CN105046690A - Organ sectional image digitizing method - Google Patents

Abstract

The present invention provides an organ sectional image digitizing method. The method comprises the steps of scanning to obtain an organ image; based on the differential principle, segmenting the organ image into at least two layers of sectional images, and marking the sectional images; dividing the sectional images of each layer into at least four rectangular grid cells by a grid division method; based on the differential principle, marking the rectangular grid cells; based on the differential principle, marking at least one pixel of the rectangular grid cells; utilizing a binary number to mark the color of the pixel; using the sectional images, the rectangular grid cells, the pixels of the rectangular grid cells, the color sets of the pixels to form the digitizing mark of the organ image. According to the present invention, the digitization expression of the organ sectional image is realized by the differential principle, and the operation, processing and analysis of the organ image big data are convenient.

Description

Organ fault image digitizing solution

Technical field

The present invention relates to medical science fault image process field, particularly a kind of organ fault image digitizing solution.

Background technology

BigBrain project, from 2003, completed the three-dimensional collection of illustrative plates of high-resolution human brain in 2013.This project goes through 10 years, cost 1,000,000,000 Euros, but this just starts.At the beginning of 2013, EU Committee announces, " human brain engineering " is selected in " following emerging flagship technological project ", obtains the subsidy for research of 1,000,000,000 Euros.And then, White House discloses " advancing innovation Neural Technology brain project " (being called for short " brain plan "), estimates to drop into 3,000,000,000 dollars in 10 years.IBM promises to undertake investment 1,000,000,000 dollars of commercializations for its cognitive computing platform Watson; Corporate investment of Baidu 3.5 hundred million dollars carries forward vigorously " Baidu's brain " project ...2015, " Chinese brain plan " fermented startup, will launch from understanding brain, protection brain and simulation brain three directions.In brain research, brain model is set up in primary element task.

In the reconstruction of brain model, in order to understand the relation between brain morphosis and function, brain model needs less tomography layer distance and larger image resolution.In time doing like this, data volume just becomes increasing, and final large to being undertaken analyzing, cannot processing on a common computer by main software, this makes human organ become " large data ".The large data of organ fault image can only rely on supercomputer, as the three-dimensional reconstruction work of BigBrain, enable Germany and the Canadian multiple stage supercomputer several years consuming time be exactly an illustration.Even like this, still have individual problem: the ability of supercomputer is also limited, as in BigBrain project, researchist can with the resolution scan brain region of 1 micron.But the another one collection of illustrative plates completed with high resolving power like this, even about 2 hundred million hundred million byte datas---current state-of-the-art supercomputer is also difficult to effectively process so large data volume by creation.This brings huge challenge to the large data processing of organ fault image.But even the resolution of 1 micron also will stop at the observation of organelle, nanometer will be only the main yardstick of cell, molecular morphology research.

Supercomputer belongs to scarce resource, and the programming for supercomputer is a professional very strong job, and during the machine of costliness, expense also hampers the development process that brain model is rebuild.But if the physical model that neither one is concrete, the research of the fundamental research of the cognitive principle of brain, brain major disease and class brain artificial intelligence will lack necessary " propping material ", because morphology provides the most reliable factual evidence.

How to break away from the large data processing of human organ to the dependence of supercomputer and to break the limited processing power bottleneck of existing supercomputer be a gordian technique urgently broken through.As can be seen here, setting up a kind of is can be the problem needing when implementing " brain plan ", " brain engineering " to solve for the method for common computer process large for organ fault image data transformations.

Publication number be the Chinese invention patent of CN102096106B disclose a kind of based on spatial gridding algorithm containing trap-up isoline drawing practice, comprise the steps: the spatial gridding containing tomography and trap-up: point, interpolation are sought in initialization, the gridding of interpolation point; Contour tracing containing tomography and trap-up: grid edge is marked, line, drawing process end at tomography.Employing spatial gridding algorithm carries out gridding to layer bit data, adopts Kriging regression algorithm to follow the trail of spatial gridding data, and the isoline realized containing tomography and trap-up is drawn, and does not consider pixel and the color of drawing.

