CN110163925A - A kind of method for visualizing and device of long-range projection neuron - Google Patents
A kind of method for visualizing and device of long-range projection neuron Download PDFInfo
- Publication number
- CN110163925A CN110163925A CN201910271433.3A CN201910271433A CN110163925A CN 110163925 A CN110163925 A CN 110163925A CN 201910271433 A CN201910271433 A CN 201910271433A CN 110163925 A CN110163925 A CN 110163925A
- Authority
- CN
- China
- Prior art keywords
- neuron
- fiber
- subsegment
- obtains
- neighborhood
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/003—Reconstruction from projections, e.g. tomography
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T15/00—3D [Three Dimensional] image rendering
- G06T15/005—General purpose rendering architectures
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Graphics (AREA)
- Image Processing (AREA)
Abstract
The invention discloses the method for visualizing and device of a kind of long-range projection neuron, method includes being split the neuron skeletal tree of manual trace to obtain more root section fibres；It extracts the neighborhood three dimensional image signals of the subsegment fiber and projects；The projected image for merging the subsegment fiber obtains the complete image signal of neuron.Device includes the first acquisition module, and the projected image of neuron subsegment fiber is obtained for the neuron skeletal tree based on manual trace；Second obtains module, obtains neuron complete image signal for merging all neuron subsegment fiber projected images.The present invention can facilitate the raw image data of the entire neuron of Neuroscience Research person's quick visualization according to the complete image information of full brain neuron morphological data collection and the single neuron of neuron skeleton file rapidly extracting of manual trace.
Description
Technical field
The invention belongs to Neuscience art of image analysis, more particularly, to a kind of the visual of long-range projection neuron
Change method and device.
Background technique
One of the core objective that neural circuit figure is modern neuro science is drawn, it is dependent on single neuron within the scope of full brain
The fine reconstruction of form and the analysis and research to projecting pattern and Synaptic junction.Fluorescence sparse markup and large volume are fine in recent years
The research of imaging technique makes it possible the acquisition of the full brain neuron morphological data collection of neuromorphic under submicron resolution,
These data provide the intact form information of neuron.Analyzing and studying neuron morphology, projecting pattern and Synaptic junction
When, we usually require the skeleton structure of tracking neuron, additionally need to extract the original image number for having tracked neuron
According to and be able to carry out visualization.The skeleton structure of neuron can react the whole projected path of neuron, in addition to projected path,
Neuroscience Research person also needs to study the original image of neuron, these images can provide the Synaptic junction on fiber
Information, fiber width and fiber gray-scale intensity information of node etc., while also contributing to having helped Neuroscience Research person's detection
Whether the neuron skeleton through manual trace is matched with raw image data.But a set of full mouse brain sub-micron data are usual
Up to several TB to tens TB, while the total length of the neuron of long-range projection can achieve several centimetres to tens centimetres, from
Extracted in the data set of this scale of construction image data of single neuron and visualize it is very troublesome, it is current lack facilitate it is handy
Tool.
Business software Amira is a kind of visual tool of effective 3 d image data, but is directly read without providing
Single Neuronal images data and visual function.Indirect scheme be artificially divide in Amira along neurofibril it is more
The image of part near secondary load fibers.This method cannot once extract all image datas of single neuron, and right
In the neuron of long-range projection, about needs artificially to divide secondary load image data up to a hundred and each load requires to calculate Local map
As the position offset of relatively full brain neuron morphological data collection, therefore it is not easy to very much Neuroscience Research person's use.In addition
Topography not only include further include other neighbouring neurons to visual neuron, this can interfere Current neural member can
Depending on changing.
Summary of the invention
In view of the drawbacks of the prior art, the purpose of the present invention is to provide a kind of visualization sides of long-range projection neuron
Method, it is intended to solve the problems, such as that current three-dimensional large volume image data visualization software is difficult to directly handle long-range projection neuron.
To achieve the above object, the present invention provides a kind of method for visualizing of long-range projection neuron, comprising:
The neuron skeletal tree of manual trace is split to obtain more root section fibres；
It extracts the neighborhood three dimensional image signals of subsegment fiber and projects；
The projected image for merging all subsegment fibers obtains the complete image of neuron.
Preferably, neuron skeletal tree is considered as a binary tree structure, each side of binary tree corresponds to the one of neuron
Root fiber.To single fiber, calculates its length and be split into several subsegment fibers by scheduled subsegment fibre length.Specifically
If T is neuron skeleton tree construction, T={ E1, E2..., EnIndicate neuron skeleton tree construction splitting into n root fiber.Assuming that Ei
Length be li, the length of subsegment fiber is d, then the subsegment fibre number m that this root fiber is spliti=[li/ d], Ei={ S1 i,
S2 i..., Smi iIndicate a fiber splitting into miRoot section fibre, wherein [] indicates to round up.Due to being rounded,
The length of last root section fibre is typically less than equal to d's.
Preferably, the skeleton of subsegment fiber is usually and is formed by connecting by a series of broken line points, in order to extract its neighborhood image
Signal, the present invention first carry out equidistant interpolation to it.Then the spheric region centered on the point is extracted to each equidistant interpolation point.
All spheric regions are grouped together into the cylindrical body picture signal region using skeleton as center line.Specifically set Sj iIndicate mind
Jth root section fibre on i-th fiber through member tree, it is assumed that Sj iLength be lj i, the spacing of interpolation point is r, then needs to insert
The number m of the point of valuej i=[lj i/ r] -1, wherein [] indicates to round up.The spacing of interpolation point is r, while the radius of ball is arranged
For r, then all adjacent ball structures and an approximate cylindrical region is formed into together.
Preferably, according to the corresponding direction xy of projected image of the neighborhood 3-D image of each subsegment fiber relative to full brain
The global position offset of neuron morphology data set exists the projected image splicing of the neighborhood 3-D image of all subsegment fibers
Together.For the place of coincidence, the maximum value in all values is taken.And the big figure for splicing and obtaining is calculated relative to full brain neuron
The position offset in the direction xy of morphological data collection.
Other side according to the invention provides a kind of visualization device of long-range projection neuron, comprising:
First obtains module, and the perspective view of neuron subsegment fiber is obtained for the neuron skeletal tree based on manual trace
Picture；
Second obtains module, obtains the complete image letter of neuron for merging all neuron subsegment fiber projected images
Number.
Preferably, the first acquisition module includes:
Split cells obtains more root section fibres for splitting neuron skeletal tree；
Extraction unit, for extracting the neighborhood three dimensional image signals of the subsegment fiber and projecting.
Contemplated above technical scheme through the invention, can obtain compared with prior art it is following the utility model has the advantages that
1, it is difficult to directly handle the problem of long-range projection neuron for current three-dimensional large volume image viewing software, this
Invention provides a kind of image data quick visualization method of long-range projection neuron, and this method can be quickly according to full cranial nerve
The neighborhood image information of first morphological data collection and the single neuron of neuron skeleton file rapidly extracting of manual trace, facilitates mind
Raw image data through the entire neuron of scientific researcher quick visualization；
2, method proposed by the present invention can be by the 3 d image data of long-range projection neuron from full brain neuron form
It extracts in data set and is visualized in a manner of two-dimensional projection, because this method only extracts neuron skeleton lines neighborhood
3-D image, so other interference of Neuronal images data to Current neural member can be excluded；Long-range projection neuron is complete
The projection scope of intracerebral can reach in cubic millimeter magnitude, the voxel number for covering the 3-D image of single long-range projection neuron
Hundred billion, it is difficult to read and visualize, the present invention is by the three-dimensional image projection of neighborhood at two dimensional image, and processing can be by neuron in this way
Raw image data can be stored entirely in the two dimensional image spliced, easily facilitate Neuroscience Research person visualize it is single
The raw image data of a neuron, and include nerve node required for Neuroscience Research person in the two dimensional image spliced
The information such as structure information, including Synaptic junction node, fiber width and fiber gray-scale intensity.
Detailed description of the invention
Fig. 1 is a kind of flow diagram of the method for visualizing of long-range projection neuron provided in an embodiment of the present invention；
Fig. 2 is that neuron skeleton provided in an embodiment of the present invention splits schematic diagram；
Fig. 3 is the schematic diagram of subsegment fiber neighborhood image signal extraction provided in an embodiment of the present invention；
Fig. 4 is the long-range cone mind that a kind of method for visualizing of long-range projection neuron provided in an embodiment of the present invention extracts
Effect picture through member.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.As long as in addition, technical characteristic involved in the various embodiments of the present invention described below
Not constituting conflict each other can be combined with each other.
The size for the neuron morphology data set that full mouse brain sub-micron is differentiated in the present embodiment is 30380 × 49000 ×
8179(0.2×0.2×1.0um3), about 7.24TB in total.Concentrating in the mouse brain neuron morphological data includes people
The long-range of work tracking projects cone neurone, length about 104um。
It is as shown in Figure 1 a kind of process signal of the method for visualizing of long-range projection neuron provided in an embodiment of the present invention
Scheme, in method shown in Fig. 1, comprising the following steps:
The neuron skeleton of manual trace is split to obtain more root section fibres；
It extracts the neighborhood three dimensional image signals of subsegment fiber and projects；
The projected image for merging all subsegment fibers obtains the complete image signal of neuron.
Specifically, neuron skeletal tree is considered as a binary tree structure, each side of binary tree corresponds to the one of neuron
Root fiber.To single fiber, calculates its length and be split into several subsegment fibers by scheduled subsegment fibre length.Specifically
If T is neuron skeleton tree construction, T={ E1, E2..., EnIndicate neuron skeleton tree construction splitting into n root fiber.Assuming that Ei
Length be li, the length of subsegment fiber is d, then the subsegment fibre number m that this root fiber is spliti=[li/ d], Ei={ S1 i,
S2 i..., Smi iIndicate a fiber splitting into miRoot section fibre, wherein [] indicates to round up.Due to being rounded,
The length of last root section fibre is typically less than equal to d's.D is set as 80um in the present embodiment.As Fig. 2 illustrates nerve
The method for splitting of the skeletal tree of member, control binary tree neuron have 7 fibers, are split as isometric 6 for fiber 1
Subsegment fiber, other fibers are also split according to same method.
Specifically, the skeleton of subsegment fiber is usually and is formed by connecting by a series of broken line points, in order to extract its neighbouring figure
As signal, we first carry out equidistant interpolation to it.Then the spheric region centered on the point is extracted to each equidistant interpolation point.
All spheric regions are grouped together into the cylindrical body picture signal region using skeleton as center line.Specifically set Sj iIndicate mind
Jth root section fibre on i-th fiber through member tree, it is assumed that Sj iLength be lj i, the spacing of interpolation point is r, then needs to insert
The number m of the point of valuej i=[lj i/ r] -1, wherein [] indicates to round up.The spacing of interpolation point is r, while the radius of ball is arranged
For r, then all adjacent balls constitute an approximate cylindrical region.General axis dash forward synaptic knob radius in 0.5um~1um,
Therefore it is 1.5um that radius of a ball r, which is arranged, in we.It not only can guarantee in this way and do not missed potential aixs cylinder knot, but also reduced to the greatest extent and need to identify
Image volume.The resolution ratio of full brain data x/y plane is in 0.2um in the present embodiment, and z, in 1um, therefore is arranged r and is to resolution ratio
The reason of [8 8 3], the direction z takes 3 is that the voxel size in the direction z is bigger than x/y plane.As Fig. 3 illustrates to extract subsegment fiber neighborhood
The method of 3-D image carries out equidistant interpolation to the subsegment fiber that is connected by scatterplot first, and the subsegment fiber in Fig. 3 is from 4 points
19 points are interpolated to, then each point is taken using interpolation spacing as the spheric region picture signal of radius centered on the point, most
The spheric region of all points is combined the cylindrical body picture signal region to be formed using skeleton as center line, i.e. subsegment fiber eventually
Neighborhood three dimensional image signals.
Specifically, according to the global position in the corresponding direction xy of projected image of the neighborhood 3-D image of each subsegment fiber
Offset (relative to full brain neuron morphological data collection) splices the projected image of the neighborhood 3-D image of all subsegment fibers
Together.For the place of coincidence, the maximum value in all values is taken.And the big figure for splicing and obtaining is calculated relative to full cranial nerve
The position offset in the direction xy of first morphological data collection.
The spliced map of 16172 × 16255 (0.2um × 0.2um), the overall situation direction xy position are finally obtained in the present embodiment
Offset is 842.4um and 4207.8um, as shown in Figure 4.The complete image signal of neuron shown in Fig. 4, both it can be seen that
The global pattern of entire neuron, and can see the neighborhood original image of every fiber, including needed for Neuroscience Research person
The information such as Synaptic junction node, fiber width and fiber gray-scale intensity.
As it will be easily appreciated by one skilled in the art that the foregoing is merely illustrative of the preferred embodiments of the present invention, not to
The limitation present invention, any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should all include
Within protection scope of the present invention.
Claims (8)
1. a kind of method for visualizing of long-range projection neuron characterized by comprising
The neuron skeletal tree of manual trace is split to obtain more root section fibres；
It extracts the neighborhood three dimensional image signals of the subsegment fiber and projects；
The projected image for merging the subsegment fiber obtains the complete image signal of neuron.
2. the method as described in claim 1, which is characterized in that the neuron skeletal tree is binary tree structure, the y-bend
Each side of tree is a fiber of neuron.
3. the method as described in claim 1, which is characterized in that the equal length of the subsegment fiber.
4. the method as described in claim 1, which is characterized in that the neighborhood three dimensional image signals for extracting the subsegment fiber
Including carrying out equidistant interpolation to the subsegment fiber, extract region be the spheric region centered on each equidistant interpolation point and
Collection.
5. the method as described in claim 1, which is characterized in that the neighborhood three dimensional image signals for extracting the subsegment fiber
Making the projection plane projected is x/y plane.
6. a kind of device based on the method described in claim 1 to 5 characterized by comprising
First obtains module, and the projected image of neuron subsegment fiber is obtained for the neuron skeletal tree based on manual trace；
Second obtains module, obtains the complete image signal of neuron for merging all neuron subsegment fiber projected images.
7. device as claimed in claim 6, which is characterized in that described first, which obtains module, includes:
Split cells obtains more root section fibres for splitting neuron skeletal tree；
Extraction unit, for extracting the neighborhood three dimensional image signals of the subsegment fiber and projecting.
8. a kind of computer readable storage medium, the computer-readable recording medium storage has computer program, and feature exists
In method of the realization as described in still one in claim 1 to 5 when the computer program is executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910271433.3A CN110163925B (en) | 2019-04-04 | 2019-04-04 | Visualization method and device for long-range projection neuron |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910271433.3A CN110163925B (en) | 2019-04-04 | 2019-04-04 | Visualization method and device for long-range projection neuron |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110163925A true CN110163925A (en) | 2019-08-23 |
| CN110163925B CN110163925B (en) | 2021-01-05 |
Family
ID=67638853
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910271433.3A Active CN110163925B (en) | 2019-04-04 | 2019-04-04 | Visualization method and device for long-range projection neuron |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110163925B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100074486A1 (en) * | 2006-11-22 | 2010-03-25 | Max-Planck-Gesellschaft Zur Forderung Der Wissenschaften E.V., A Corporation Of Germany | Reconstruction and visualization of neuronal cell structures with bright-field mosaic microscopy |
| CN104331892A (en) * | 2014-11-05 | 2015-02-04 | 南京理工大学 | Morphology-based neuron recognizing and analyzing method |
| CN108053391A (en) * | 2017-11-22 | 2018-05-18 | 华中科技大学 | A kind of method for identifying neuron reconstruction errors |
| CN108550147A (en) * | 2018-04-13 | 2018-09-18 | 华中科技大学苏州脑空间信息研究院 | Full cranial nerve beam projected path method for automatic tracking and system |
-
2019
- 2019-04-04 CN CN201910271433.3A patent/CN110163925B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100074486A1 (en) * | 2006-11-22 | 2010-03-25 | Max-Planck-Gesellschaft Zur Forderung Der Wissenschaften E.V., A Corporation Of Germany | Reconstruction and visualization of neuronal cell structures with bright-field mosaic microscopy |
| CN104331892A (en) * | 2014-11-05 | 2015-02-04 | 南京理工大学 | Morphology-based neuron recognizing and analyzing method |
| CN108053391A (en) * | 2017-11-22 | 2018-05-18 | 华中科技大学 | A kind of method for identifying neuron reconstruction errors |
| CN108550147A (en) * | 2018-04-13 | 2018-09-18 | 华中科技大学苏州脑空间信息研究院 | Full cranial nerve beam projected path method for automatic tracking and system |
Non-Patent Citations (3)
| Title |
|---|
| CHEN ZHANG 等: "A Stereological Quantitative Analysis of the Brain-wide Distribution of SOM Neurons in Whole Mouse Brain", 《14TH INTERNATIONAL CONFERENCE ON PHOTONICS AND IMAGING IN BIOLOGY AND MEDICINE》 * |
| 骆清铭: "脑空间信息学—连接脑科学与类脑人工智能的桥梁", 《中国科学：生命科学》 * |
| 龚辉 等: "单神经元分辨水平的小鼠全脑网络可视化", 《生命科学》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110163925B (en) | 2021-01-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Schmitt et al. | New methods for the computer-assisted 3-D reconstruction of neurons from confocal image stacks | |
| CN110853022B (en) | Pathological section image processing method, device and system and storage medium | |
| Chothani et al. | Automated tracing of neurites from light microscopy stacks of images | |
| US8076937B2 (en) | Fibre tracking on the basis of macroscopic information | |
| CN102737250B (en) | To the method and system of spinal injury automatic detection in 3D medical images | |
| CN101027692B (en) | System and method for object characterization of toboggan-based clusters | |
| Banerjee et al. | Semantic segmentation of microscopic neuroanatomical data by combining topological priors with encoder–decoder deep networks | |
| Santamaría-Pang et al. | Automatic morphological reconstruction of neurons from multiphoton and confocal microscopy images using 3D tubular models | |
| Sathyan et al. | Lung nodule classification using deep ConvNets on CT images | |
| Yang et al. | A distance-field based automatic neuron tracing method | |
| Pfister et al. | Visualization in connectomics | |
| Everts et al. | Exploration of the brain’s white matter structure through visual abstraction and multi-scale local fiber tract contraction | |
| CN113781387A (en) | Model training method, image processing method, device, equipment and storage medium | |
| CN115349863A (en) | Magnetoencephalography device and MRI registration method | |
| CN113012129A (en) | System and device for counting area positioning and marked nerve cells of brain slice image | |
| CN106886754B (en) | Object identification method and system under a kind of three-dimensional scenic based on tri patch | |
| Amelio et al. | Skin lesion image segmentation using a color genetic algorithm | |
| CN110163925A (en) | A kind of method for visualizing and device of long-range projection neuron | |
| Su et al. | A novel method for dendritic spines detection based on directional morphological filter and shortest path | |
| Aliaga Maraver et al. | A method for the symbolic representation of neurons | |
| CN111583281B (en) | Brain region dividing method of multi-mode connection mode | |
| Soltanian-Zadeh et al. | 3-D quantification and visualization of vascular structures from confocal microscopic images using skeletonization and voxel-coding | |
| Wang et al. | Cell Image Segmentation of Biomedical Microscopes. | |
| Yin et al. | Three-dimensional super line-localization in low signal-to-noise microscope images via prior-apprised unsupervised learning (PAUL) | |
| CN111429343B (en) | Rapid detection method for branch point in three-dimensional digital image |
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 |