Focus state evaluation method, stained image analysis method, and computer-readable program

Abstract

To precisely evaluate the focus state of a stained region of a stained image.SOLUTION: A focus state evaluation method of evaluating the focus state of a stained image by an immunostaining method comprises: a step (step S13) of extracting a stained region from a stained image of a sample by the immunostaining method; steps (steps S15 to S17) of setting an evaluation region including the stained region in the stained image, and performing mask processing to exclude an exclusion region as a region other than the evaluation region to acquire a processed image; and a step (step S18) of examining a focus state of the evaluation region in the processed image to acquire a focus evaluation value indicative of the focus state of the stained region. Consequently, the focus state of the stained region in the stained image can be precisely evaluated.SELECTED DRAWING: Figure 4A

Description

[Technical Field]

[0001] The present invention relates to a technique for evaluating the focus state of a stained image obtained by immunostaining. [Background technology]

[0002] In recent years, in medical and biological research, single-cell analysis of biological specimens such as tissue and cell specimens has been performed with the aim of elucidating disease mechanisms, biological mechanisms, drug action mechanisms, etc. For example, in immunostaining analysis, quantitative and qualitative analysis is performed by acquiring stained images of biological substances such as proteins and measuring the staining state through image analysis.

[0003] When performing the image analysis, the stained image must be in good focus to ensure the accuracy of the analysis. For this reason, in devices that acquire stained images, the stained image is focused using an autofocus mechanism. In addition, the quality of the focus state of the stained image is also determined.

[0004] Conventional autofocus methods achieve average focus for the entire stained image. Conventional focus state determinations determine an average focus state for the entire stained image. However, in analyzing stained images, it is usually important to analyze the stained region (hereinafter also referred to as the "stained region"). Therefore, even if the entire stained image is in focus on average, if the focus state in the stained region is poor, the analysis accuracy may decrease. Such poor focus in the stained region is likely to occur, for example, when the proportion of the stained region in the stained image is relatively low. In this case, the focus is adjusted to, for example, a region in the stained image where no cells exist or to the cell wall of an unstained cell.

[0005] Therefore, Patent Document 1 discloses a technology for determining the defocus amount of a sample by focusing on the sample within the field of view in an autofocus device using a phase-contrast optical system. Specifically, the microscope control device in Patent Document 1 divides a pair of phase-contrast images acquired by the autofocus device into multiple local regions and determines the variance of brightness values in each local region. Then, a set of local regions where the variance of brightness values is equal to or greater than a threshold is determined as a region in which the sample is imaged, and the phase difference between the pair of phase-contrast images for that region is calculated. Based on the phase difference, the amount of deviation between the position of the sample and the focal position of the microscope is calculated. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Japanese Patent Application Laid-Open No. 2012-42668 Summary of the Invention [Problem to be solved by the invention]

[0007] In the above-described immunostaining analysis, there is a need to determine whether a stained region in a pre-acquired stained image is in a focus state suitable for analysis. The technology of Patent Document 1, as described above, is a technology that calculates the amount of deviation between the position of the sample and the focal position of the microscope using a pair of different phase-contrast images having a phase difference before acquiring an image of the sample, and therefore cannot meet the above-described need.

[0008] The present invention has been made in consideration of the above-mentioned problems, and has as its object to accurately evaluate the focus state of a stained region in a stained image. [Means for solving the problem]

[0009] A first aspect of the present invention is a focus state evaluation method for evaluating the focus state of an image stained by immunostaining, comprising the steps of: a) extracting a stained region from an image of a specimen stained by immunostaining; b) setting an evaluation region including the stained region in the stained image and performing masking to exclude an exclusion region that is a region other than the evaluation region, thereby obtaining an image to be processed; and c) examining the focus state for the evaluation region in the processed image and obtaining a focus evaluation value that indicates the focus state of the stained region.

[0010] A second aspect of the present invention is the focus state evaluation method of the first aspect, wherein the step c) comprises: c1) extracting an edge of the dyed region from the processed image and determining a pixel value distribution indicating a distribution of pixel values associated with the edge; and c2) obtaining the focus evaluation value based on the pixel value distribution.

[0011] A third aspect of the present invention relates to the focus state evaluation method of the second aspect, wherein, in a state in which pixel values of the edge are higher than pixel values of areas other than the edge, the pixel value distribution includes a first peak portion which is a peak portion in an area near a maximum pixel value, and a second peak portion which is a peak portion in an area where pixel values are smaller than those of the first peak portion. The focus evaluation value is a ratio of the number of pixels included in the first peak portion to the number of pixels included in the second peak portion. The focus state of the dyed area improves as the focus evaluation value increases.

[0012] A fourth aspect of the present invention relates to the focus state evaluation method of the second aspect, wherein, when the pixel values of the edge are lower than the pixel values of the region other than the edge, the pixel value distribution includes a first peak portion that is a peak portion in a region near a minimum pixel value, and a second peak portion that is a peak portion in a region where the pixel values are higher than the first peak portion. The focus evaluation value is a ratio of the number of pixels included in the first peak portion to the number of pixels included in the second peak portion. The focus state of the dyed region improves as the focus evaluation value increases.

[0013] A fifth aspect of the present invention is the focus state evaluation method of any one of the first to fourth aspects, wherein the step a) includes a step of binarizing the stained image to obtain a binarized image, and the step b) includes a step of b1) performing a noise removal process on the binarized image, b2) setting the evaluation area in the binarized image and forming a mask image that masks the excluded area, and b3) obtaining the processed image by overlaying the mask image on the stained image.

[0014] A sixth aspect of the present invention is a focus state evaluation method according to any one of the first to fourth aspects (or any one of the first to fifth aspects), wherein the evaluation area includes the dyed area and a peripheral area of a predetermined width extending from the boundary of the dyed area.

[0015] A seventh aspect of the present invention is the focus state evaluation method of any one of Aspects 1 to 4 (or any one of Aspects 1 to 6), wherein in the step c), the focus evaluation value is acquired for each of a plurality of divided regions obtained by dividing the stained image. A processed stained image is generated in which, of the plurality of divided regions of the stained image, divided regions whose focus evaluation value is less than a predetermined threshold are not displayed.

[0016] Aspect 8 of the present invention is a focus state evaluation method according to any one of aspects 1 to 4 (which may be any one of aspects 1 to 7), in which focus evaluation values obtained by performing steps a), b), and c) on multiple stained images of the specimen having different focus states are compared, and one stained image having the best focus state is selected from the multiple stained images.

