/4h2 | 9/4h2 |
| S2 | 9/4h2 | 9/8h2 | 9/8h2 |
| S3 | 9/4h2 | 9/8h2 | 9/8h2 |
| S4 | 9/4h2 | 9/8h2 | 9/8h2 |
| S5 | 9/8h2 | 9/12h2 | 9/12h2 |
| S6 | 9/8h2 | 9/12h2 | 9/12h2 |
For a certain point, the time cost of 6 modalities is calculated respectively, and the minimum value is taken as the time cost of the point. During the whole iteration process of the multi-modal fast marching method, each grid point x is assigned one of the following three labels:
tag 2, temporary point, point x may change in value;
the value of tag 3, unknown point, point x has not been calculated.
The iterative process is as follows:
initially, the starting point is marked as a known point, the time cost of the points in the 26-neighborhood is calculated and marked as a temporary point:
step S51, finding the minimum value in all the point sets with temporary labels, and changing the labels to known points;
step S52, finding out the neighboring point of the label in the 26-neighborhood of the point as the unknown point or the temporary point;
step S53, calculating the time cost of these points and updating their values;
in step S54, the process returns to step S51.
And calculating projection paths from the starting point to other positions according to the cost matrix, and calculating a unit vector of a point in the space, wherein the point points to the point with the minimum time cost in the neighborhood of 26 according to the cost matrix, so as to obtain a vector field traced back to the searching direction of the starting point. For any point in space where there is a signal and the signal density is higher than a certain threshold (the threshold needs to be able to distinguish whether there is a signal at the point), iterative search is performed along the search direction to obtain a path from the point to the starting point, and the search algorithm may select a fourth-order Runge-Kutta method, etc., where the fourth-order Runge-Kutta method is selected, as shown below:
K1＝f(pi-1)
K4＝f(pi-1+hK3)
wherein p isiIs the position of the search point and f is the search direction. If it falls within the voxel, then tri-linear interpolation is used, h being the search step.
And step S6, path integration and visualization, wherein all the calculated connection paths of the starting point and other positions are converted into a standard tree structure and stored as a label file, the starting point is used as a root node, and the paths are gradually added as branches.
Specifically, when calculating the path, the position of the next point to be searched is determined by the position of the current point and the search direction, so that as long as two paths coincide at a certain point, the following paths both coincide, so the path adding process is as follows: and after the bifurcation point of the current path and the existing path is found, taking the bifurcation point as a father node, and adding the path between the bifurcation point and the terminal point.
The invention adopts the idea of dichotomy search to position the cross point, and the specific process is as follows:
step S61, calculating the distance between the middle point of the current path and the existing path;
step S62, if the distance between the point and the existing path is less than a certain threshold, the point is considered to be intersected with the existing path, and the distance between the middle point of the starting point and the existing path is continuously calculated;
step S63, if the distance between the point and the existing path is larger than the certain threshold, the point is considered not intersected with the existing path, and the distance between the point and the middle point of the end point and the existing path is continuously calculated;
step S64, repeating steps S62 and S63 until a bifurcation point is found, i.e. the path between the point and the starting point coincides with the existing path, but the path between the point and the ending point does not coincide with the existing path.
The threshold is typically very small.
And storing the result obtained by path integration as a mark file, and loading the mark file into corresponding software for visualization. In an embodiment of the present invention, the imaging data of the mouse whole brain neurofibrillary virus marker is taken as an example, the injection point of the virus is set as the starting point, the three-dimensional space is divided into data blocks of 10 × 10 × 10 μm, the signal density threshold is set to 0.01, and the nerve bundle projection path in the whole brain range is calculated, and the result is shown in fig. 3. As can be seen, there is a significant nerve fiber connection from the site of marker injection to the brain region, such as the superior colliculus, motor cortex, etc.
As shown in fig. 4, the present invention further discloses an automatic tracking system for projection path of global brain nerve bundle, comprising:
the data preprocessing unit is used for separating the brain contour on the brain imaging projection drawing, applying contour information to a corresponding original image and setting the gray level of an area outside the contour to be 0;
the signal extraction unit is used for obtaining and subtracting background noise from the image from which the noise outside the brain contour is removed through convolution calculation, filtering after removing the influence of the background noise, and then performing binarization by using threshold segmentation to obtain a foreground signal;
the data space partitioning unit is used for partitioning the three-dimensional data space into a plurality of data blocks with equal size;
the signal density calculating unit is used for calculating the signal density of each data block, and the signal density is the ratio of the number of pixels occupied by the signals in the data block to the total number of pixels of the data block to obtain a signal density matrix;
the projection path calculation unit is used for calculating a time cost matrix from the starting point to other positions according to the signal density matrix, obtaining a vector field of a search direction tracing the starting point from a point in the space according to the time cost matrix, and then calculating the projection path by using an iterative search method;
and the path integration and visualization unit is used for converting all the connection paths of the starting point and other positions obtained by calculation into a standard tree structure and storing the standard tree structure as a marker file, wherein the starting point is used as a root node, and the paths are gradually added as branches.
