WO2019135268A1

WO2019135268A1 - Cell image processing device, cell image processing method, and cell image processing program

Info

Publication number: WO2019135268A1
Application number: PCT/JP2018/000024
Authority: WO
Inventors: 朗松下; 良介村田
Original assignee: オリンパス株式会社
Priority date: 2018-01-04
Filing date: 2018-01-04
Publication date: 2019-07-11
Also published as: JPWO2019135268A1; JP6891299B2

Abstract

A cell image processing device (1) that comprises: a vector calculation part (4) that receives as input a plurality of images that have been acquired by photographing cells at time intervals, divides each of any two of the inputted images into a plurality of units, and calculates a movement vector for the cells within each unit; a stable region finding part (5) that finds a stable region in which at least a peripheral edge is constituted by stable units that have movement vectors, as calculated by the vector calculation part (4), that have magnitudes that are smaller than a prescribed threshold value; and a cell counting part (6) that sets the stable region found by the stable region finding part (5) as a counting region and counts the cells in the counting region.

Description

Cell image processing apparatus, cell image processing method and cell image processing program

The present invention relates to a culture image processing apparatus, a cell image processing method, and a cell image processing program.

There is known a technique of analyzing an image of a cell and acquiring a parameter representing a change in the state of the cell in order to quantitatively observe a time-dependent change of cell proliferation (see, for example, Patent Document 1). Specifically, the timing of passage is determined based on the confluence ratio, which is the ratio of the area occupied by cells in the image.

In addition, in order to observe the relationship between cell aggregation and cell division, such as the formation of colonies, there is known a technique for acquiring statistical quantities of the positional relationship between a pixel of interest and peripheral pixels as a parameter representing cell aggregation (for example, , Patent Document 2).

WO 2016/098271 JP, 2009-229274, A

In order to manage the culture of cells precisely, it is necessary to accurately detect whether the cells are proliferating or not. However, there is a disadvantage that the proliferation of cells can not be detected with high accuracy only by acquiring parameters such as the confluence rate as in Patent Document 1 and the aggregation of cells around a target cell as in Patent Document 2.

The present invention has been made in view of the above-described circumstances, and a cell image processing device, a cell image processing method, and a cell image capable of accurately detecting cell proliferation and accurately managing cell culture. The purpose is to provide a processing program.

According to an aspect of the present invention, a plurality of images acquired by capturing cells at time intervals are input, and any two of the images among the input images are input into a plurality of units. Divided by the vector calculation unit for calculating the movement vector of the cells in each unit, and at least the periphery by the stable unit in which the magnitude of the movement vector calculated in the vector calculation unit is smaller than a predetermined threshold value. A stable region searching unit configured to search for a stable region, and a cell number counting unit configured to set the stable region searched by the stable region searching unit as a counting region and count the number of cells in the counting region. It is a cell image processing device.

According to this aspect, when a plurality of images acquired by capturing at intervals of time are input, any two images of the input images are respectively processed by the plurality of units in the vector calculation unit. The cell movement vector in each unit is calculated. The calculated movement vector is input to the stable area search unit, and a stable area whose periphery is constituted by a stable unit whose magnitude of the movement vector is smaller than a predetermined threshold value is searched. Then, the searched stable area is set as a counting area in the cell counting section, and the number of cells in the counting area is counted.

That is, since the stable region searched is constituted at least by the stable unit, the inflow of cells to the stable region and the outflow of cells from the stable region are limited to a certain level or less. Therefore, by setting the searched stable area as a counting area for counting the number of cells, it is possible to accurately count the number of cells while minimizing the influence of cell migration.

In the above aspect, the stable area search unit may classify in advance whether or not the plurality of units are the stable unit, and search for the stable area.
By doing this, it is possible to easily search for a stable area whose peripheral edge is constituted by the stable unit.

