Presumption method

Abstract

This technology provides a method for accurately estimating the volume of a 3D cell aggregate from a 2D image. [Solution] An estimation method for estimating the volume of a cell aggregate, comprising: step S101 of acquiring a two-dimensional image of the cell aggregate; step S102 of extracting an object region corresponding to the cell aggregate from the two-dimensional image; step S103 of finding a reference point of the object region; step S103 of estimating a reference height which is the height of the reference point; step S105 of calculating a first distance between the reference point and the target point for each target point of the object region and a second distance between the reference point and the periphery of the object region; step S106 of calculating the height of the target point from the reference height, the first distance and the second distance; and step S107 of estimating the volume of the cell aggregate from the heights of all target points. As a result, even if the cell aggregate is deviated from a perfect sphere and the distance from the reference point to the periphery of the object region is not constant, the volume of the cell aggregate can be estimated with high accuracy.

Description

Technical Field

[0001] The present invention relates to a technique for estimating the volume of three-dimensionally cultured cell aggregates.

Background Art

[0002] Three-dimensional cell aggregates such as spheroids and organoids are known to mimic tissue characteristics better than conventional two-dimensional cell cultures. For this reason, as one of the applications, it may be used as a tissue model in screening. Screening is a method of measuring the amount of ATP activity in cells, obtaining a dose-response curve based on the change in the amount of activity according to the concentration of a drug or reagent, calculating IC50 (50% inhibitory concentration) and EC50 (50% effective concentration), and searching for drugs or reagents with a large effect based on these values. The ATP activity value is used as a method for quantifying the relative amount of living cells.

[0003] On the other hand, when screening is performed by a method other than the ATP activity value, there is a method of observing the increase or decrease of living cells in three-dimensional cell aggregates. As a method for evaluating three-dimensional cell aggregates, for example, it is described in Patent Document 1. In Patent Document 1, a three-dimensional image of a three-dimensional cell aggregate is acquired, and the cell aggregate is evaluated by analyzing the three-dimensional image. However, since a special device is required to acquire a three-dimensional image and it also takes a long time to shoot, it is preferable that it can be analyzed from a two-dimensional image such as a microscope image.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, since microscope images are two-dimensional images, while area values ​​can be calculated, it is difficult to accurately calculate the volume of a three-dimensional cell aggregate. For this reason, conventionally, the area value of a three-dimensional cell aggregate composed of living cells was used to quantify the amount of living cells from two-dimensional images such as microscope images.

[0006] Alternatively, the volume of a 3D cell cluster was calculated based on the average of the 3D cell clusters obtained from 2D images, assuming the 3D cell cluster was a perfect sphere. In that case, the radius r was calculated from the area A of the 3D cell cluster obtained from the 2D image, as shown in Equation 1, and then the volume V was calculated using that radius r, as shown in Equation 2.

number

number

[0007] However, three-dimensional cell aggregates often do not take on a perfect sphere shape, resulting in a discrepancy between their perceived volume and their actual volume.

[0008] This invention has been made in view of these circumstances, and aims to provide a technique for accurately estimating the volume of a three-dimensional cell aggregate from a two-dimensional image. [Means for solving the problem]

[0009] To solve the above problems, the first invention of the present application is an estimation method for estimating the volume of a three-dimensionally cultured cell aggregate, comprising: a) a step of acquiring a two-dimensional image of the cell aggregate; b) a step of extracting an object region corresponding to the cell aggregate from the two-dimensional image; c) a step of determining a reference point of the object region; d) a step of estimating a reference height which is the height of the reference point; e) a step of calculating a first distance which is the distance from the reference point to the calculation target point and a second distance which is the distance from the reference point to the periphery of the object region on the extension of the calculation target point, with each point of the object region as a calculation target point; f) a step of calculating the height of each calculation target point from the reference height calculated in step d), the first distance and the second distance calculated in step e); and g) a step of estimating the volume of the cell aggregate based on the heights of all the calculation target points calculated in step f).

[0010] The second invention of this application is the estimation method of the first invention, wherein in step c), the reference point is the centroid of the object region.

[0011] The third invention of this application is an estimation method for the first or second invention, wherein in step d), the reference height is calculated based on the average radius of the object region centered on the reference point.

[0012] The fourth invention of this application is a method for estimating any one of the first to third inventions, wherein in step f), the height of the point to be calculated is calculated by assuming that the vertical cross-section from the reference point to the periphery of the object region on the extension line of the point to be calculated is elliptical.