The present invention has been illustrated by the above embodiments, but it should be understood that the above embodiments are for illustrative and descriptive purposes only and are not intended to limit the invention to the scope of the described embodiments. Furthermore, it will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, and that many variations and modifications may be made in accordance with the teachings of the present invention, which variations and modifications are within the scope of the present invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (11)
1. A method for automatically tracking the projection path of a global cranial nerve bundle is characterized by comprising the following steps:
step S1, a data preprocessing step, namely, separating a brain contour on the brain imaging projection drawing, applying contour information to a corresponding original image, and setting the gray level of an area outside the contour as 0;
step S2, a signal extraction step, namely, obtaining and subtracting background noise from the image from which the noise outside the brain contour is removed through convolution calculation, filtering after removing the influence of the background noise, and then performing binarization by using threshold segmentation to obtain a foreground signal;
step S3, data space blocking step, dividing the three-dimensional data space into a plurality of data blocks with equal size;
step S4, a signal density calculation step, in which the signal density of each data block is calculated, and the signal density is the ratio of the number of pixels occupied by the signals in the data block to the total number of pixels of the data block, so as to obtain a signal density matrix;
step S5, a projection path calculation step, namely calculating a time cost matrix from the starting point to other positions according to the signal density matrix, obtaining a vector field of a search direction of tracing the point in the space to the starting point according to the time cost matrix, and calculating the projection path by using an iterative search method;
and step S6, path integration and visualization, namely converting all the calculated connecting paths of the starting point and other positions into a standard tree structure to be stored as a label file, wherein the starting point is used as a root node, and the paths are gradually added as branches.
2. The method for automatically tracking the projection path of the global brain nerve bundle as claimed in claim 1, wherein in step S2, the image is subjected to a small operation with an approximately estimated gray threshold for distinguishing the foreground from the background, and then subjected to a plurality of convolutions with the mean template to obtain a noise background image, and the noise background image is subtracted from the original image.
3. The method for automatically tracking the projection path of the whole brain nerve bundle according to claim 1, wherein in step S3, the size of the data block is set according to the resolution requirement.
4. The method for automatically tracking the projection path of the global brain nerve bundle according to claim 1, wherein in step S3, the size of the data block is in the order of 10 μm square.
5. The method for automatically tracking the projection path of the global brain nerve bundle according to claim 1, wherein in step S5, a time cost matrix is calculated by a multi-modal fast marching method, and 26 neighborhood iterations are used.
6. The method for automatically tracking the projection path of the global brain nerve bundle according to claim 5, wherein the step S5 includes:
when a multi-mode fast marching method is used, information provided by points on a diagonal line needs to be considered, 6 modes exist, time cost under the 6 modes needs to be calculated for a certain point in a space respectively, and the minimum value is taken as the time cost of the point;
and calculating the unit vector of the point pointing to the point with the minimum time cost in the neighborhood of 26 according to the cost matrix, thereby obtaining a vector field of the searching direction traced back to the starting point.
7. The method of claim 6, wherein for any point in space where there is a signal and the signal density is higher than a threshold, an iterative search is performed along the search direction to find the path from the point to the starting point.
8. The method of claim 7, wherein the threshold is set according to the need of distinguishing whether there is a signal at the point.
9. The method for automatically tracking the projection path of the global brain nerve bundle according to claim 1, wherein in step S6, the bifurcation point is accurately located by using a bisection method.
10. The method for automatically tracking the projection path of the global brain nerve bundle according to claim 9, wherein the step S6 includes:
when calculating the path, the position of the next point to be searched is determined by the position of the current point and the searching direction, and the path adding process is as follows: after a bifurcation point between the current path and the existing path is found, taking the bifurcation point as a father node, and adding the path between the bifurcation point and the terminal point;
the idea of binary search is adopted to locate the intersection, and the process is as follows:
step S61, calculating the distance between the middle point of the current path and the existing path;
step S62, if the distance between the point and the existing path is less than a certain threshold, the point is considered to be intersected with the existing path, and the distance between the middle point of the starting point and the existing path is continuously calculated;
step S63, if the distance between the point and the existing path is larger than the certain threshold, the point is considered not intersected with the existing path, and the distance between the point and the middle point of the end point and the existing path is continuously calculated;
and step S64, repeating the step S62 and the step S63 until the bifurcation point is found.