Further, in the above aspect, the non-stable unit, wherein the stable region search unit has a magnitude of the orthogonal component of any one of the movement vectors in any of the units calculated by the vector calculation unit equal to or larger than a predetermined threshold. And the motion vector in the other unit adjacent to the direction of the orthogonal component larger than the threshold value of the non-stable unit is calculated by the vector calculation unit, and the magnitudes of both of the orthogonal components are predetermined. The unit smaller than the threshold may be set as the stable unit.
By doing this, the stable region can be searched until the stable unit arranged around the searched astable unit is searched, and the unit which calculates the movement vector to set the counting region You need less.

Further, in the above aspect, the vector calculation unit may calculate the movement vector based on a correlation value between images in the same unit in two images.
By doing this, the movement vector can be easily calculated.

Furthermore, in the above aspect, the units may have the same shape.
By doing this, it is possible to easily divide the vector calculation unit into a plurality of units.

Moreover, in the above aspect, the stable region may be configured only by the stable unit.
By doing this, it is possible to count the number of cells with high accuracy using a stable area with less cell migration.

Further, in the above aspect, the stable area is configured by surrounding the periphery of one or more of the non-stable units whose magnitude of the orthogonal component of any of the movement vectors is equal to or larger than a predetermined threshold by the stable unit. May be
In this way, even if the stable region contains an unstable unit, the number of cells can be accurately counted using the stable region with less inflow and outflow of cells as a whole.

Further, in the above aspect, the stable area search unit has a function of outputting data of the stable area and the movement vector in a display device that displays the movement vector of the unit included in the stable area. It is also good.

In addition, another aspect of the present invention includes a processor into which a plurality of images acquired by capturing a cell at intervals of time are input, and the processor is any of the input images Each of the two images is divided into a plurality of units, the movement vector of the cells in each unit is calculated, and at least a stable unit in which the calculated magnitude of the movement vector is smaller than a predetermined threshold value. It is a cell image processing apparatus which searches for a stable area having a peripheral edge, sets the searched stable area as a counting area, and counts the number of cells in the counting area.

Further, according to another aspect of the present invention, a plurality of images acquired by capturing a cell at intervals of time are input, and any one of the two images among the input images is input. A vector calculation step of dividing into a plurality of units and calculating movement vectors of the cells in each of the units, and a stable unit in which the size of the movement vector calculated in the vector calculation step is smaller than a predetermined threshold value A stable area searching step of searching for a stable area having at least a peripheral edge, and a cell counting step of setting the stable area searched by the stable area searching step as a counting area and counting the number of cells in the counting area And a cell image processing method.

Further, according to another aspect of the present invention, a plurality of images acquired by capturing a cell at intervals of time are input, and any one of the two images among the input images is input. A vector calculation step of dividing into a plurality of units and calculating movement vectors of the cells in each of the units, and a stable unit in which the size of the movement vector calculated in the vector calculation step is smaller than a predetermined threshold value A stable area searching step of searching for a stable area having at least a peripheral edge, and a cell counting step of setting the stable area searched by the stable area searching step as a counting area and counting the number of cells in the counting area And a cell imaging program that causes a computer to execute the program.

According to the present invention, it is possible to accurately detect cell proliferation and to manage cell culture with high accuracy.

FIG. 1 is an overall configuration diagram showing a cell image processing apparatus according to an embodiment of the present invention. It is a figure which shows the image input into the cell image processing apparatus of FIG. 1, and the unit dividedly formed in the image. It is a figure which shows the example of the stable area | region set by the cell image processing apparatus of FIG. It is a flowchart explaining the cell image processing method by the cell image processing apparatus of FIG. It is a flowchart explaining the cell image processing method by the cell image processing apparatus concerning the modification of FIG. It is a figure explaining the process in which a stable area | region is searched in a one-dimensional direction by the cell image processing method of FIG. It is a figure explaining the process in which a stable area | region is searched in a two-dimensional direction by the cell image processing method of FIG. It is a figure explaining the process in which a stable area | region is searched in a two-dimensional direction by the cell image processing method of FIG. It is a figure explaining the process in which a stable area | region is searched in a two-dimensional direction by the cell image processing method of FIG. It is a figure explaining the process in which a stable area | region is searched in a two-dimensional direction by the cell image processing method of FIG. It is a figure which shows an example of the stable area | region where the cell of FIG. 5 was searched by the image processing method.