[0013] The fifth invention of this application is a method for estimating any one of the first to third inventions, wherein in step f), the height of the point to be calculated is calculated by assuming that the upper and lower halves of the vertical cross-section from the reference point to the periphery of the object region on the extension line of the point to be calculated are each extension lines of half a cycloid arch.

[0014] The sixth invention of this application is a method for estimating any one of the first to third inventions, wherein in step f), the height of the point to be calculated is calculated by assuming that the vertical cross-section from the reference point to the periphery of the object region on the extension line of the point to be calculated is parabolic. [Effects of the Invention]

[0015] According to the first to sixth inventions of this application, even if the shape of the cell aggregate deviates from a perfect sphere, the volume of the three-dimensional cell aggregate can be accurately estimated from a two-dimensional image. [Brief explanation of the drawing]

[0016] [Figure 1] This is a schematic diagram of the image processing device according to the first embodiment. [Figure 2] This is a functional block diagram of the image processing apparatus according to the first embodiment. [Figure 3] This is a flowchart illustrating the process for estimating the volume of a cell aggregate. [Figure 4] This is a two-dimensional image of a cell aggregate. [Figure 5] This is an image of an object region extracted from a 2D image of a cell aggregate. [Figure 6] This figure shows an example of an object region and calculation target points. [Figure 7] This figure shows an example of a hypothetical cross-section according to the first embodiment. [Figure 8] This diagram shows the extension of a cycloid curve by one arch. [Figure 9] This figure shows an example of the shape of the upper half of the hypothetical cross-section in the second variant. [Figure 10] This figure shows an example of the shape of the upper half of the hypothetical cross-section in the third variant. [Figure 11] This figure shows an example of the shape of the upper half of the hypothetical cross-section in the third variant. [Modes for carrying out the invention]

[0017] Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings.

[0018] <1. First Embodiment> <1-1. Configuration of the image processing device> Figure 1 is a schematic diagram of the image processing device 1 according to the first embodiment. Figure 2 is a functional block diagram of the image processing device 1. This image processing device 1 performs volume estimation processing of the cell aggregate 9 by processing a two-dimensional image of the cell aggregate 9.

[0019] As shown in Figure 1, this image processing device 1 comprises a computer body 11, a display unit 12, and an input unit 13. The computer body 11 includes a processor 111 such as a CPU, memory 112 such as RAM, and a storage unit 113 such as a hard disk drive. The display unit 12 displays images output from the computer body 11. The display unit 12 is, for example, a liquid crystal display such as a PC monitor. The input unit 13 can receive commands to the computer body 11. The input unit 13 is, for example, a keyboard and a mouse.

[0020] Memory 112 and storage unit 113 are connected to the processor 111 via bus wiring (not shown). Storage unit 113 stores a computer program Pg. The processor 111 loads the computer program Pg stored in storage unit 113 into memory 112 and executes the code contained in the computer program Pg sequentially. As a result, the image processing device 1 performs volume estimation processing on the two-dimensional image Di to be processed and estimates the volume of the three-dimensional cell aggregate 9 depicted in the two-dimensional image Di.

[0021] The computer program Pg is application software that causes the computer main unit 11 to perform various processes related to volume estimation. The computer program Pg is read from a storage medium M such as a CD or DVD and installed on the computer main unit 11. However, the computer program Pg may also be downloaded to the computer main unit 11 via a network N such as the Internet.

[0022] Figure 2 is a functional block diagram of the image processing device 1. The computer main unit 11 is connected to the imaging device 20, which captures a two-dimensional image Di of the cell aggregate 9, either directly via a wired connection or via a network N.

[0023] The computer main unit 11 has an image processing unit 30 as a functional unit realized by the processor 111 executing the computer program Pg. The image processing unit 30 includes a two-dimensional image acquisition unit 31, an object region extraction unit 32, a reference point determination unit 33, a reference height calculation unit 34, a distance calculation unit 35, a point height calculation unit 36, and an estimated volume calculation unit 37.

[0024] The 2D image acquisition unit 31 acquires a 2D image Di of the cell aggregate 9 input from the imaging device 20 to the image processing unit 30, and passes it on to the object region extraction unit 32.

[0025] The object region extraction unit 32 extracts the object region R corresponding to the cell cluster 9 from the 2D image Di. The object region extraction unit 32 then passes the object region R to the reference point determination unit 33, the reference height calculation unit 34, and the distance calculation unit 35.

[0026] The reference point determination unit 33 determines the reference point O in the object region R. The reference point determination unit 33 then passes the reference point O to the reference height calculation unit 34 and the distance calculation unit 35.