[0017] A ninth aspect of the present invention is the focus state evaluation method of any one of aspects 1 to 4 (or any one of aspects 1 to 8), further comprising, prior to step a), a step of performing color separation processing on the stained image. In step a), the stained region is extracted from the stained image after the color separation processing.

[0018] A tenth aspect of the present invention is a stained image analysis method for analyzing a plurality of stained images obtained by a multiple immunostaining method, comprising the steps of: d) obtaining a focus evaluation value of the stained region for each of the plurality of stained images obtained by the multiple immunostaining method using the focus state evaluation method of any one of aspects 1 to 4 (or any one of aspects 1 to 9); and e) analyzing the plurality of stained images taking into account the focus evaluation value for each stained image.

[0019] An eleventh aspect of the present invention is a computer-readable program that causes a computer to evaluate the focus state of a stained image obtained by immunostaining, and when the program is executed by a computer, the following steps are performed: a) extracting a stained area from an image of a specimen stained by immunostaining; b) performing a masking process on the stained image to exclude an exclusion area, which is an area other than an evaluation area that includes the stained area, to obtain a processed image; and c) examining the focus state for the evaluation area in the processed image to obtain a focus evaluation value that indicates the focus state of the stained area. [Effects of the Invention]

[0020] According to the present invention, the focus state of a stained region in a stained image can be evaluated with high accuracy. [Brief explanation of the drawings]

[0021] [Figure 1] 1 is a diagram showing a configuration of an image display device according to an embodiment; [Figure 2] FIG. 2 is a block diagram showing a functional configuration of the image display device. [Figure 3] FIG. 10 shows stained images. [Figure 4A] FIG. 10 is a diagram showing a flow of evaluation of the focus state of a stained image. [Figure 4B] FIG. 10 is a diagram showing a flow of evaluation of the focus state of a stained image. [Figure 5] FIG. 10 is a diagram showing a binarized image. [Figure 6] FIG. 10 is a diagram showing a mask image. [Figure 7] FIG. 2 is a diagram showing a processed image. [Figure 8] FIG. 10 is a diagram illustrating an example of a pixel value distribution. [Figure 9] FIG. 10 is a diagram illustrating an example of a pixel value distribution. [Figure 10] FIG. 10 is a diagram illustrating an example of a pixel value distribution. [Figure 11] FIG. 10 is a diagram illustrating an example of a pixel value distribution. [Figure 12] FIG. 2 is a diagram showing a processed image. [Figure 13] FIG. 10 shows a processed stained image. [Figure 14] FIG. 1 is a diagram showing the flow of analysis of stained images. DETAILED DESCRIPTION OF THE INVENTION

[0022] FIG. 1 is a diagram showing the configuration of a computer that functions as a focus state evaluation device 1 according to one embodiment of the present invention. The focus state evaluation device 1 is a device that evaluates the focus state of a stained image of a biological specimen obtained by immunostaining. The focus state is an index that indicates the degree to which an object in a stained image is in focus. When the object is in focus, the focus state is considered to be good. On the other hand, when the object is not in focus (i.e., out of focus), the focus state is considered to be poor. Furthermore, the greater the deviation from the focus of the object, the worse the focus state becomes.

[0023] 1 is a diagram showing the configuration of a computer that functions as a focus state evaluation device 1. The computer has the configuration of a typical computer system including a CPU 51, a ROM 52, a RAM 53, a fixed disk 54, a display 55, an input unit 56, a reading device 57, a communication unit 58, a GPU 59, and a bus 50. The CPU 51 performs various types of arithmetic processing. The GPU 59 performs various types of arithmetic processing related to image processing. The ROM 52 stores a basic program. The RAM 53 stores various types of information. The fixed disk 54 stores information. The display 55 is a display unit that displays various types of information such as images.

[0024] The input unit 56 includes a keyboard 56a and a mouse 56b that accept input from an operator. The reading device 57 reads information from a computer-readable recording medium 571, such as an optical disk, a magnetic disk, a magneto-optical disk, or a memory card. The display 55, the keyboard 56a, the mouse 56b, and the reading device 57 are connected to the bus 50 via an interface I / F. The communication unit 58 transmits and receives signals to and from devices external to the focus state evaluation device 1. The bus 50 is a signal circuit that connects the CPU 51, the GPU 59, the ROM 52, the RAM 53, the fixed disk 54, the display 55, the input unit 56, the reading device 57, and the communication unit 58.

[0025] In the focus state evaluation device 1, a program 572 is read in advance from a recording medium 571 via a reading device 57 and stored on a fixed disk 54. The program 572 may be stored on the fixed disk 54 via a network. The CPU 51 and the GPU 59 execute arithmetic processing using the RAM 53 and the fixed disk 54 in accordance with the computer-readable program 572. The CPU 51 and the GPU 59 function as a calculation unit in the focus state evaluation device 1. Other components that function as a calculation unit may be employed in addition to the CPU 51 and the GPU 59.

[0026] 2 is a block diagram showing the functional configuration of the focus state evaluation device 1 realized by the above-described computer executing arithmetic processing and the like in accordance with the program 572. This functional configuration includes a storage unit 501, a stained region extraction unit 502, an evaluation region setting unit 503, a mask image formation unit 504, a processed image acquisition unit 505, and a focus state evaluation unit 506. All or part of these functions may be realized by dedicated electrical circuits. Alternatively, these functions may be realized by multiple computers.

[0027] 2, storage unit 501 is mainly realized by RAM 53 and fixed disk 54. Furthermore, stained region extraction unit 502, evaluation region setting unit 503, mask image formation unit 504, processed image acquisition unit 505, and focus state evaluation unit 506 are realized by CPU 51, GPU 59, ROM 52, RAM 53, fixed disk 54, and their peripheral components.

[0028] The storage unit 501 stores in advance stained images of a biological specimen (hereinafter also simply referred to as a "specimen"). FIG. 3 is a diagram schematically showing an example of a stained image 91 of a specimen stored in the storage unit 501. The stained image 91 is an image obtained by capturing an image of a specimen stained by immunostaining. In FIG. 3, the stained image 91 is drawn at a lower resolution (i.e., with a larger pixel size) than the actual stained image 91.

[0029] The stained image 91 is, for example, a single-stained image captured after staining a specimen using a single staining method. The staining method is, for example, a staining method using a CD45 antibody that detects CD45 protein, which is one of the CD classifications. In this case, the stained image 91 shows cells detected by the CD45 antibody among many cells contained in the specimen, stained in a single color. Note that the specimen may be stained using various staining methods that use antibodies, or may be stained using various staining methods that do not use antibodies.