A cell image processing apparatus 1, a cell image processing method, and a cell image processing program according to an embodiment of the present invention will be described below with reference to the drawings.
In the cell image processing apparatus 1 according to the present embodiment, as shown in FIG. 1, a plurality of images (see FIG. 2) G obtained by imaging with a time interval opened by an imaging unit 50 such as a microscope are input. An apparatus for outputting the number of cells in the counting area to a display unit 60 such as a monitor.

The cell image processing apparatus 1 includes a memory 2 that sequentially stores the image G input from the imaging unit 50 in association with time information when the image G was captured, and a processing unit that processes the image G stored in the memory 2 It has 3 and.
The processing unit 3 divides the image G stored in the memory 2 into a plurality of units P, and calculates a movement vector of cells in each unit P, based on the calculated movement vector and the vector calculation unit 4 A stable area search unit 5 for searching for a stable area, and a cell number counting section 6 for setting the searched stable area as a counting area and counting the number of cells in the counting area are provided.

For example, the vector calculation unit 4 divides two images G taken out of the memory 2 into a plurality of units P, for example, a plurality of square units arranged in a square as shown in FIG. By calculating the correlation value of the two images G in the image by a known method, it is possible to calculate the movement vector of the subject by a method of evaluating the portion where the correlation value is high as an area where the same subject is photographed. It has become.

In this embodiment, movement vectors are calculated for all units P formed in each image G or units P within a designated predetermined range.
The stable area search unit 5 compares the movement vector calculated for each unit P with a predetermined threshold to classify the movement vector as a stable unit equal to or larger than the threshold and the stable unit where the movement vector is smaller than the threshold. It has become.

Then, the stable area search unit 5 searches for a stable area in which an edge is formed by the stable unit over the entire circumference. One or more stable regions may be searched in the image G.
Also, as the stable region, as shown in FIG. 3, two cases are considered: in the case of the region consisting only of the stable unit and in the case of the region surrounded by the stable unit all around the non-stable unit. Be In any case, the stable region is a region in which the number of inflow and outflow of cells is small.

The cell number counting unit 6 sets the searched stable area as a counting area, and counts the number of cells in the counting area by image processing.
When a plurality of search areas are searched by the stable area search unit 5, a plurality of counting areas are set in the same image G. This allows the growth characteristics of cells to be monitored at multiple locations.

By counting the number of cells in the set counting area in two images G separated by a time interval, the difference in the number of cells in the counting area of each image G is divided by the time interval between the two images G By simply doing this, it is possible to easily determine the cell proliferation rate, which is the change in the number of cells per unit time.

A cell image processing method using the cell image processing apparatus 1 according to the present embodiment configured as described above will be described below.
In the cell image processing method according to the present embodiment, as shown in FIG. 4, first, any two images G of the input images are divided into a plurality of units P (step S1).

Next, one of the units P divided in step S1 is selected (step S2). Then, the movement vector of the cells in the unit P selected in step S2 is calculated (vector calculation step S3).

After this, the magnitude of the movement vector calculated in the vector calculation step S3 is compared with a predetermined threshold (step S4), and as a result of comparison in step S4, if the movement vector is smaller than the predetermined threshold, It sets (step S5), and when a movement vector is more than a predetermined | prescribed threshold value, it sets as a non-stable step (step S6). Then, it is determined whether or not the classification is completed for all units P (step S7), and if not completed, the process from step S2 is repeated.

When classification is completed for all units P, a stable area at least the periphery of which is constituted by a stable unit is searched (stable area search step S8), and the stable area searched in stable area search step S8 is used as a counting area. The number of cells in the counting area is counted (cell counting step S9).