[0027] The reference height calculation unit 34 determines the reference height, which is the height of the cell aggregate 9 at the reference point O. The reference height H shown below is oThis represents half the height of the cell aggregate 9 at reference point O. The reference height calculation unit 34 calculates the reference height H o The data is then passed to the point height calculation unit 36.

[0028] The distance calculation unit 35 calculates a first distance OP and a second distance OE for each point in the object region R, treating each point as a target point P. The first distance OP is the distance from the reference point O to the target point P. The second distance OE is the distance from the reference point O to point E on the periphery of the object region R, which lies on the extension of the straight line drawn from the reference point O to the target point P. The distance calculation unit 35 passes the first distance OP and the second distance OE for all target points P to the point height calculation unit 36.

[0029] Note that a "point" that can be a calculation target point P within the object region R is, for example, each individual pixel. However, the calculation target point P is not limited to a single pixel; for example, multiple pixels such as 2x2 pixels can be treated as a single unit and used as the calculation target point P.

[0030] The point height calculation unit 36 ​​calculates the reference height H for each target point P in the object region R. o Then, the height H of each target point P is calculated from the first distance OP and the second distance OE. The point height calculation unit 36 ​​passes the height H for all target points P to the estimated volume calculation unit 37.

[0031] The estimated volume calculation unit 37 estimates the volume of the cell aggregate 9 based on the heights H of all calculation target points P in the object region R.

[0032] <1-2. Volume estimation process for cell aggregates> Figure 3 is a flowchart showing the flow of the volume estimation process for cell aggregates 9 in the image processing device 1.

[0033] In the cell aggregate volume estimation process, first, the 2D image acquisition unit 31 acquires a 2D image Di of the 3D cultured cell aggregate 9 from the imaging device 20 (Step S101: 2D image acquisition step). Specifically, for example, a 2D image is acquired by performing 2D imaging of the cell aggregate using bright-field observation with an optical microscope. Then, the 2D image acquisition unit 31 hands over the 2D image Di to the object region extraction unit 32.

[0034] Next, the object region extraction unit 32 extracts object regions R corresponding to the cell clusters 9 from the 2D image Di (Step S102: Object Region Extraction Step). The object region extraction unit 32 may, for example, define the region of the 2D image Di that is above / below a predetermined brightness threshold as the object region R. Alternatively, for example, the region above / below a predetermined brightness threshold and the region inside that region may be defined as the object region R. Or, the object region extraction unit 32 may extract the object region R using a machine learning model that has learned the regions of cell clusters using 2D images of multiple cell clusters.

[0035] Here, examples of object region extraction using a luminance threshold are shown in Figures 4 and 5. Figure 4 is an example of an image diagram of a 2D image Di. Figure 5 is an example of an image diagram of an object region. In the 2D image Di of the example in Figure 4, the object region extraction unit 32 extracts the region above a predetermined luminance threshold and the region inside that region as the object region R. As a result, the extracted object region R is extracted as a region as shown in Figure 5.

[0036] Next, the reference point determination unit 33 determines the reference point O of the object region R (step S103: reference point setting step). The reference point O of the object region R is, for example, the centroid of the object region R. Note that the reference point O of the object region R may be a point determined by other criteria, such as the center of the circumscribed circle of the object region R.

[0037] The reference height calculation unit 34 determines the reference height, which is the height of the cell aggregate 9 at the reference point O (step S104: reference height calculation step). The reference height is calculated based on, for example, the average radius of the object region R centered on the reference point O. Specifically, the reference height H, which is half of the height of the cell aggregate 9 at the reference point O, o is obtained by multiplying the average value H of the distance from the reference point O to the periphery of the object region R, as shown in Equation 3, α by a constant multiple. Note that the constant α is a positive number.

Equation

[0038] Here, the constant α varies depending on the type of the cell aggregate 9. Generally, the cell aggregate 9 obtained by three-dimensional culture often has a smaller size in the vertical direction (perpendicular direction) than in the horizontal direction. In such a case, the constant α is less than 1.

[0039] Thereafter, the distance calculation unit 35 calculates the first distance OP and the second distance OE for one calculation target point P in the object region R (step S105: distance calculation step).

[0040] FIG. 6 is a diagram showing an example of the object region R and the calculation target point P. As shown in FIG. 6, first, the object region R is arranged on the xy coordinate axes with the reference point O as the origin. Hereinafter, the calculation target point P arranged at (x, y) on this xy coordinate is referred to as P (x,y) . And the first distance OP for the calculation target point is referred to as OP (x,y) , and the second distance OE is referred to as OE (x,y) .