[0030] The stained image 91 shown in FIG. 3 includes two stained regions 92. The stained regions 92 are regions where cells stained by the staining method described above are gathered. In FIG. 3, the stained regions 92 are indicated by hatched lines. In the stained image 91, regions other than the stained regions 92 are regions where no cells originally exist, or regions where unstained cells that are not stained by the staining method exist.

[0031] 4A and 4B are diagrams showing the flow of evaluation of the focus state of a stained image 91 by the focus state evaluation device 1. When evaluating the focus state, first, a specimen stained by immunostaining is imaged in a device other than the focus state evaluation device 1 (for example, a microscope device, etc.) to obtain a stained image 91. The stained image 91 is sent to a storage unit 501 (see FIG. 2) of the focus state evaluation device 1 and stored in the storage unit 501 (FIG. 4A: step S11). Note that the focus state evaluation device 1 may be part of the above-mentioned microscope device, etc.

[0032] Next, stained image 91 is subjected to color separation processing by stained region extraction unit 502 (see FIG. 2 ), and a specific color tone to be analyzed in stained image 91 is extracted (step S12). In step S12, color separation processing is performed using a known method, for example, so that the pixel value (i.e., luminance value) of stained region 92 is increased. In this case, in stained image 91 after color separation processing, the pixel value of stained region 92 is higher than the pixel value of the area other than stained region 92. On the other hand, if color separation processing is performed so that the pixel value (i.e., luminance value) of stained region 92 is decreased, the pixel value of stained region 92 is lower than the pixel value of the area other than stained region 92 in stained image 91 after color separation processing. The color separation processing may be performed using commercially available image analysis software. In this embodiment, stained image 91 after color separation processing is an image expressed in 256 gradations, with the minimum and maximum pixel values being 0 and 255, respectively.

[0033] Next, evaluation area setting unit 503 binarizes stained image 91 after color separation processing to obtain a binarized image. As a result, stained area 92 is extracted from stained image 91 (step S13). Then, noise removal processing is performed on the binarized image (step S14). The noise removal processing may be performed by various known methods, such as expansion / contraction processing or a median filter. For example, if stained area 92 is represented in black on the binarized image, the size of each black area may be measured, and areas smaller than a predetermined size may be removed as noise (i.e., changed to white). On the other hand, if stained area 92 is represented in white on the binarized image, the size of each white area may be measured, and areas smaller than a predetermined size may be removed as noise (i.e., changed to black). The noise removal processing may be performed using commercially available image analysis software.

[0034] When the noise removal process is completed, evaluation regions 94 including each stained region 92 are set in a binarized image 93 of the stained image 91 shown in FIG. 5 (step S15). In the example shown in FIG. 5, two evaluation regions 94 each including two stained regions 92 are set on the binarized image 93. In FIG. 5, the outer edges of the evaluation regions 94 are indicated by two-dot chain lines. Each evaluation region 94 includes one entire stained region 92. Each evaluation region 94 includes the stained region 92 and a peripheral region 941 extending around the stained region 92. The peripheral region 941 is a region that extends from the boundary 921 of the stained region 92 around the entire periphery of the stained region 92 by a predetermined width (i.e., a predetermined number of pixels) toward a background region 95 outside the stained region 92. In other words, the peripheral region 941 is a portion of the background region 95, which is the region other than the stained region 92, that is located around the entire periphery of the stained region 92. In the example shown in FIG. 5, peripheral region 941 is a region that extends outward by two pixels from boundary 921 between stained region 92 and background region 95.

[0035] Note that when multiple stained regions 92 are arranged closely to each other on binarized image 93, the entirety of each of the multiple stained regions 92 may be included in one evaluation region 94. Furthermore, evaluation region 94 does not necessarily need to include peripheral region 941, and evaluation region 94 may be set so that the shape of evaluation region 94 is the same as the overall shape of stained region 92. Furthermore, the above-mentioned noise removal process (step S14) may be performed between step S15 and step S16, which will be described later.

[0036] Once the evaluation area 94 has been set, a mask image 97 that masks an exclusion area 96, which is an area of the binarized image 93 other than the evaluation area 94, is formed by the mask image forming unit 504 (see FIG. 2) (step S16), as shown in FIG. 6. In FIG. 6, the exclusion area 96 of the mask image 97 is marked with diagonal lines (the same applies to FIG. 7).

[0037] Next, processed image acquisition unit 505 (see FIG. 2) superimposes mask image 97 on stained image 91 shown in FIG. 3, thereby excluding exclusion region 96 in stained image 91 and acquiring processed image 98 consisting only of evaluation region 94, as shown in FIG. 7 (step S17). That is, in step S17, mask processing is performed on stained image 91 to exclude exclusion region 96. In this embodiment, processed image 98 is an image expressed in 256 gradations, with the minimum and maximum pixel values being 0 and 255, respectively.

[0038] In the masking process, for example, if a heavily stained portion in stained image 91 is represented in a color close to black (i.e., a color with a pixel value close to 0), exclusion region 96 is changed to white (i.e., a pixel value of 255). On the other hand, if a heavily stained portion in stained image 91 is represented in a color close to white (i.e., a color with a pixel value close to 255), exclusion region 96 is changed to black (i.e., a pixel value of 0) by the masking process.

[0039] Alternatively, in the masking process, when a pixel value distribution is obtained in obtaining a focus evaluation value in step S18 described later, a process of excluding the pixel values of the pixels in the exclusion area 96 so that they are not included in the pixel value distribution may be performed.

[0040] Next, focus state evaluation unit 506 (see FIG. 2) examines the focus state for evaluation region 94 of processed image 98, and obtains a focus evaluation value indicating the focus state of stained region 92 (step S18).

[0041] In step S18, first, edges located in each evaluation region 94 of the processed image 98 are extracted. The edges refer to regions in each evaluation region 94 where the color and / or brightness changes significantly. The edges include the boundary 921 of the dyed region 92. The edges may also exist inside the dyed region 92 (i.e., the region inside the boundary 921 of the dyed region 92). The edge extraction may be performed by applying a known edge extraction filter such as a Sobel filter or a Laplacian filter to the processed image 98. The edge extraction may also be performed by other known edge extraction methods. Alternatively, if an edge has been extracted when the evaluation region 94 was set (step S15), the edge may be used.