Summary of the invention

Technical matters to be solved by this invention is: the organ fault image digitizing solution providing a kind of parameterisable.

In order to solve the problems of the technologies described above, the technical solution used in the present invention is:

A kind of organ fault image digitizing solution, described method is:

Scanning obtains organ image;

Based on Differential Principle, organ Image Segmentation is become at least two-layer fault image, and fault image is marked;

The rectangular node unit of the fault image one-tenth at least four of every layer is divided with Meshing Method;

Based on Differential Principle, rectangular node unit is marked;

Based on Differential Principle, the pixel of at least one of rectangular node unit is marked;

The color of binary number to pixel is utilized to mark;

Described fault image, rectangular node unit, the pixel of rectangular node unit, the color set of pixel are formed the digitized markers of described organ image.

Beneficial effect of the present invention is: with Differential Principle, organ image is divided at least two-layer fault image, the fault image of every layer carries out gridding division and obtains rectangular node unit, rectangular node unit is again with the element marking of at least one, the feature color value of pixel marks, fault image, rectangular node unit, pixel, color value all mark by numeral, be convenient to the digitizing realizing organ image, be convenient to the image computer process that rapidity and convenience realize this large data of organ image.

Accompanying drawing explanation

Fig. 1 is organ fault image digitizing solution process flow diagram of the present invention;

Fig. 2 is the schematic diagram that the organ Image Segmentation of the brain of the embodiment of the present invention one becomes fault image;

Fig. 3 is that the gridding of the fault image of the organ image of the brain of the embodiment of the present invention one divides schematic diagram.

Embodiment

By describing technology contents of the present invention in detail, realized object and effect, accompanying drawing is coordinated to be explained below in conjunction with embodiment.

The design of most critical of the present invention is: utilize Differential Principle, organ image is divided at least two-layer fault image, fault image is carried out gridding division, color value corresponding to the grid cell, the pixel of grid cell and the pixel that fault image, fault image gridding are obtained after dividing represents with scale-of-two, fault image, grid cell, pixel, color value all one_to_one corresponding, organ image indirectly formula is convenient to accurately to express and analyze, be convenient to realize the accurate expression to organ image, be convenient to the digitizing realizing organ image.

The explanation of technical terms that the present invention relates to is in table 1:

Table 1

Please refer to Fig. 1, the specific embodiment of the present invention is:

A kind of organ fault image digitizing solution, described method is:

Scanning obtains organ image;

Based on Differential Principle, organ Image Segmentation is become at least two-layer fault image, and fault image is marked;

The rectangular node unit of the fault image one-tenth at least four of every layer is divided with Meshing Method;

Based on Differential Principle, rectangular node unit is marked;

Based on Differential Principle, the pixel of at least one of rectangular node unit is marked;

The color of binary number to pixel is utilized to mark;

Described fault image, rectangular node unit, the pixel of rectangular node unit, the color set of pixel are formed the digitized markers of described organ image.

From foregoing description, beneficial effect of the present invention is: with Differential Principle, organ image is divided at least two-layer fault image, the fault image of every layer carries out gridding division and obtains rectangular node unit, rectangular node unit is again with the element marking of at least one, the feature color value of pixel marks, fault image, rectangular node unit, pixel, color value all represent by numeral, be convenient to the digitizing that rapidity and convenience realize organ image, be convenient to the image computer process realizing this large data of organ image.

Further, described organ fault image digitizing solution is:

Scanning obtains organ image;

Based on Differential Principle, organ Image Segmentation is become at least two-layer fault image, and fault image is marked, be specially:

$organ=\underset{k}{\Σ}tomography\left(k\right);$

Wherein, organ is digitized organ image, and tomography is one deck fault image of organ image, and k is the level number of fault image;

The rectangular node unit of the fault image one-tenth at least four of every layer is divided with Meshing Method;

Based on Differential Principle, rectangular node unit is marked, is specially:

$tomography=\underset{m}{\Σ}\underset{n}{\Σ}grid(m,n);$

Wherein, grid is the rectangular node unit of fault image, and m is the line number of rectangular node unit, and n is the row number of rectangular node unit;

