JP2011043350A - Cell image analyzer, cell image analyzing method, and computer program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To precisely detect a cell membrane from a cell image. <P>SOLUTION: The cell image obtained with imaging a cell stained with a fluorescent substance is inputted, and a cell nucleus region is extracted from the inputted cell image. The cell membrane region positioned outside the cell nucleus region is extracted on the basis of the brightness value of the cell image. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a cell image analysis apparatus, a cell image analysis method, and a computer program for analyzing a cell image in which a cell organ is stained with a fluorescent substance.

Conventionally, a technique has been proposed in which cell organs such as cell membranes and cell nuclei are stained with a fluorescent substance, imaged with a microscope, and the captured cell image is analyzed (see, for example, Patent Document 1). In the conventional cell image analysis technique, binarization is performed by setting an appropriate luminance threshold for an image, setting 1 to a point where the luminance is higher than the threshold, and setting 0 to other points. Each cell organ is recognized by classifying a region that is a set of points set to 1 individually.
For example, a technique has been proposed in which a cell membrane is stained with a fluorescent substance by attaching an antibody to the surface of the cell membrane, the stained cells are imaged with a microscope, and the cell membrane is detected from the captured cell image.

JP 2006-79196 A

However, the brightness may be greatly different between a plurality of cell membranes included in the same cell image due to factors such as a difference in the degree of staining and a mixture of a plurality of types of cells. In particular, it is difficult to uniformly dye cell membranes due to their properties, and uneven staining often occurs. In such a case, the conventional method of analyzing a cell image using a threshold value set for an image causes problems such as oversight of a cell membrane with low luminance.
In addition, depending on the staining method, only the cell membrane may not be stained with high accuracy, and even the cytoplasm may be stained. In such a case, it is difficult to recognize the boundary between the cytoplasm and the cell membrane. It was.

  In view of the above circumstances, an object of the present invention is to provide a cell image analysis apparatus, a cell image analysis method, and a computer program that can accurately detect a cell membrane from a cell image.

  [1] One aspect of the present invention is a cell image analysis device, which includes an image input unit that inputs a cell image obtained by imaging cells stained with a fluorescent substance, and a cell nucleus region extracted from the input cell image And a cell membrane region extraction unit that extracts a cell membrane region located outside the cell nucleus region based on the luminance value of the cell image.

  [2] One aspect of the present invention is the above-described cell image analysis device, wherein the cell membrane region extraction unit is configured by a set of pixels satisfying a predetermined condition in which a change in luminance value with respect to a periphery in the cell image is set in advance. A cell membrane region located outside the cell nucleus region is extracted.

  [3] One aspect of the present invention is the above-described cell image analysis device, wherein the cell membrane region extraction unit is a region existing in a belt-like region set by a predetermined reference based on a boundary line of the cell nucleus region In this case, a region constituted by a set of pixels whose luminance value change with the surroundings satisfies a predetermined condition set in advance is extracted as a cell membrane region.

  [4] One aspect of the present invention is the above-described cell image analysis apparatus, wherein when the cell membrane region of adjacent cells is detected by the cell membrane region extraction unit, the cell membrane of a portion that is not in contact with each cell It further includes a correction unit that detects the thickness of each region and cuts the cell membrane region in contact with each cell membrane region according to the detected thickness.

  [5] One aspect of the present invention is a cell image analysis method, in which the cell image analysis apparatus inputs a cell image obtained by imaging a cell stained with a fluorescent substance, and the cell image analysis apparatus A cell nucleus region extracting step for extracting a cell nucleus region from the input cell image, and a cell membrane for extracting a cell membrane region located outside the cell nucleus region based on a luminance value of the cell image by the cell image analyzer A region extraction step.

  [6] One embodiment of the present invention is a computer program, an image input step of inputting a cell image in which cells stained with a fluorescent material are imaged, and a cell nucleus that extracts a cell nucleus region from the input cell image The information processing apparatus is configured to execute a region extraction step and a cell membrane region extraction step of extracting a cell membrane region located outside the cell nucleus region based on the luminance value of the cell image.

  According to the present invention, it is possible to accurately detect a cell membrane from a cell image.

It is a schematic block diagram showing the functional composition of the cell image analysis device which is the first embodiment of the cell image analysis device. It is a flowchart showing the operation example of a cell image analyzer. It is a figure showing the outline | summary of the luminance value of a cell nucleus dyeing | staining image. It is a figure showing the outline | summary of the luminance value of a cell membrane dyeing | staining image. It is the schematic which shows the outline of a strip | belt-shaped area | region. It is a figure which shows the example of the cell membrane area | region detected. It is a schematic block diagram showing the function structure of the cell image analyzer which is 2nd embodiment of a cell image analyzer. It is a figure showing the outline of a process of a correction part. It is a figure which shows the example of a cell image in case a some cell adjoins mutually and forms one lump.

[First embodiment]
FIG. 1 is a schematic block diagram showing a functional configuration of a cell image analysis apparatus 1a which is a first embodiment of a cell image analysis apparatus. The cell image analysis apparatus 1 a includes an image input unit 101, a cell nucleus region extraction unit 102, and a cell membrane region extraction unit 103. The cell image analysis apparatus 1a may be configured using an information processing apparatus such as a personal computer or a workstation, or may be configured as a dedicated apparatus incorporated in a microscope or the like.