[0042] Next, a pixel value distribution indicating the distribution of pixel values related to the edge is obtained (FIG. 4B: step S181). The pixel value distribution indicates the magnitude of change in luminance value near the edge (i.e., the edge portion), and can be obtained by various methods. For example, the above-mentioned edge extraction filter is applied to the processed image 98, and the distribution of pixel values of all pixels in the generated edge image (i.e., an image of the edge degree in the processed image 98) is obtained as the pixel value distribution.

[0043] Alternatively, as shown in Fig. 7, first, in an edge pixel group 923 which is a collection of a plurality of edge pixels 922 that form an edge, the pixel value (i.e., luminance value) of each edge pixel 922 is acquired. Then, the distribution of the pixel values of the edge pixels 922 in the edge pixel group 923 may be obtained as the pixel value distribution. Note that in Fig. 7, only the edge pixels 922 out of the pixels that form the processed image 98 are illustrated as small rectangles.

[0044] FIG. 8 is a diagram showing an example of the pixel value distribution obtained in step S181. The graph shown in FIG. 8 is a histogram showing the distribution of pixel values for all pixels in the edge image. In the edge image, pixel values at edges are higher than pixel values in regions other than the edges. The horizontal axis of the histogram represents pixel values, and the vertical axis represents the number of pixels corresponding to the pixel values. The histogram includes a peak portion (hereinafter also referred to as a "first peak portion 924") in a region near the maximum pixel value, and another peak portion (hereinafter also referred to as a "second peak portion 925") in a region where the pixel value is smaller than that of the first peak portion 924. As described above, when the distribution of pixel values of edge pixels 922 in the edge pixel group 923 is obtained as the pixel value distribution, a histogram substantially similar to that shown in FIG. 8 is obtained.

[0045] The first peak portion 924 is a portion where the number of pixels (i.e., the number of pixels corresponding to each pixel value) forms a peak that is generally convex upward in a pixel value range of a predetermined width near the maximum value (i.e., 255) of the gradation that represents the processed image 98. For example, when the number of pixels in the pixel value range of the predetermined width is curve-fitted using the least squares method or the like, the first peak portion 924 is a portion where the fitted curve monotonically increases as the pixel value increases, reaches a maximum number of pixels, and then monotonically decreases as the pixel value increases. The first peak portion 924 mainly corresponds to pixels that are in good focus (i.e., in focus) near the edge.

[0046] The width of the pixel value range of the first peak portion 924 (i.e., the difference between the maximum pixel value and the minimum pixel value in the pixel value range) is set arbitrarily, for example, within a range of 10 to 30 gradations. The pixel value range is preferably set so that the pixel value at which the number of pixels in the first peak portion 924 is greatest is approximately in the center of the pixel value range. In this embodiment, the first peak portion 924 is a portion corresponding to a pixel value range of 230 to 250, and the width of the pixel value range of the first peak portion 924 is 20 gradations. The pixel value range corresponding to the first peak portion 924 is set, for example, between the maximum pixel value (i.e., 255) and 204, which is 80% of the maximum pixel value. Note that the pixel value range corresponding to the first peak portion 924 does not include the maximum pixel value (i.e., 255).

[0047] The second peak portion 925 is located in a region where the pixel value is smaller than the pixel value range (e.g., 230 to 250) corresponding to the first peak portion 924. The second peak portion 925, similar to the first peak portion 924, is a portion where the number of pixels in a pixel value range of a predetermined width forms a peak portion where the number of pixels is generally convex upward. For example, when the number of pixels in the pixel value range of the predetermined width is curve-fitted using the least squares method or the like, the second peak portion 925 is a portion where the fitted curve monotonically increases as the pixel value increases, reaches a maximum number of pixels, and then monotonically decreases as the pixel value increases. The second peak portion 925 mainly corresponds to pixels near the edge that are not in good focus (i.e., out of focus). The second peak portion 925 is usually lower than the first peak portion 924. In other words, the maximum number of pixels in the second peak portion 925 is usually smaller than the maximum number of pixels in the first peak portion 924.

[0048] The width of the pixel value range of second peak portion 925 (i.e., the difference between the maximum pixel value and the minimum pixel value in the pixel value range) is set arbitrarily, for example, within a range of 10 to 30 gradations. The pixel value range is preferably set so that the pixel value with the largest number of pixels in second peak portion 925 is approximately in the center of the pixel value range. In this embodiment, second peak portion 925 is a portion corresponding to a pixel value range of 110 to 130, and the width of the pixel value range of second peak portion 925 is 20 gradations. In this embodiment, the width of the pixel value range of second peak portion 925 is the same as the width of the pixel value range of first peak portion 924, but may be different. Note that the pixel value range corresponding to second peak portion 925 does not include the minimum pixel value (i.e., 0).

[0049] Once the pixel value distribution (see FIG. 8 ) is determined in step S181, a focus evaluation value indicating the focus state of dyed region 92 is acquired based on the pixel value distribution (step S182). Specifically, the number of pixels included in first peak portion 924 and the number of pixels included in second peak portion 925 are determined. Then, a value (hereinafter also referred to as the “edge pixel number ratio”) is calculated by dividing the number of pixels in first peak portion 924 by the number of pixels in second peak portion 925, and this edge pixel number ratio is used as the focus evaluation value. The edge pixel number ratio is the ratio of the number of pixels included in first peak portion 924 to the number of pixels included in second peak portion 925.

[0050] When the number of pixels in a good focus state is relatively large and the number of pixels in a bad focus state (i.e., defective) is relatively small near the edge of stained image 91, the edge pixel number ratio (i.e., focus evaluation value) in the pixel value distribution described above becomes large. On the other hand, when the number of pixels in a good focus state is relatively small and the number of pixels in a bad focus state is relatively large near the edge of stained image 91, the edge pixel number ratio (i.e., focus evaluation value) in the pixel value distribution described above becomes small. In other words, when the focus state of stained region 92 in stained image 91 is good, the edge pixel number ratio becomes large, and when the focus state is poor, the edge pixel number ratio becomes small.

[0051] The pixel value distribution shown in Fig. 8 described above corresponds to stained image 91, which has a relatively large number of pixels in a good focus state and is suitable for image analysis. On the other hand, the pixel value distribution shown in Fig. 9 corresponds to stained image 91, which has a relatively large number of pixels in a bad focus state and is not very suitable for image analysis. In the pixel value distribution shown in Fig. 9, the difference between the maximum number of pixels in first peak portion 924 and the maximum number of pixels in second peak portion 925 is smaller than in the pixel value distribution shown in Fig. 8.