11. An automatic tracking system for projection path of global brain nerve bundle, comprising:
the data preprocessing unit is used for separating the brain contour on the brain imaging projection drawing, applying contour information to a corresponding original image and setting the gray level of an area outside the contour to be 0;
the signal extraction unit is used for obtaining and subtracting background noise from the image from which the noise outside the brain contour is removed through convolution calculation, filtering after removing the influence of the background noise, and then performing binarization by using threshold segmentation to obtain a foreground signal;
the data space partitioning unit is used for partitioning the three-dimensional data space into a plurality of data blocks with equal size;
the signal density calculating unit is used for calculating the signal density of each data block, and the signal density is the ratio of the number of pixels occupied by the signals in the data block to the total number of pixels of the data block to obtain a signal density matrix;
the projection path calculation unit is used for calculating a time cost matrix from the starting point to other positions according to the signal density matrix, obtaining a vector field of a search direction traced back to the starting point by a point in space according to the time cost matrix, and calculating a projection path by using an iterative search method;
and the path integration and visualization unit is used for converting all the connection paths of the starting point and other positions obtained by calculation into a standard tree structure and storing the standard tree structure as a marker file, wherein the starting point is used as a root node, and the paths are gradually added as branches.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810329472.XA CN108550147B (en) | 2018-04-13 | 2018-04-13 | Automatic tracking method and system for whole brain nerve bundle projection path |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810329472.XA CN108550147B (en) | 2018-04-13 | 2018-04-13 | Automatic tracking method and system for whole brain nerve bundle projection path |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108550147A CN108550147A (en) | 2018-09-18 |
| CN108550147B true CN108550147B (en) | 2021-07-06 |
Family
ID=63514998
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810329472.XA Active CN108550147B (en) | 2018-04-13 | 2018-04-13 | Automatic tracking method and system for whole brain nerve bundle projection path |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108550147B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110163925B (en) * | 2019-04-04 | 2021-01-05 | 华中科技大学 | Visualization method and device for long-range projection neuron |
| CN113470073A (en) * | 2021-07-06 | 2021-10-01 | 浙江大学 | Animal center tracking method based on deep learning |
| CN117648538B (en) * | 2024-01-29 | 2024-03-29 | 小舟科技有限公司 | Training and application method, device, equipment and medium of electroencephalogram signal denoising model |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107256408A (en) * | 2017-05-12 | 2017-10-17 | 常州大学 | A kind of critical path searching method of cerebral function network |
-
2018
- 2018-04-13 CN CN201810329472.XA patent/CN108550147B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107256408A (en) * | 2017-05-12 | 2017-10-17 | 常州大学 | A kind of critical path searching method of cerebral function network |
Non-Patent Citations (2)
| Title |
|---|
| 双斑蟋蟀触角传入神经纤维在脑内投射分布研究;武健文等;《现代医学生物进展》;20121230;第12卷(第29期);第5443-5446页 * |
| 基于MOST技术的小鼠全脑数据集三维配准方法研究;徐正超;《万方学术期刊》;20151016;1 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108550147A (en) | 2018-09-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Türetken et al. | Reconstructing curvilinear networks using path classifiers and integer programming | |
| CN108550147B (en) | Automatic tracking method and system for whole brain nerve bundle projection path | |
| CN111256594B (en) | Method for measuring physical characteristics of surface state of aircraft skin | |
| Chothani et al. | Automated tracing of neurites from light microscopy stacks of images | |
| US20210358101A1 (en) | Processing image data sets | |
| CN111768411B (en) | Coronary centerline extraction method, device, computer equipment and storage medium | |
| US20070036436A1 (en) | Method for improved image segmentation | |
| Rakhmadi et al. | Connected component labeling using components neighbors-scan labeling approach | |
| CN108648256B (en) | Grayscale rock core three-dimensional reconstruction method based on super-dimension | |
| Li et al. | SparseTracer: the reconstruction of discontinuous neuronal morphology in noisy images | |
| CN110349170B (en) | Full-connection CRF cascade FCN and K mean brain tumor segmentation algorithm | |
| Huang et al. | Weakly supervised learning of 3D deep network for neuron reconstruction | |
| CN112801075B (en) | Automatic rural road boundary line extraction method based on aerial image | |
| Koyuncu et al. | Object‐oriented segmentation of cell nuclei in fluorescence microscopy images | |
| CN116523898A (en) | Tobacco phenotype character extraction method based on three-dimensional point cloud | |
| Ghahremani et al. | NeuroConstruct: 3D reconstruction and visualization of neurites in optical microscopy brain images | |
| Liou et al. | A parallel technique for signal-level perceptual organization | |
| Drees et al. | Hierarchical random walker segmentation for large volumetric biomedical images | |
| CN103871089A (en) | Image superpixel meshing method based on fusion | |
| Luengo-Oroz et al. | Methodology for reconstructing early zebrafish development from in vivo multiphoton microscopy | |
| Wang et al. | Memory and time efficient 3D neuron morphology tracing in large-scale images | |
| CN116452604A (en) | Complex substation scene segmentation method, device and storage medium | |
| CN116452826A (en) | Coal gangue contour estimation method based on machine vision under shielding condition | |
| Basu et al. | Retracted: Localizing and extracting filament distributions from microscopy images | |
| CN114626513A (en) | Neuron reconstruction method and device based on whole brain optical imaging data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |