CN110163925B - Visualization method and device for long-range projection neuron - Google Patents

Visualization method and device for long-range projection neuron Download PDF

Info

Publication number
CN110163925B
CN110163925B CN201910271433.3A CN201910271433A CN110163925B CN 110163925 B CN110163925 B CN 110163925B CN 201910271433 A CN201910271433 A CN 201910271433A CN 110163925 B CN110163925 B CN 110163925B
Authority
CN
China
Prior art keywords
neuron
sub
fibers
segment
projection
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
Application number
CN201910271433.3A
Other languages
Chinese (zh)
Other versions
CN110163925A (en
Inventor
骆清铭
龚辉
程胜华
王小俊
刘钰蓉
曾绍群
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huazhong University of Science and Technology
Original Assignee
Huazhong University of Science and Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Huazhong University of Science and Technology filed Critical Huazhong University of Science and Technology
Priority to CN201910271433.3A priority Critical patent/CN110163925B/en
Publication of CN110163925A publication Critical patent/CN110163925A/en
Application granted granted Critical
Publication of CN110163925B publication Critical patent/CN110163925B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

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 a visualization method and a visualization device for long-range projection neurons, wherein the method comprises the steps of splitting an artificially tracked neuron skeleton tree to obtain a plurality of sub-segment fibers; extracting neighborhood three-dimensional image signals of the sub-section fibers and projecting; and combining the projection images of the sub-segment fibers to obtain a complete image signal of the neuron. The device comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for acquiring a projection image of the fiber of the sub-segment of the neuron based on the artificially tracked neuron skeleton tree; and the second acquisition module is used for combining the fiber projection images of all the sub-segments of the neuron to obtain a complete image signal of the neuron. The invention can rapidly extract the complete image information of a single neuron according to the full brain neuron morphological data set and the artificially tracked neuron skeleton file, and is convenient for a neuroscience researcher to rapidly visualize the original image data of the whole neuron.

Description

Visualization method and device for long-range projection neuron
Technical Field
The invention belongs to the field of image analysis of neuroscience, and particularly relates to a visualization method and device for long-range projection neurons.
Background
The mapping of neural circuits is one of the core goals of modern neuroscience, and relies on the fine reconstruction of single neuron morphology and the analytical study of projection patterns and synaptic connections throughout the brain. Recent studies of fluorescent sparse labeling and large-volume fine imaging techniques have enabled the acquisition of a full brain neuronal morphology dataset of neuronal morphology at sub-micron resolution, which provides complete morphological information of neurons. When analyzing and studying neuron morphology, projection patterns and synaptic connections, we usually need to track the skeletal structure of neurons, and furthermore need to extract raw image data already tracked neurons and to be able to visualize. The skeleton structure of the neuron can reflect the whole projection path of the neuron, and besides the projection path, a neuroscience researcher needs to research raw images of the neuron, the images can provide information of synaptic connection nodes on fibers, fiber width information, fiber gray intensity information and the like, and meanwhile, the neuroscience researcher can help to detect whether the artificially tracked neuron skeleton is matched with the raw image data. However, a set of rat brain submicron data is usually as high as several TBs to several tens of TBs, and the total length of a neuron projected in a long range can reach several centimeters to several tens of centimeters, so that it is very troublesome to extract image data of a single neuron from a data set of such a size and visualize the image data, and a convenient and easy-to-use tool is not available at present.
Commercial software Amira is an effective tool for visualization of three-dimensional image data, but does not provide the functionality of directly reading single neuron image data and visualization. An indirect approach is to manually load the image of the local vicinity of the fiber multiple times along the neuron fiber in Amira. This method cannot extract all the image data of a single neuron at a time, and for neurons projected over a long range, image data needs to be loaded manually about hundreds of times and each loading requires calculation of the position offset of a local image with respect to a full brain neuron morphological data set, and thus is very inconvenient for neuroscience researchers to use. In addition, the partial image contains not only the neurons to be visualized but also other neurons in the vicinity, which can interfere with the visualization of the current neuron.
Disclosure of Invention
Aiming at the defects of the prior art, the invention aims to provide a visualization method of a long-range projection neuron, and aims to solve the problem that the current three-dimensional large-volume image data visualization software is difficult to directly process the long-range projection neuron.
In order to achieve the above object, the present invention provides a method for visualizing a long-range projection neuron, comprising:
splitting the artificially tracked neuron skeleton tree to obtain a plurality of sub-segment fibers;
extracting neighborhood three-dimensional image signals of the sub-section fibers and projecting;
the projection images of all sub-segment fibers are combined to obtain a complete image of the neuron.
Preferably, the neuron skeleton tree is regarded as a binary tree structure, and each edge of the binary tree corresponds to one fiber of the neuron. For a single fiber, its length is calculated and split into several sub-segment fibers according to a predetermined sub-segment fiber length. Specifically, let T be a neuron skeleton tree structure, T ═ E1，E2，…，EnDenotes the splitting of the neuron skeletal tree structure into n fibers. Suppose EiHas a length of liAnd the length of the sub-section fiber is d, the number m of the sub-section fibers split by the fiberi＝[li/d]，Ei＝{S1 i，S2 i，…，Smi iMeans breaking a fiber into miA root segment fiber, wherein]Indicating rounding up. For rounding reasons, the length of the fiber of the last subsection is generally less than or equal to d.
Preferably, the skeleton of the sub-segment fiber is generally formed by connecting a series of broken line points, and in order to extract the neighborhood image signal, the invention performs equidistant interpolation on the sub-segment fiber. A spherical region centered on each equidistant interpolation point is then extracted. All the spherical areas are combined together to form a cylindrical image signal area with the skeleton as a central line. Is concretely provided with Sj iRepresents the j root sub-segment fiber on the ith root fiber of the neuron tree, assuming Sj iHas a length of lj iAnd the distance between interpolation points is r, the number of points needing interpolation is mj i＝[lj i/r]-1, wherein]Indicating rounding up. The distance between the interpolation points is r, and the radius of the sphere is r, so that all adjacent spheres are combined to form an approximate cylinder area.
Preferably, the projection images of the neighborhood three-dimensional images of all the sub-segment fibers are spliced together according to the global position offset of the xy direction corresponding to the projection image of the neighborhood three-dimensional image of each sub-segment fiber relative to the whole brain neuron morphological data set. For the place of coincidence, take the maximum of all values. And calculating the position offset of the spliced big graph relative to the xy direction of the whole brain neuron morphological data set.
According to another aspect of the present invention, there is provided a visualization apparatus of a long-range projection neuron, including:
the first acquisition module is used for acquiring a projection image of the sub-segment fiber of the neuron based on the artificially tracked neuron skeleton tree;
and the second acquisition module is used for combining the fiber projection images of all the sub-segments of the neuron to obtain a complete image signal of the neuron.
Preferably, the first obtaining module includes:
the splitting unit is used for splitting the neuron skeleton tree to obtain a plurality of sub-segment fibers;
and the extraction unit is used for extracting the neighborhood three-dimensional image signals of the sub-section fibers and projecting the neighborhood three-dimensional image signals.
Through the technical scheme, compared with the prior art, the invention has the following beneficial effects:
1. aiming at the problem that the current three-dimensional large-volume image visualization software is difficult to directly process long-range projection neurons, the invention provides a method for quickly visualizing image data of the long-range projection neurons, which can quickly extract neighborhood image information of a single neuron according to a whole brain neuron form data set and a manually tracked neuron skeleton file, and is convenient for a neuroscience researcher to quickly visualize original image data of the whole neuron;
2. the method provided by the invention can extract the three-dimensional image data of the long-range projection neuron from the whole brain neuron form data set and visualize the three-dimensional image data in a two-dimensional projection mode, and because the method only extracts the three-dimensional image of the neuron skeleton line neighborhood, the interference of other neuron image data on the current neuron can be eliminated; the projection range of the long-range projection neuron in the whole brain is in the order of cubic millimeter, the number of voxels of a three-dimensional image covering a single long-range projection neuron can reach billions, and the voxel is difficult to read and visualize.
Drawings
Fig. 1 is a schematic flowchart of a visualization method for long-range projection neurons according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a neuron scaffold split provided by an embodiment of the invention;
FIG. 3 is a schematic diagram of sub-segment fiber neighborhood image signal extraction provided by an embodiment of the present invention;
fig. 4 is an effect diagram of a long-range pyramidal neuron extracted by the visualization method for a long-range projection neuron according to the embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In addition, the technical features involved in the embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
The size of the neuron morphological data set for the whole-rat brain submicron resolution in the embodiment is 30380 × 49000 × 8179(0.2 × 0.2 × 1.0 um)3) Total of about 7.24 TB. The morphological data set of mouse brain neurons comprises an artificially tracked long-range projection pyramidal neuron of about 10 mm in length4um。
Fig. 1 is a schematic flow chart of a visualization method for long-range projection neurons according to an embodiment of the present invention, where the method shown in fig. 1 includes the following steps:
splitting the artificially tracked neuron skeleton to obtain a plurality of sub-segment fibers;
extracting neighborhood three-dimensional image signals of the sub-section fibers and projecting;
the projection images of all sub-segment fibers are combined to obtain a complete image signal of the neuron.
Specifically, the neuron skeleton tree is regarded as a binary tree structure, and each edge of the binary tree corresponds to one fiber of the neuron. For a single fiber, its length is calculated and split into several sub-segment fibers according to a predetermined sub-segment fiber length. Specifically, let T be a neuron skeleton tree structure, T ═ E1，E2，…，EnDenotes the splitting of the neuron skeletal tree structure into n fibers. Suppose EiHas a length of liAnd the length of the sub-section fiber is d, the number m of the sub-section fibers split by the fiberi＝[li/d]，Ei＝{S1 i，S2 i，…，Smi iMeans breaking a fiber into miA root segment fiber, wherein]Indicating rounding up. For rounding reasons, the length of the fiber of the last subsection is generally less than or equal to d. In this embodiment d is set to 80 um. As shown in fig. 2, the method for splitting the skeleton tree of the neuron is illustrated, the neuron has 7 fibers compared with the binary tree, the neuron is split into 6 sub-segment fibers with equal length aiming at the fiber 1, and other fibers are split according to the same method.
Specifically, the skeleton of the sub-segment fibers is generally formed by connecting a series of broken line points, and we interpolate the broken line points equidistantly in order to extract the image signals adjacent to the broken line points. A spherical region centered on each equidistant interpolation point is then extracted. All the spherical areas are combined together to form a cylindrical image signal area with the skeleton as a central line. Is concretely provided with Sj iRepresents the j root sub-segment fiber on the ith root fiber of the neuron tree, assuming Sj iHas a length of lj iAnd the distance between interpolation points is r, the number of points needing interpolation is mj i＝[lj i/r]-1, wherein]Indicating rounding up. The distance between the interpolation points is r, and the radius of the sphere is set to be r, so that all adjacent spheres form an approximate cylinder area. The radius of the axonal synaptosomal junction is generally between 0.5um and 1um, so we set the spherical radius r to 1.5 um. This ensures that no potential leakage occursAxon nodules, in turn, minimize the volume of images that need to be identified. In this embodiment, the resolution of the xy plane of the whole brain data is 0.2um, and the resolution in the z direction is 1um, so that r is set to [ 883 ]]The reason why the z direction takes 3 is that the voxel size in the z direction is larger than the xy plane. For example, fig. 3 illustrates a method for extracting a three-dimensional image of a neighborhood of a sub-segment fiber, which includes the steps of performing equidistant interpolation on the sub-segment fibers connected by scattered points, interpolating the sub-segment fibers in fig. 3 from 4 points to 19 points, then taking a spherical region image signal with the point as a center and the interpolation distance as a radius for each point, and finally combining the spherical regions of all the points to form a cylindrical image signal region with a skeleton as a center line, namely, a three-dimensional image signal of a neighborhood of the sub-segment fiber.
Specifically, the projection images of the neighborhood three-dimensional images of all the sub-segment fibers are spliced together according to the global position offset (relative to the whole brain neuron morphological data set) in the xy direction corresponding to the projection image of the neighborhood three-dimensional image of each sub-segment fiber. For the place of coincidence, take the maximum of all values. And calculating the position offset of the spliced big graph relative to the xy direction of the whole brain neuron morphological data set.
In this embodiment, a stitched image of 16172 × 16255(0.2um × 0.2um) is finally obtained, and the global xy direction position offsets of the stitched image are 842.4um and 4207.8um, as shown in fig. 4. The complete image signal of the neuron shown in fig. 4 can see not only the global morphology of the whole neuron, but also the neighborhood original image of each fiber, including information such as synaptic connection nodes, fiber width, and fiber gray intensity required by the neuroscience researcher.
It will be understood by those skilled in the art that the foregoing is only a preferred embodiment of the present invention, and is not intended to limit the invention, and that any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the scope of the present invention.

Claims (6)

1. A method for visualizing long-range projection neurons, comprising:
splitting the artificially tracked neuron skeleton tree to obtain a plurality of sub-segment fibers; the neuron skeleton tree is of a binary tree structure, and each edge of the binary tree is a fiber of a neuron;
extracting neighborhood three-dimensional image signals of the sub-section fibers and projecting; the extraction of the neighborhood three-dimensional image signals of the subsegment fibers comprises equidistant interpolation of the subsegment fibers, and an extraction region is a union set of spherical regions taking each equidistant interpolation point as a center;
and combining the projection images of the sub-segment fibers to obtain a complete image signal of the neuron.
2. The method of claim 1, wherein the sub-segment fibers are of equal length.
3. The method according to claim 1, wherein the projection plane for extracting the neighborhood three-dimensional image signals of the sub-segment fibers for projection is an xy plane.
4. An apparatus based on the method of any one of claims 1 to 3, comprising:
the first acquisition module is used for acquiring a projection image of the sub-segment fiber of the neuron based on the artificially tracked neuron skeleton tree;
and the second acquisition module is used for combining the fiber projection images of all the sub-segments of the neuron to obtain a complete image signal of the neuron.
5. The apparatus of claim 4, wherein the first obtaining module comprises:
the splitting unit is used for splitting the neuron skeleton tree to obtain a plurality of sub-segment fibers;
and the extraction unit is used for extracting the neighborhood three-dimensional image signals of the sub-section fibers and projecting the neighborhood three-dimensional image signals.
6. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1 to 3.
CN201910271433.3A 2019-04-04 2019-04-04 Visualization method and device for long-range projection neuron Active CN110163925B (en)

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 CN110163925A (en) 2019-08-23
CN110163925B true 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)

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2008061548A1 (en) * 2006-11-22 2008-05-29 MAX-PLANCK-Gesellschaft zur Förderung der Wissenschaften e.V. Reconstruction and visualization of neuronal cell structures with brigh-field mosaic microscopy
CN104331892B (en) * 2014-11-05 2017-04-19 南京理工大学 Morphology-based neuron recognizing and analyzing method
CN108053391B (en) * 2017-11-22 2020-06-23 华中科技大学 Method for identifying neuron reconstruction errors
CN108550147B (en) * 2018-04-13 2021-07-06 华中科技大学苏州脑空间信息研究院 Automatic tracking method and system for whole brain nerve bundle projection path

Also Published As

Publication number Publication date
CN110163925A (en) 2019-08-23

Similar Documents

Publication Publication Date Title
Liu et al. MDC-net: A new convolutional neural network for nucleus segmentation in histopathology images with distance maps and contour information
CN111292314B (en) Coronary artery segmentation method, device, image processing system and storage medium
US20100079140A1 (en) Fibre tracking on the basis of macroscopic information
Lee et al. Robust tumor localization with pyramid grad-cam
Li et al. Automated measurement network for accurate segmentation and parameter modification in fetal head ultrasound images
CN112380900A (en) Deep learning-based cervical fluid-based cell digital image classification method and system
GB2472313A (en) Analyzing A Selected Region Of Interest In Medical Image Data
US20060018549A1 (en) System and method for object characterization of toboggan-based clusters
CN101836235A (en) Imaging method for sampling a cross-section plane in a three-dimensional (3d) image data volume
Xie et al. Anisotropic path searching for automatic neuron reconstruction
CN110969623B (en) Lung CT multi-symptom automatic detection method, system, terminal and storage medium
US11471096B2 (en) Automatic computerized joint segmentation and inflammation quantification in MRI
Wu et al. uRP: an integrated research platform for one-stop analysis of medical images
Li et al. SparseTracer: the reconstruction of discontinuous neuronal morphology in noisy images
Rey-Villamizar et al. Large-scale automated image analysis for computational profiling of brain tissue surrounding implanted neuroprosthetic devices using Python
Roudot et al. u-track 3D: measuring and interrogating dense particle dynamics in three dimensions
US20220222812A1 (en) Device and method for pneumonia detection based on deep learning
Nurmaini et al. An improved semantic segmentation with region proposal network for cardiac defect interpretation
CN110163925B (en) Visualization method and device for long-range projection neuron
US20200349706A1 (en) Method and system for detecting chest x-ray thoracic diseases utilizing multi-view multi-scale learning
CN113012129A (en) System and device for counting area positioning and marked nerve cells of brain slice image
Su et al. A novel method for dendritic spines detection based on directional morphological filter and shortest path
Bonda et al. 3D Quantification of vascular-like structures in z stack confocal images
Krentzel et al. CLEM-Reg: An automated point cloud based registration algorithm for correlative light and volume electron microscopy
Guo et al. A novel glomerular basement membrane segmentation using neutrsophic set and shearlet transform on microscopic images

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