The image input unit 101 inputs digital data of a cell image obtained by capturing cells stained with a fluorescent substance with a microscope to the cell image analyzer 1.
The cell nucleus region extraction unit 102 extracts a region where the cell nucleus exists (hereinafter referred to as “cell nucleus region”) from the input cell image, and acquires information indicating the position and range of the cell nucleus region.

The cell membrane region extraction unit 103 extracts a cell membrane region (hereinafter referred to as “cell membrane region”) from the input cell image, and acquires information representing the position and range of the cell membrane region.
Specifically, the cell membrane region extraction unit 103 is located outside the cell nucleus region extracted by the cell nucleus region extraction unit 102 based on the brightness value of the cell image (hereinafter, “brightness value” is referred to as “brightness value”). The located cell membrane region is extracted. Note that the luminance here is a value representing light and shade in a light and shade image, for example, and is a value representing the brightness of each pixel (pixel) in the image.

  FIG. 2 is a flowchart showing an operation example of the cell image analysis apparatus 1a. Hereinafter, an operation example of the cell image analysis apparatus 1a will be described with reference to FIG. First, the image input unit 101 inputs the digital data of the cell image to the cell image analyzer 1a (step S101). Next, the cell nucleus region extraction unit 102 performs a cell nucleus region extraction process to extract a cell nucleus region (step S102). Execution of the cell nucleus region extraction process reveals the number of cells existing in the cell image, that is, the number of cell membranes. Next, the cell membrane region extraction unit 103 performs a cell membrane region extraction process to extract a cell membrane region (step S103). Then, the cell membrane region extraction unit 103 outputs cell membrane region information (step S104). The cell membrane region information includes information indicating the position and range of each extracted cell membrane, information indicating the number of extracted cell membranes, and the like.

[[Nuclear region extraction processing]]
In step S102, the cell nucleus region extraction process may be realized by any of the following four processes. Which process is used may be selected by the user. Hereinafter, four patterns will be described for specific processing contents of the cell nucleus region extraction processing.

[First nuclear region extraction process]
In the first cell nucleus region extraction process, a cell nucleus stained image is used. The cell nucleus stained image is input by the image input unit 101 as a cell image. FIG. 3 is a diagram showing an outline of the luminance value of the cell nucleus stained image. In the cell nucleus staining image, cell nuclei are stained using a fluorescent material that selectively stains cell nuclei. For this reason, in the cell nucleus stained image, the region where the cell nucleus does not exist (A1 region) is imaged with a low luminance value, and the region where the cell nucleus exists (A2 region) is imaged with a high luminance value.

  The cell nucleus region extraction unit 102 stores a preset first threshold value for cell nuclei. The cell nucleus region extraction unit 102 extracts all regions having a luminance value higher than the first threshold value for cell nuclei as cell nucleus regions in the cell nucleus staining image. Note that the value of the first threshold value for cell nuclei is optimally determined according to the experimental environment by performing a plurality of experiments in advance and taking statistics of the luminance value of the region where the cell nuclei are present in the cell nucleus stained image.

At this time, the cell nucleus region extraction unit 102 may store a preset first threshold for cell nucleus area. In this case, the cell nucleus region extraction unit 102 includes all regions having a luminance value higher than the first threshold value for cell nuclei in the cell nucleus staining image and having an area larger than the first threshold value for cell nucleus area. Extract as In this case as well, the value of the first threshold value for the cell nucleus area is optimally determined according to the experimental environment by conducting a plurality of experiments in advance and taking statistics of the size of the area of the cell nucleus in the stained image. Is done.
In the cell nucleus staining image, the cell nucleus region is typically fixed, and the statistical value works effectively. Therefore, the method using the first threshold for cell nucleus as described above is effective.

[Second nuclear region extraction process]
In the second to fourth cell nucleus region extraction processes, cell nucleus stained images are not used, but cell membrane stained images are used. The cell membrane stained image is input by the image input unit 101 as a cell image. FIG. 4 is a diagram showing an outline of the luminance value of the cell membrane stained image. In the cell membrane stained image, the cell membrane is stained by a method using a fluorescent substance that selectively stains the cell membrane, a method of attaching an antibody to the surface of the cell membrane, or the like. Therefore, in the cell membrane stained image, the cell membrane is imaged with the highest luminance value, the cell membrane shifted from the cytoplasm and the focal plane is imaged with the next highest luminance value, and the region where the cell nucleus is present is the next highest An image is captured with a luminance value, and an area where no cells are present is imaged with the lowest luminance value.

  In the second cell nucleus region extraction process, the cell nucleus region extraction unit 102 stores a preset threshold value for cells. The cell threshold value is optimally determined according to the experimental environment by performing a plurality of experiments in advance and taking statistics on the luminance value of the area where the cells exist. The cell region is a region where cells including a cell membrane, a cytoplasm, and a cell nucleus are present. The cell nucleus region extraction unit 102 stores a preset second threshold value for cell nuclei. The value of the second threshold value for cell nuclei is optimally determined according to the experimental environment by conducting a plurality of experiments in advance and taking statistics of the luminance value of the region where the cell nuclei are present in the cell membrane stained image.