Since a plurality of units P divided in step S1 are classified into stable units and non-stable units in steps S4 to S6, a stable region search step S8 is to easily search for a stable region surrounded by the stable units. Can. Then, since the cell number is counted only for the stable region searched, the inflow and outflow of cells are small compared to the case where the cell number in the region where the unstable unit is arranged at the edge is counted. The number of cells can be counted for the area.
That is, according to the image processing method of the cell image processing apparatus 1 and cells relating to the present embodiment, cells for calculating the proliferation of cells by counting the number of cells in the stable region where the inflow and outflow of cells are small. There is an advantage that the number can be accurately determined with a small amount of noise.

In the present embodiment, the movement vector is calculated by obtaining the correlation value between the two images G for each unit P, but instead, the entire two images G or a set of a plurality of units P are calculated. A correlation value may be obtained between unit groups and the movement vector may be calculated for the entire image.

Further, in the present embodiment, each image G is divided into a plurality of square units P arranged in a square, but the present invention is not limited to this, and may be divided into units of other shapes. It may be divided into units of different sizes. For example, the central portion of the image G to be observed in detail may be a small unit, and the periphery of the image G may be divided into large units.

Further, in the present embodiment, for all units P generated in the image G, the stable unit or the non-stable unit is classified in advance, and then the stable region whose peripheral edge is constituted by the stable unit is searched. However, instead of this, the stable region may be searched by the method shown in FIG.

That is, first, as shown in FIG. 6A, the image G is divided into a plurality of units P (step S1), and the observation-required observation unit in the image G is selected (step S11). Next, the movement vector of the cells in the observation required unit is calculated based on the two images G separated by the time interval acquired for the selected observation required unit (vector calculation step S3). Then, it is determined whether or not either the calculated magnitude of the movement vector or the magnitude of two orthogonal components of the movement vector is equal to or greater than a predetermined threshold (step S12). In the example shown in FIG. 6, it is illustrated whether or not the magnitude of the movement vector itself in one direction is greater than or equal to the threshold value.

As shown in FIG. 6B, when it is determined in step S12 that the magnitude of the movement vector or the magnitude of any orthogonal component is equal to or greater than a predetermined threshold value, the observation-required unit is not It is determined that the unit is a stable unit (step S6), and if both the movement vector and the two orthogonal components are smaller than a predetermined threshold, it is determined that the unit is a stable unit (step S5).

As shown in FIG. 6C, when it is determined in step S6 that the unit is an unstable unit, units adjacent to both sides in the direction of the movement vector or orthogonal component determined to be equal to or greater than the threshold value It is newly established as an observation required unit (step S13).
If it is determined in step S13 that a unit requiring observation is newly formed and if it is determined that the unit is a stable unit in step S5, it is determined whether there is another unit requiring observation (step S14). As shown in FIGS. 6 (d) and 6 (e), the process from step S11 is repeated, and if not present, the number of cells in the counting area is counted (cell number counting step S9), Finish.

In the upper part of FIG. 6 (d), when all the observation units required in FIG. 6 (c) are non-stable units, in the middle part, the upper observation units newly added are the stability unit, and in the lower part. When the observation unit is a non-stabilizing unit, the lower part shows the case where the newly established observation unit is a stable unit and the stable region is set.
Further, FIG. 6E shows a state in which a new observation unit is newly formed at the adjacent position of the observation unit determined to be the non-stable unit in the upper stage and middle stage of FIG. 6D.

Also, FIGS. 7A to 7D illustrate an example of new establishment of the observation unit when one of the magnitude of the motion vector and the magnitude of the two orthogonal components of the motion vector has a magnitude equal to or greater than a predetermined threshold. . According to this, since a unit adjacent to a plurality of directions of a single required observation unit is newly established as a new required observation unit, the stable region is searched so as to expand in a two-dimensional manner. As a result, the stable region is searched as shown in FIG. 8, for example. Arrows in FIG. 8 indicate movement vectors in the stable unit or the non-stable unit.

According to such a method, it is not necessary to classify stable units and non-stable units in advance for all units P, and it is limited until the peripheral units are surrounded by the stable units based on the selected observation unit. It is sufficient if classification is performed for a number of units P. As a result, there is an advantage that the calculation area can be reduced and the counting area including the observation required unit can be set more easily.

Further, in the present embodiment, the case where the movement vector is calculated using two images G separated by a time interval to search for a stable region has been described. Actually, since the image G which changes every moment is sequentially input and stored in the memory 2, the movement vector is calculated sequentially using two different images G, and the searched stable region is updated each time It will be done. Therefore, the stable area searched as shown in FIG. 3 or 8 may be displayed on the display unit 60 so as to change from moment to moment.

Further, in the present embodiment, the processing unit 3 may include the vector calculating unit 4, the stable area searching unit 5, and the cell counting unit 6. The processing unit 3 may be configured by an electric circuit. You may comprise by the above processor.
When configured by a processor, the present invention may be configured by a cell image processing program that can be executed by a computer in vector calculation step S3, stable region search step S8 and cell count step S9.

DESCRIPTION OF SYMBOLS 1 cell image processing apparatus 4 vector calculation part 5 stable area | region search part 6 cell number counting part G image P unit S3 vector calculation step S8 stable area search step S9 cell number counting step

Claims

A plurality of images acquired by capturing cells at time intervals are input,
A vector calculation unit that divides each of any two of the input images into a plurality of units and calculates movement vectors of the cells in each of the units;
A stable area search section for searching for a stable area having at least a peripheral edge formed by a stable unit in which the magnitude of the movement vector calculated by the vector calculation section is smaller than a predetermined threshold value;
A cell image processing apparatus comprising: a cell number counting unit that sets the stable region searched by the stable region searching unit as a counting region and counts the number of cells in the counting region.
The cell image processing apparatus according to claim 1, wherein the stable area search unit classifies in advance whether or not the plurality of units are the stable unit, and searches for the stable area.
The stable region search unit searches for a non-stable unit in which the magnitude of any orthogonal component of the movement vector in any of the units calculated by the vector calculation unit is equal to or greater than a predetermined threshold. The cell image processing apparatus according to claim 1, wherein the stable region is searched until another unit adjacent in the direction of an orthogonal component larger than the threshold value of is the stable unit.
The cell image processing apparatus according to any one of claims 1 to 3, wherein the vector calculation unit calculates the movement vector based on a correlation value between images in the same unit in two images.
The cell imaging device according to any one of claims 1 to 4, wherein the units have the same shape.
The cell image processing apparatus according to any one of claims 1 to 5, wherein the stable area is constituted only by the stable unit.
4. The stable unit according to claim 3, wherein the stable unit is configured to surround the periphery of one or more of the non-stable units whose magnitudes of any orthogonal components of the movement vector are equal to or greater than a predetermined threshold. Cell imaging device.
The cell image processing according to claim 1, wherein the stable area search unit has a function of outputting data of the stable area and the movement vector on a display device that displays the movement vector of the unit included in the stable area. apparatus.
It has a processor to which a plurality of images acquired by capturing cells at time intervals are input,
The processor divides each of any two of the input images into a plurality of units, calculates a movement vector of the cells in each unit, and calculates the movement vector A stable unit having a size smaller than a predetermined threshold searches for a stable area having at least the periphery, sets the searched stable area as a counting area, and counts the number of cells in the counting area. Processing unit.
A plurality of images acquired by capturing cells at time intervals are input,
A vector calculation step of dividing each of any two of the input images into a plurality of units, and calculating a movement vector of the cells in each of the units;
A stable area search step of searching for a stable area having at least a periphery formed by a stable unit in which the magnitude of the movement vector calculated in the vector calculation step is smaller than a predetermined threshold value;
A cell number counting step of setting the stable region searched in the stable region searching step as a counting region and counting the number of cells in the counting region.
A plurality of images acquired by capturing cells at time intervals are input,
A vector calculation step of dividing each of any two of the input images into a plurality of units, and calculating a movement vector of the cells in each of the units;
A stable area search step of searching for a stable area having at least a periphery formed by a stable unit in which the magnitude of the movement vector calculated in the vector calculation step is smaller than a predetermined threshold value;
A cell image processing program that causes a computer to execute the cell counting step of setting the stable region searched in the stable region searching step as a counting region and counting the number of cells in the counting region.