[0052] The pixel value distribution acquired in step S181 is not necessarily limited to the histograms illustrated in FIGS. 8 and 9. For example, if the pixel values of the edges in the edge image generated by applying the edge extraction filter described above to the processed image 98 are lower than the pixel values in the non-edge region, the histograms illustrated in FIGS. 10 and 11 are obtained. FIGS. 10 and 11 correspond to the histograms illustrated in FIGS. 8 and 9, respectively. The horizontal axis of the histograms illustrated in FIGS. 10 and 11 indicates the pixel value, and the vertical axis indicates the number of pixels corresponding to the pixel value. The histograms illustrated in FIGS. 10 and 11 each include a first peak 924, which is a peak in a region near the minimum pixel value, and a second peak 925, which is another peak in a region with a pixel value greater than that of the first peak 924. As described above, when the distribution of pixel values of edge pixels 922 in the edge pixel group 923 is calculated as the pixel value distribution, histograms substantially similar to those illustrated in FIGS. 10 and 11 are obtained.

[0053] 10 and 11, the first peak portion 924 is a portion where the number of pixels (i.e., the number of pixels corresponding to each pixel value) forms a peak that is generally convex upward in a pixel value range of a predetermined width near the minimum value (i.e., 0) of the gradation that represents the processed image 98. For example, when the number of pixels in the pixel value range of the predetermined width is curve-fitted using the least squares method or the like, the first peak portion 924 is a portion where the fitted curve monotonically increases as the pixel value increases, reaches a maximum number of pixels, and then monotonically decreases as the pixel value increases. The first peak portion 924 mainly corresponds to pixels that are in good focus (i.e., in focus) near the edge.

[0054] The width of the pixel value range of the first peak portion 924 (i.e., the difference between the maximum pixel value and the minimum pixel value in the pixel value range) is set arbitrarily, for example, within a range of 10 to 30 gradations. The pixel value range is preferably set so that the pixel value with the largest number of pixels in the first peak portion 924 is approximately in the center of the pixel value range. In this embodiment, the first peak portion 924 is a portion corresponding to a pixel value range of 5 to 25, and the width of the pixel value range of the first peak portion 924 is 20 gradations. The pixel value range corresponding to the first peak portion 924 is set, for example, between the minimum pixel value (i.e., 0) and 51, which is 20% of the maximum pixel value (i.e., 255). Note that the pixel value range corresponding to the first peak portion 924 does not include the minimum pixel value (i.e., 0).

[0055] The second peak portion 925 is located in a region where the pixel value is greater than the pixel value range (e.g., 5 to 25) corresponding to the first peak portion 924. The second peak portion 925, similar to the first peak portion 924, is a portion where the number of pixels in a pixel value range of a predetermined width forms a peak portion where the number of pixels is generally convex upward. For example, when the number of pixels in the pixel value range of the predetermined width is curve-fitted using the least squares method or the like, the second peak portion 925 is a portion where the fitted curve monotonically increases as the pixel value increases, reaches a maximum number of pixels, and then monotonically decreases as the pixel value increases. The second peak portion 925 mainly corresponds to pixels near the edge that are not well-focused (i.e., out of focus). The second peak portion 925 is usually lower than the first peak portion 924. In other words, the maximum number of pixels in the second peak portion 925 is usually smaller than the maximum number of pixels in the first peak portion 924.

[0056] The width of the pixel value range of the second peak portion 925 (i.e., the difference between the maximum pixel value and the minimum pixel value in the pixel value range) is set arbitrarily, for example, within a range of 10 to 30 gradations. The pixel value range is preferably set so that the pixel value with the largest number of pixels in the second peak portion 925 is approximately in the center of the pixel value range. In this embodiment, the second peak portion 925 is a portion corresponding to a pixel value range of 125 to 145, and the width of the pixel value range of the second peak portion 925 is 20 gradations. In this embodiment, the width of the pixel value range of the second peak portion 925 is the same as the width of the pixel value range of the first peak portion 924, but may be different. Note that the pixel value range corresponding to the second peak portion 925 does not include the maximum pixel value (i.e., 255).

[0057] Once the pixel value distribution (see FIGS. 10 and 11) has been determined in step S181, a focus evaluation value indicating the focus state of dyed region 92 is acquired based on the pixel value distribution, as described above (step S182). Specifically, the number of pixels included in first peak portion 924 and the number of pixels included in second peak portion 925 are determined. Then, a value (i.e., edge pixel number ratio) is calculated by dividing the number of pixels in first peak portion 924 by the number of pixels in second peak portion 925, and this edge pixel number ratio is used as the focus evaluation value. The edge pixel number ratio is the ratio of the number of pixels included in first peak portion 924 to the number of pixels included in second peak portion 925.

[0058] When the number of pixels in a good focus state is relatively large and the number of pixels in a bad focus state (i.e., defective) is relatively small near the edge of stained image 91, the edge pixel number ratio (i.e., focus evaluation value) in the pixel value distribution described above becomes large, as illustrated in Fig. 10. On the other hand, when the number of pixels in a good focus state is relatively small and the number of pixels in a bad focus state is relatively large near the edge of stained image 91, the edge pixel number ratio (i.e., focus evaluation value) in the pixel value distribution described above becomes small, as illustrated in Fig. 11. In the pixel value distribution shown in Fig. 11, the difference between the maximum number of pixels in first peak portion 924 and the maximum number of pixels in second peak portion 925 is smaller than in the pixel value distribution shown in Fig. 10.

[0059] In this way, when the focus state of stained region 92 in stained image 91 is good, the edge pixel number ratio is large as shown in Fig. 10, and stained image 91 is suitable for image analysis. On the other hand, when the focus state of stained image 91 is poor, the edge pixel number ratio is small as shown in Fig. 11, and stained image 91 is not very suitable for image analysis.

[0060] When step S18 (i.e., steps S181 and S182) is completed, focus state evaluation unit 506 (see FIG. 2) compares the focus evaluation value (i.e., edge pixel number ratio) acquired in step S18 with a predetermined threshold value (hereinafter also referred to as the "evaluation threshold value") pre-stored in storage unit 501. If the focus evaluation value is equal to or greater than the evaluation threshold value, the focus state of stained region 92 in stained image 91 is determined to be good (FIG. 4A: steps S19 and S20). On the other hand, if the focus evaluation value is less than the evaluation threshold value, the focus state of stained region 92 in stained image 91 is determined to be poor (steps S19 and S21).

[0061] In the above-mentioned step S18, it is not necessary to acquire the edge pixel number ratio as the focus evaluation value, and other parameters may be acquired as the focus evaluation value based on the pixel value distribution. Also, the focus evaluation value may be acquired by other known methods without using the pixel value distribution.

[0062] Other parameters obtained based on the pixel value distribution described above are obtained, for example, as follows. First, a pixel value range (e.g., a range of pixel values from 120 to 190) is selected from the pixel value distribution shown in FIG. 8 that is greater than the pixel value indicating the maximum number of pixels in second peak portion 925 and in which the number of pixels generally decreases as the pixel value increases. Then, in that pixel value range, the slope of the number of pixels versus pixel value is calculated by linear approximation, and that slope is obtained as the focus evaluation value. In this case, the focus evaluation value is a negative value. If the absolute value of the focus evaluation value is relatively large, the focus state in stained region 92 of stained image 91 is good, and if the absolute value of the focus evaluation value is relatively small, the focus state is poor. This is because a poor focus state tends to result in a small number of pixels in the region between first peak portion 924 and second peak portion 925.

[0063] As described above, the focus state evaluation method for evaluating the focus state of a stained image 91 using an immunostaining method includes the steps of extracting a stained region 92 from the stained image 91 of a specimen using an immunostaining method (step S13), setting an evaluation region 94 including the stained region 92 in the stained image 91 and performing masking to exclude an exclusion region 96, which is the region other than the evaluation region 94, to obtain a processed image 98 (steps S15 to S17), and examining the focus state for the evaluation region 94 in the processed image 98 to obtain a focus evaluation value indicating the focus state of the stained region 92 (step S18).

[0064] In stained image 91, cell walls of many unstained cells (i.e., unstained cells) are generally present in exclusion region 96. Because the cell walls have high contrast, if the cell walls are included in the focus state examination target as in the past, stained image 91 in which the cell walls are in focus is likely to be determined to have a good focus state. However, in the analysis of stained image 91, stained region 92 is the analysis target. Therefore, it is not appropriate to determine that stained image 91 in which the cell walls of unstained cells are in focus has a good focus state, regardless of whether stained region 92 is in focus or not.

[0065] In the focus state evaluation method described above, a mask process is performed on an exclusion region 96 of the stained image 91 (i.e., a region other than the evaluation region 94), and the exclusion region 96 is essentially excluded from the focus state considerations. This eliminates the influence of cell walls of unstained cells present in the exclusion region 96 (i.e., the influence of regions unnecessary for the analysis of the stained image 91), allowing the focus state of the stained region 92 to be evaluated with high accuracy. Furthermore, as described above, when the distribution of pixel values of the edge pixels 922 is obtained as the pixel value distribution, the number of pixels to be analyzed in the stained image 91 is reduced, and therefore the amount of processing (i.e., the amount of calculation) required for evaluating the focus state can also be reduced.

[0066] The step of acquiring the focus evaluation value (step S18) described above preferably includes the steps of: extracting edges of stained region 92 from processed image 98 and determining a pixel value distribution indicating the distribution of pixel values associated with the edges (step S181); and acquiring a focus evaluation value based on the pixel value distribution (step S182). In this way, by focusing on edges with relatively high contrast in stained region 92 and using the pixel value distribution of the edges, the focus evaluation value can be suitably acquired. Furthermore, by using the pixel value distribution of the edges rather than the pixel value distribution of pixels throughout stained region 92, the amount of processing (i.e., the amount of calculation) required to acquire the focus evaluation value can be reduced.

[0067] 8 and 9, in a state in which the pixel values of the edge are higher than the pixel values of the region other than the edge, the pixel value distribution includes a first peak portion 924, which is a peak portion in a region near the maximum pixel value, and a second peak portion 925, which is a peak portion in a region where the pixel value is smaller than that of the first peak portion 924. The focus evaluation value is preferably the ratio of the number of pixels included in the first peak portion 924 to the number of pixels included in the second peak portion 925 (i.e., the edge pixel number ratio). The focus state of the dyed region 92 improves as the focus evaluation value increases. This makes it possible to acquire a focus evaluation value that can preferably indicate the focus state of the dyed region 92.

[0068] 10 and 11 , in a state in which the pixel values of the edge are lower than the pixel values of the region other than the edge, the pixel value distribution includes a first peak portion 924, which is a peak portion in a region near the minimum pixel value, and a second peak portion 925, which is a peak portion in a region where the pixel values are higher than those of the first peak portion 924. The focus evaluation value is preferably the ratio of the number of pixels included in first peak portion 924 to the number of pixels included in second peak portion 925 (i.e., the edge pixel number ratio). The focus state of dyed region 92 improves as the focus evaluation value increases. As described above, this makes it possible to acquire a focus evaluation value that can preferably indicate the focus state of dyed region 92.

[0069] As described above, the step of extracting stained region 92 (step S13) preferably includes a step of binarizing stained image 91 to obtain binarized image 93. Furthermore, the step of obtaining raw image 98 preferably includes a step of performing a noise reduction process on binarized image 93 (step S14), a step of setting evaluation region 94 in binarized image 93 and forming mask image 97 that masks exclusion region 96 (steps S15 to S16), and a step of obtaining raw image 98 by overlaying mask image 97 on stained image 91 (step S17). This allows the acquisition of raw image 98 from which noise has been reduced. As a result, the focus state of stained region 92 can be evaluated with even greater accuracy. Extraction of stained region 92 in step S13 may be performed by various known methods other than binarizing stained image 91.

[0070] As described above, it is preferable that evaluation region 94 includes stained region 92 and peripheral region 941 of a predetermined width extending from boundary 921 of stained region 92 to the periphery. This increases the contrast of boundary 921 of stained region 92, allowing for more accurate evaluation of the focus state of stained region 92. Furthermore, it is possible to prevent or suppress weakly stained portions on the periphery of stained cells from being included in exclusion region 96 and being excluded from consideration when evaluating the focus state. As a result, it is possible to more accurately evaluate the focus state of stained region 92.

[0071] As described above, it is preferable that the focus state evaluation method further includes a step (step S12) of performing color separation processing on stained image 91 before step S13. In this case, in step S13, stained region 92 is extracted from stained image 91 after the color separation processing. As a result, stained image 91 from which a specific color tone to be analyzed has been extracted is used in step S13, and therefore stained region 92 can be suitably extracted from stained image 91.