[0041] In step S105, the distance calculation unit 35 first calculates the first distance OP (x,y) from the coordinates (x, y) of the calculation target point P (x,y) . On the other hand, the distance calculation unit 35 is a point on the periphery of the object region R and is a point E (x,y) on the extension line of the line segment extended from the reference point O to the calculation target point P (x,y)Obtain point E. (x,y) From the coordinates, the second distance OE (x,y) Calculate.

[0042] Next, the point height calculation unit 36 ​​calculates the height of the point P (x,y) In contrast, the standard height H o And, the first distance OP (x,y) and second distance OE (x,y) Therefore, the calculation target point P (x,y) Calculate the height H (Step S106: Step to calculate the height of each point).

[0043] Figure 7 shows an example of a hypothetical cross-section of a cell aggregate 9 passing through the reference point O and the calculation target point P. In this embodiment, the reference point O and each calculation target point P (x,y) Assuming that the hypothetical cross-section of the cell aggregate 9 passing through is elliptical, the calculation target point P is calculated as follows: (x,y) Height H (x,y) Calculate.

[0044] First, as shown in Figure 6, the point height calculation unit 36 ​​uses a reference point O and a calculation target point P. (x,y) A coordinate axis s-axis with reference point O as the origin is placed on a straight line passing through [the specified point]. Then, as shown in Figure 7, with the s-axis as the horizontal axis and the vertical coordinate axis z-axis as the vertical axis, the line segment OE of the cell aggregate 9 is defined. (x,y) Assuming a longitudinal section in the region, the hypothetical section C (x,y) Create.

[0045] In this embodiment, the target point P is calculated from the reference point O. (x,y) Periphery E of object region R on the extension of the line (x,y) This is calculated assuming that the longitudinal section up to point C is elliptical. That is, assumed section C (x,y) The s-intercept is ±OE (x,y) Therefore, the z-intercept is ±H o It is an ellipse. In this case, the assumed cross section C (x,y) The peripheral shape is shown by the following equation 4. Transforming equation 4 yields equation 5.

number

number

[0046] From here, the point P to be calculated (x,y) Height H is half the height of cell aggregate 9 in the cell aggregate 9. (x,y) This is shown by equation 6.

number

[0047] In this way, the point height calculation unit 36 ​​calculates each target point P (x,y) For the given cross-section C (x,y) The calculation targets P are defined and (x,y) Height H (x,y) The following is calculated. Note that in Figure 6, different calculation target points P' are shown. (x,y) The corresponding peripheral point E' (x,y) An example of the coordinate axis s' is also shown for reference. In this way, the target point P (x,y) The direction of the s-axis differs depending on the position.

[0048] The image processing unit 30 then determines whether or not it has calculated the height H for all of the calculation target points P (step S107). If the image processing unit 30 determines in step S107 that it has not yet calculated the height H for all of the calculation target points P (No in step S107), it returns to step S105 and performs steps S105 and S106 for the calculation target points P for which the height H has not yet been calculated.

[0049] Furthermore, if the image processing unit 30 determines in step S107 that it has calculated the height H for all calculation target points P (Yes in step S107), it proceeds to step S108. Then, the estimated volume calculation unit 37 calculates the volume V of the cell aggregate 9 based on the height H of all calculation target points P in the object region R, as shown in Equation 7 (step S108).

number

[0050] In this way, the image processing unit 30 assumes a longitudinal section passing through the reference point O and the calculation target points P, and estimates the volume V. As a result, as shown in Figure 6, even if the shape of the cell aggregate 9 deviates from a perfect sphere and the distance from the reference point O to the periphery of the object region R is not constant, the volume V of the cell aggregate 9 can be estimated with high accuracy.

[0051] <2. Variant> Although one embodiment of the present invention has been described above, the present invention is not limited to the above-described embodiment.

[0052] In the above embodiment, the calculation target point P (x,y) Assumed cross section C (x,y) Assuming the shape is elliptical, each of the calculation target points P (x,y) Height H is half the height of cell aggregate 9 in the cell aggregate 9. (x,y The present invention seeks to achieve this, but is not limited to this. Variations of the shape of the assumed cross-section are shown in the following first, second, and third modified examples.