Based on Differential Principle, the pixel of at least one of rectangular node unit is marked, is specially:

$grid=\underset{i}{\Σ}\underset{j}{\Σ}pexel(i,j);$

Wherein, pixel is the pixel of the rectangular node unit of fault image, and i is the line number of pixel, and j is the row number of pixel;

Utilize the color of binary number to pixel to mark, be specially:

The color of each pixel of organ image is represented with 24 bits respectively, and 24 bits is divided into low eight-digit binary number a, middle eight-digit binary number b and high eight-bit binary number c, be specially:

$\left\{\begin{array}{c}a=colour\\ b=colour/256\\ c=colour/65536\end{array}\right.;$

$pixel=\underset{a}{\Σ}\underset{b}{\Σ}\underset{c}{\Σ}colour(a,b,c);$

Wherein, colour is the color value of pixel;

Described fault image, rectangular node unit, the pixel of rectangular node unit, the color set of pixel are formed the digitized markers of described organ image, are specially:

$organ=\underset{k}{\Σ}tomography\left(\underset{m}{\Σ}\underset{n}{\Σ}grid\right(\underset{i}{\Σ}\underset{j}{\Σ}pixel\left(\underset{a}{\Σ}\underset{b}{\Σ}\underset{c}{\Σ}colour\right(a,b,c)\left)\right)).$

Seen from the above description, fault image, rectangular node unit, pixel, color value all decimally or binary number represent, Differential Principle is adopted fault image, rectangular node unit, pixel, color value to be showed by different level, data form reasonable simplicity, be convenient to the digitizing realizing organ image, the image being convenient to realize this large data of organ image processes on a common computer, is convenient to the image computer process that rapidity and convenience realize this large data of organ image.

Please refer to Fig. 1, embodiments of the invention one are:

A kind of organ fault image digitizing solution is:

Please refer to Fig. 2, obtain the organ image of brain by CT scan;

Based on Differential Principle, the organ Image Segmentation of brain is become 7404 layers of fault image, the thickness of every layer of fault image is 0.02 millimeter, marks, be specially fault image:

$organ=\underset{k}{\Σ}tomography\left(k\right);$

Wherein, organ is digitized organ image, and tomography is one deck fault image of organ image, and k is the level number of fault image, k=1,2 ..., 7404;

Please refer to Fig. 3, the fault image dividing in Cerebrum image every layer with Meshing Method becomes 16X16=256 rectangular node unit;

Based on Differential Principle, rectangular node unit is marked, is specially:

$tomography=\underset{m}{\Σ}\underset{n}{\Σ}grid(m,n);$

Wherein, grid is the rectangular node unit of fault image, and m is the line number of rectangular node unit, m=1,2 ..., 16, n is the row number of rectangular node unit, n=1,2 ..., 16;

Based on Differential Principle, the pixel of at least one of rectangular node unit is marked, in Cerebrum image, the width of fault image of every layer is 6560 pixels, be highly 5696 pixels, the fault image of every layer is divided into 256 rectangular node unit, then the width of each rectangular node unit is 410 pixels, is highly 356 pixels, is specially:

$grid=\underset{i}{\Σ}\underset{j}{\Σ}pixel(i,j);$

Wherein, pixel is the pixel of the rectangular node unit of fault image, and i is the line number of pixel, i=1,2 ..., 356, j is the row number of pixel, j=1,2 ..., 410;

Utilize the color of binary number to pixel to mark, be specially:

The color of each pixel of organ image is represented with 24 bits respectively, and 24 bits is divided into low eight-digit binary number a, middle eight-digit binary number b and high eight-bit binary number c, be specially:

$\left\{\begin{array}{c}a=colour\\ b=colour/256\\ c=colour/65536\end{array}\right.;$

$pixel=\underset{a}{\Σ}\underset{b}{\Σ}\underset{c}{\Σ}colour(a,b,c);$

Wherein, colour is the color value of pixel;

Described fault image, rectangular node unit, the pixel of rectangular node unit, the color set of pixel are formed the digitized markers of described organ image, are specially:

$organ=\underset{k=1}{\overset{7404}{\Σ}}tomography\left(\underset{m=1}{\overset{16}{\Σ}}\underset{n=1}{\overset{16}{\Σ}}grid\right(\underset{i=1}{\overset{356}{\Σ}}\underset{j=1}{\overset{410}{\Σ}}pixel\left(\underset{a=1}{\overset{8}{\Σ}}\underset{b=1}{\overset{8}{\Σ}}\underset{c=1}{\overset{8}{\Σ}}colour\right(a,b,c)\left)\right)).$

In sum, organ fault image digitizing solution provided by the invention, with Differential Principle, organ image is divided at least two-layer fault image, the fault image of every layer carries out gridding division and obtains rectangular node unit, rectangular node unit is again with the element marking of at least one, the feature color value of pixel marks, fault image, rectangular node unit, pixel, color value all decimally or binary number represent, adopt Differential Principle by fault image, rectangular node unit, pixel, color value shows by different level, data form reasonable simplicity, be convenient to the digitizing realizing organ image, the image being convenient to realize this large data of organ image processes on a common computer, be convenient to realize computer disposal fast and easily.

The foregoing is only embodiments of the invention; not thereby the scope of the claims of the present invention is limited; every equivalents utilizing instructions of the present invention and accompanying drawing content to do, or be directly or indirectly used in relevant technical field, be all in like manner included in scope of patent protection of the present invention.

Claims (2)

1. an organ fault image digitizing solution, is characterized in that, described method is:

Scanning obtains organ image;

Based on Differential Principle, organ Image Segmentation is become at least two-layer fault image, and fault image is marked;

The rectangular node unit of the fault image one-tenth at least four of every layer is divided with Meshing Method;

Based on Differential Principle, rectangular node unit is marked;

Based on Differential Principle, the pixel of at least one of rectangular node unit is marked;

The color of binary number to pixel is utilized to mark;

Described fault image, rectangular node unit, the pixel of rectangular node unit, the color set of pixel are formed the digitized markers of described organ image.

2. organ fault image digitizing solution according to claim 1, it is characterized in that, described method is:

Scanning obtains organ image;

Based on Differential Principle, organ Image Segmentation is become at least two-layer fault image, and fault image is marked, be specially:

$organ=\underset{k}{\Σ}tomography\left(k\right);$

Wherein, organ is digitized organ image, and tomography is one deck fault image of organ image, and k is the level number of fault image;

The rectangular node unit of the fault image one-tenth at least four of every layer is divided with Meshing Method;

Based on Differential Principle, rectangular node unit is marked, is specially:

$tomography=\underset{m}{\Σ}\underset{n}{\Σ}grid(m,n);$

Wherein, grid is the rectangular node unit of fault image, and m is the line number of rectangular node unit, and n is the row number of rectangular node unit;

Based on Differential Principle, the pixel of at least one of rectangular node unit is marked, is specially:

$grid=\underset{i}{\Σ}\underset{j}{\Σ}pixel(i,j);$

Wherein, pixel is the pixel of the rectangular node unit of fault image, and i is the line number of pixel, and j is the row number of pixel;

Utilize the color of binary number to pixel to mark, be specially:

The color of each pixel of organ image is represented with 24 bits respectively, and 24 bits is divided into low eight-digit binary number a, middle eight-digit binary number b and high eight-bit binary number c, be specially:

$\left\{\begin{array}{c}a=colour\\ b=colour/256\\ c=colour/65536\end{array}\right.;$

$pixel=\underset{a}{\Σ}\underset{b}{\Σ}\underset{c}{\Σ}colour(a,b,c);$

Wherein, colour is the color value of pixel;

Described fault image, rectangular node unit, the pixel of rectangular node unit, the color set of pixel are formed the digitized markers of described organ image, are specially:

$organ=\underset{k}{\Σ}tomography\left(\underset{m}{\Σ}\underset{n}{\Σ}grid\right(\underset{i}{\Σ}\underset{j}{\Σ}pixel\left(\underset{a}{\Σ}\underset{b}{\Σ}\underset{c}{\Σ}colour(a,b,c)\right)\left)\right).$