[0072] As described above, program 572 is a computer-readable program that causes a computer to evaluate the focus state of a stained image obtained by immunostaining. Execution of program 572 by a computer performs the following steps: extracting a stained region 92 from a stained image 91 of a specimen obtained by immunostaining (step S13); performing masking processing on stained image 91 to exclude an exclusion region 96, which is a region other than an evaluation region 94 that includes the stained region 92, to obtain a processed image 98 (steps S15 to S17); and examining the focus state for the evaluation region 94 in processed image 98 to obtain a focus evaluation value that indicates the focus state of stained region 92 (step S18). As described above, this allows the focus state of stained region 92 to be evaluated with high accuracy.

[0073] In the above example, in step S18, one focus evaluation value is obtained for one stained image 91, but this is not limiting. For example, in step S18, a focus evaluation value may be obtained for each of a plurality of divided regions obtained by dividing one stained image 91.

[0074] Specifically, first, steps S12 to S17 are performed on one stained image 91 in the same manner as described above to obtain one raw image 98. Next, as shown in FIG. 12, the raw image 98 is divided into a plurality of divided regions 991. In the example shown in FIG. 12, the raw image 98 is divided into 12 divided regions 991 arranged in a matrix in the vertical and horizontal directions in the figure. Each divided region 991 has a square shape, for example. Each divided region 991 is larger than each pixel constituting the raw image 98 or the stained image 91, for example. In other words, each divided region 991 includes a plurality of pixels. Furthermore, each divided region 991 is larger than each cell included in the stained image 91, for example. Note that each divided region 991 may have various shapes other than a square (for example, a rectangle).

[0075] Next, the above-described step S18 is performed for each of the plurality of divided regions 991, thereby obtaining a focus evaluation value for each of the plurality of divided regions 991. In the focus state evaluation device 1, as shown in Fig. 13, of the plurality of divided regions 991 in the stained image 91, the divided regions 991 whose focus evaluation values are less than the above-described evaluation threshold are hidden, and a processed stained image 99 is generated in which only the divided regions 991 whose focus evaluation values are equal to or greater than the evaluation threshold are displayed.

[0076] The processed stained image 99 is displayed, for example, on the display 55 (see FIG. 1). The operator of the focus state evaluation device 1 can observe only the displayed divided area 991 of the processed stained image 99 displayed on the display 55. Therefore, the operator can analyze only the divided area 991 of the stained image 91 that has a good focus state, and can efficiently analyze the stained image 91.

[0077] In the above example, steps S12 to S17 are performed collectively for one stained image 91, and step S18 is performed individually for each divided region 991, but this is not limiting. For example, between step S11 and step S12, stained image 91 may be divided into a plurality of divided regions 991, and steps S12 to S18 may be performed for each divided region 991, to obtain a focus evaluation value for each divided region 991.

[0078] As described above, in the focus state evaluation method, it is also preferable that in step S18, a focus evaluation value is obtained for each of a plurality of divided regions 991 obtained by dividing stained image 91. Then, it is preferable to generate a processed stained image 99 in which, of the plurality of divided regions 991 of stained image 91, divided regions 991 having focus evaluation values less than a predetermined threshold value (i.e., evaluation threshold value) are not displayed. This makes it possible to observe only regions in stained image 91 that are in a good focus state.

[0079] The focus state evaluation device 1 described above can also be used to analyze a plurality of stained images 91 acquired by a multiple immunostaining method. Fig. 14 shows the flow of this analysis. In this analysis, for example, a plurality of stained images 91 acquired by a sequential multiple immunostaining method are prepared (step S31). The sequential multiple immunostaining method is a multiple immunostaining method in which immunostaining, observation (e.g., imaging), and color removal for a single specimen are repeated multiple times while changing the antibodies, etc. used for immunostaining.

[0080] Next, the focus state evaluation device 1 acquires a focus evaluation value of the stained region 92 for each of the plurality of stained images 91 (step S32). The method for acquiring the focus evaluation value by the focus state evaluation device 1 is similar to the above-described steps S12 to S18. Thereafter, the plurality of stained images 91 are analyzed in consideration of the focus evaluation value in each stained image 91 (step S33).

[0081] Specifically, for example, the above-mentioned multiple divided regions 991 are set in each stained image 91, and a focus evaluation value is obtained for each of the multiple divided regions 991. Then, for each divided region 991, the focus evaluation values of the multiple stained images 91 are compared with an evaluation threshold, and only divided regions 991 for which the focus evaluation values of all stained images 91 are equal to or greater than the evaluation threshold are extracted as divided regions 991 appropriate for analysis. For example, the above-mentioned multiple stained images 91 obtained by the sequential multiple immunostaining method are displayed on the display 55 of the focus state evaluation device 1, with only the divided regions 991 appropriate for analysis being displayed and the other divided regions 991 not being displayed.

[0082] As described above, the stained image analysis method for analyzing a plurality of stained images 91 acquired by the multiple immunostaining method includes a step (step S32) of acquiring a focus evaluation value for each of the plurality of stained images 91 acquired by the multiple immunostaining method using the focus state evaluation method described above, and a step (step S33) of analyzing the plurality of stained images 91 in consideration of the focus evaluation value for each stained image 91. This allows the plurality of stained images 91 acquired by the multiple immunostaining method to be suitably analyzed.

[0083] The focus state evaluation device 1 described above may be incorporated into a microscope device that acquires a stained image 91 of a specimen, and may be used for autofocusing the microscope device. For example, when the stained image 91 of the specimen is captured in the microscope device, first, a focus adjustment mechanism of the microscope device or the like is driven to provisionally acquire a plurality of stained images of the specimen with different focus states. Hereinafter, the provisionally acquired plurality of stained images will also be referred to as "provisionally acquired images."

[0084] Subsequently, the focus state evaluation device 1 performs the above-described steps S12 to S18 for each of the plurality of provisionally acquired images to acquire a focus evaluation value for each of the provisionally acquired images. Next, the focus evaluation values of the plurality of provisionally acquired images are compared, and one provisionally acquired image with the best focus state is selected from the plurality of provisionally acquired images.