[0053] <2-1. First variation> Calculation target point P (x,y) Assumed cross section C (x,y) The shape may be a deformed ellipse. In that case, the assumed cross section C (x,y) The peripheral shape is given by the following equation 8, where n is an integer. Transforming equation 8 yields equation 9.

number

number

[0054] From here, the point P to be calculated (x,y) Height H is half the height of cell aggregate 9 in the cell aggregate 9. (x,y) This is shown by the number 10.

number

[0055] <2-2. Second variation> Calculation target point P (x,y) Assumed cross section C (x,y) The peripheral shapes of the upper and lower halves may each be cycloidal curves. Specifically, the calculation target point P (x,y) Height H (x,y) However, the target point P is calculated from the reference point O. (x,y) Assumed cross-section C to the periphery of the object region R on the extension of the line. (x,y) The calculation is performed by assuming that the upper and lower halves of the peripheral shape are each extensions of one arch of a cycloid curve.

[0056] Figure 8 shows the extension line of one arch of a cycloid curve in a two-dimensional uv coordinate system. Figure 9 shows the hypothetical cross section C in the second modified example. (x,y) This figure shows the shape of the upper half. The curve shown in Figure 8 is expressed parametrically by equations 11 and 12 below for the range 0 ≤ θ ≤ 2π.

number

number

[0057] Using this, a hypothetical cross-section C as shown in Figure 9 (x,y) The peripheral shape of the upper half is given by the following equations 13 and 14 using parametric representation for the range π ≤ θ ≤ 2π.

number

number

[0058] Number 13, 1st distance OP (x,y) However, s(θ)=OP (x,y) By substituting the value of θ into equation 14, we obtain the height H (x,y) Let = z(θ) be the point P to be calculated. (x,y) Height H(x,y) It is possible to calculate this.

[0059] <2-3. Third Variation> Calculation target point P (x,y) Assumed cross section C (x,y) The peripheral shapes of the upper and lower halves may each be parabolic. Specifically, the calculation target point P (x,y) Height H (x,y) However, the target point P is calculated from the reference point O. (x,y) Assumed cross-section C to the periphery of the object region R on the extension of the line. (x,y) The calculation is performed by assuming that the upper and lower halves of the peripheral shape are each part of the shape of a parabola.

[0060] Figures 10 and 11 show hypothetical cross-section C in the third modified example, respectively. (x,y) This figure shows an example of the shape. The two curves z1 and z2 shown in Figure 10 represent the case where the integer n is even in numbers 15 and 16 below. The two curves shown in Figure 11 represent the case where the integer n is odd in numbers 15 and 16 below.

number

number

[0061] In equation 15, s has a first distance OP. (x,y) By substituting this, the height H (x,y) The point to be calculated is P, where =z. (x,y) Height H (x,y) It is possible to calculate this.

[0062] <2-4. Other variations> Furthermore, the elements that appear in the above embodiments and modifications may be combined as appropriate, to the extent that no contradictions arise. [Explanation of Symbols]

[0063] 1: Image processing device 9:Cell clumps 30: Image Processing Unit 31: 2D Image Acquisition Unit 32: Object Area Extraction Unit 33: Reference point determination section 34: Reference height calculation unit 35: Distance calculation unit 36: Point height calculation unit 37:Estimated volume calculation part

Claims

1. A method for estimating the volume of a three-dimensionally cultured cell aggregate, a) A step of acquiring a two-dimensional image of the cell aggregate, b) A step of extracting an object region corresponding to the cell aggregate from the two-dimensional image, c) A step of determining the reference point of the object region, d) A step of estimating the reference height, which is the height of the reference point, e) A step of calculating a first distance, which is the distance from the reference point to the point to be calculated, and a second distance, which is the distance from the reference point to the periphery of the object region on the extension of the point to be calculated, using each point of the object region as a point to be calculated. f) A step of calculating the height of each of the above-mentioned points to be calculated from the reference height calculated in step d) and the first distance and second distance calculated in step e), g) A step of estimating the volume of the cell aggregate based on the heights of all the calculation target points calculated in step f), An estimation method comprising:

2. The estimation method according to claim 1, An estimation method in which, in step c), the reference point is the centroid of the object region.

3. The estimation method according to claim 1, In step d) above, the reference height is calculated based on the average radius of the object region centered on the reference point, in an estimation method.

4. The estimation method according to claim 1, In step f) above, the height of the point to be calculated is calculated by assuming that the vertical cross-section from the reference point to the periphery of the object region on the extension line of the point to be calculated is elliptical, in this estimation method.

5. The estimation method according to claim 1, In step f) above, the height of the point to be calculated is calculated by assuming that the upper and lower halves of the vertical cross-section from the reference point to the periphery of the object region on the extension line of the point to be calculated are each extension lines of half a cycloid arch.

6. The estimation method according to claim 1, In step f) above, the height of the point to be calculated is calculated by assuming that the vertical cross-section from the reference point to the periphery of the object region on the extension of the point to be calculated is parabolic.

Images