[0085] If the focus evaluation value of the selected temporarily acquired image is equal to or greater than a predetermined threshold (for example, the above-mentioned evaluation threshold), the microscope apparatus acquires the selected temporarily acquired image as a stained image 91 of the specimen. On the other hand, if the focus evaluation value of the selected temporarily acquired image is less than the predetermined threshold, focus adjustment is performed based on the focus evaluation values of the multiple temporarily acquired images to improve the focus state, and then multiple new temporarily acquired images are acquired. Then, steps S12 to S18 are performed on the new multiple temporarily acquired images in the same manner as above, the focus evaluation values of the acquired temporarily acquired images are compared, and the single temporarily acquired image with the best focus state is selected.

[0086] As described above, in the focus state evaluation method, steps S12 to S18 are performed on a plurality of stained images (i.e., provisionally acquired images) of a specimen having different focus states, and the focus evaluation values obtained are compared, and one stained image having the best focus state is selected from the plurality of stained images. As described above, steps S12 to S18 enable the focus state of stained region 92 to be evaluated with high accuracy. Therefore, according to the focus state evaluation method, one stained image 91 having a good focus state of stained region 92 can be suitably selected.

[0087] The microscope apparatus repeats the above-described acquisition of the plurality of provisionally acquired images, comparison of the focus evaluation values acquired in steps S12 to S18, and selection of one provisionally acquired image until a stained image 91 having a focus evaluation value equal to or greater than a predetermined threshold is acquired. As a result, the microscope apparatus incorporating the focus state evaluation apparatus 1 can capture a stained image 91 in which the focus state of the stained region 92 is excellent.

[0088] The focus state evaluation method, stained image analysis method, and program 572 described above can be modified in various ways.

[0089] For example, step S14 (noise removal processing) does not necessarily have to be performed and may be omitted.

[0090] In step S17, the masking process for excluding the exclusion region 96 of the stained image 91 does not necessarily have to be performed by superimposing a mask image 97 on the stained image 91, and may be performed by various known methods. In this case, the formation of the mask image 97 may be omitted.

[0091] The stained image 91 does not necessarily have to be a monochromatic image, but may be a multicolor image in which, for example, multiple types of antibodies are used on a single specimen. In this case, the color separation process in step S12 may extract a color tone corresponding to a single selected antibody.

[0092] The configurations in the above-described embodiment and each modification may be combined as appropriate as long as they are not mutually contradictory. [Explanation of symbols]

[0093] 91 stained images 92 Staining area 93 Binarized Images 94 Evaluation Area 96 Exclusion area 97 Mask Images 98 processed images 99 Processed dyed images 921 Boundary (of the dyed area) 924 First Peak 925 Second Peak 941 Surrounding Area 991 Split area S11~S21, S31~S33, S181~S182 steps

Claims

1. A focus state evaluation method for evaluating a focus state of a stained image by an immunostaining method, comprising: a) extracting a stained area from an image of a specimen stained by an immunostaining method; b) setting an evaluation area including the stained area in the stained image, and performing mask processing to exclude an exclusion area that is an area other than the evaluation area, to obtain a processed image; c) examining a focus state for the evaluation region in the processed image and obtaining a focus evaluation value that indicates the focus state of the stained region; A focus state evaluation method comprising:

2. 2. The focus state evaluation method according to claim 1, The step c) c1) extracting an edge of the stained region from the processed image and determining a pixel value distribution indicating a distribution of pixel values associated with the edge; c2) obtaining the focus evaluation value based on the pixel value distribution; A focus state evaluation method comprising:

3. 3. The focus state evaluation method according to claim 2, In a state where the pixel value of the edge is higher than the pixel value of the area other than the edge, The pixel value distribution is a first peak portion which is a peak portion in an area near the maximum value of the pixel value; a second peak portion which is a peak portion in an area where the pixel value is smaller than that of the first peak portion; Equipped with the focus evaluation value is a ratio of the number of pixels included in the first peak portion to the number of pixels included in the second peak portion, A focus state evaluation method in which the focus state of the dyed region improves as the focus evaluation value increases.

4. 3. The focus state evaluation method according to claim 2, In a state where the pixel value of the edge is lower than the pixel value of the area other than the edge, The pixel value distribution is a first peak portion which is a peak portion in an area near the minimum value of the pixel value; a second peak portion which is a peak portion in an area where the pixel value is larger than that of the first peak portion; Equipped with the focus evaluation value is a ratio of the number of pixels included in the first peak portion to the number of pixels included in the second peak portion, A focus state evaluation method in which the focus state of the dyed region improves as the focus evaluation value increases.

5. 5. The focus state evaluation method according to claim 1, The step a) includes a step of binarizing the stained image to obtain a binarized image, The step b) comprises: b1) performing a noise reduction process on the binarized image; b2) setting the evaluation area in the binarized image and forming a mask image that masks the exclusion area; b3) obtaining the processed image by overlaying the mask image on the stained image; A focus state evaluation method comprising:

6. 5. The focus state evaluation method according to claim 1, The evaluation area is the stained region; a peripheral region of a predetermined width extending from the boundary of the dyed region to the periphery; A focus state evaluation method comprising:

7. 5. The focus state evaluation method according to claim 1, In the step c), the focus evaluation value is acquired for each of a plurality of divided regions obtained by dividing the stained image; A focus state evaluation method for generating a processed stained image in which, of the plurality of divided regions of the stained image, divided regions whose focus evaluation values are less than a predetermined threshold are not displayed.

8. 5. The focus state evaluation method according to claim 1, A focus state evaluation method in which focus evaluation values obtained by performing steps a), b), and c) on a plurality of stained images of the specimen with different focus states are compared, and one stained image with the best focus state is selected from the plurality of stained images.

9. 5. The focus state evaluation method according to claim 1, further comprising a step of performing color separation processing on the stained image prior to the step a), In the step a), the stained region is extracted from the stained image after the color separation process.

10. A stained image analysis method for analyzing a plurality of stained images obtained by a multiplex immunostaining method, comprising: d) acquiring the focus evaluation value for each of a plurality of stained images acquired by a multiple immunostaining method by the focus state evaluation method according to any one of claims 1 to 4; e) analyzing the plurality of stained images in consideration of the focus evaluation value in each stained image; A stained image analysis method comprising:

11. A computer-readable program that causes a computer to evaluate a focus state of a stained image by an immunostaining method, When the program is executed by a computer, a) extracting a stained area from an image of a specimen stained by an immunostaining method; b) setting an evaluation area including the stained area in the stained image, and performing mask processing to exclude an exclusion area that is an area other than the evaluation area, to obtain a processed image; c) examining a focus state for the evaluation region in the processed image and obtaining a focus evaluation value that indicates the focus state of the stained region; A computer-readable program that performs the following:

Images