The cell nucleus region extraction unit 102 extracts all regions having luminance values higher than the cell threshold and lower than the cell nucleus second threshold as cell nucleus regions in the cell membrane stained image.
At this time, the cell nucleus region extraction unit 102 may store a preset first threshold for cell nucleus area. In this case, the cell nucleus region extraction unit 102 is a region having a luminance value higher than the cell threshold and lower than the cell nucleus second threshold in the cell membrane stained image, and having an area larger than the cell nucleus area first threshold. Extract all regions having a cell nucleus region. In this case as well, the value of the first threshold value for the cell nucleus area is optimally determined according to the experimental environment by conducting a plurality of experiments in advance and taking the statistics of the size of the area where the cell nucleus is present in the cell membrane stained image. Is done.

[Third nuclear region extraction process]
In the third cell nucleus region extraction process, the cell nucleus region extraction unit 102 stores a preset threshold value for cells. The cell threshold value is optimally determined according to the experimental environment by performing a plurality of experiments in advance and taking statistics on the luminance value of the area where the cells exist. The cell region is a region where cells including a cell membrane, a cytoplasm, and a cell nucleus are present.

  The cell nucleus region extraction unit 102 uses a pixel (hereinafter referred to as a “maximum value pixel”) in which a change in luminance value with respect to the periphery is maximized from within a region having a luminance value higher than the cell threshold value in the cell membrane stained image. The enclosed area is extracted as a cell nucleus area. As is apparent from FIG. 4, in the cell membrane stained image, the luminance values of the cell nucleus region and the cytoplasm region are different from each other, so that the luminance value change between the pixel located at the boundary and the periphery becomes a maximum value. Therefore, it is possible to extract the boundary of the cell nucleus region by extracting the maximum value pixel.

  More specifically, the cell nucleus region extraction unit 102 applies, for example, an edge detection filter to a cell membrane stained image, and extracts a pixel having a luminance value that has a peak luminance value in the filtered image as a maximum value pixel. good.

  In addition, as apparent from FIG. 4, in the cell membrane stained image, the luminance values of the cytoplasm region and the cell membrane region are different from each other, so that the luminance value change between the pixel located at the boundary and the surrounding area becomes the maximum value. Therefore, when the cell nucleus region is extracted, it is necessary for the cell nucleus region extraction unit 102 not to erroneously extract a region surrounded by pixels located at the boundary between the cytoplasm region and the cell membrane region. Two specific examples of such a determination process are shown, but the determination process is not limited to the following two processes.

  In the first method, the cell nucleus region extraction unit 102 calculates the area of the extracted region, and if the calculated area is smaller than a preset second threshold for cell nucleus area, the extracted region is a cell nucleus. Judged to be an area. In this case, the cell nucleus region extraction unit 102 stores a preset second threshold for cell nucleus area. The value of the second threshold value for the cell nucleus area is optimally determined according to the experimental environment by conducting a plurality of experiments in advance and taking statistics of the area size of the area where the cell nucleus is present in the cell membrane stained image.

  In the second method, the cell nucleus region extraction unit 102 determines that the extracted region is a cell nucleus region when there are not more than a predetermined number of maximum value pixels inside the extracted region. In this case, the predetermined number of pixels may be 1 or 2 or more.

  In the third cell nucleus region extraction process, a cell nucleus region is extracted based on a change in luminance value without using an absolute threshold relating to the luminance value, so that extraction can be performed with high accuracy even when the luminance differs from cell to cell. it can.

[Fourth nuclear region extraction process]
In the fourth cell nucleus region extraction process, the cell nucleus region extraction unit 102 stores a preset threshold value for cells. The cell threshold value is optimally determined according to the experimental environment by performing a plurality of experiments in advance and taking statistics on the luminance value of the area where the cells exist. The cell region is a region where cells including a cell membrane, a cytoplasm, and a cell nucleus are present.

  The cell nucleus region extraction unit 102 detects a pixel having the lowest luminance value in a region having a luminance value higher than the cell threshold value in the cell membrane stained image, and calculates a luminance value within a predetermined range from the luminance value of this pixel. A set of pixels is extracted as a cell nucleus region. As is clear from FIG. 4, in the cell membrane stained image, the luminance value of the cell nucleus region is the lowest compared to the luminance values of other regions. Therefore, the pixel having the lowest luminance value is a pixel in the cell nucleus region, and the cell nucleus region can be extracted by extracting a set of pixels having luminance values close to this pixel.

  In the fourth cell nucleus region extraction process, the cell nucleus region is extracted based on the difference in relative brightness value in the cell region without using an absolute threshold value for the brightness value, and thus the brightness differs for each cell. Even in cases, extraction can be performed with high accuracy.

[[Cell membrane region extraction processing]]
In step S103, the cell membrane region extraction process may be realized by either of the following two processes. Which process is used may be selected by the user. In the following, two specific patterns of the cell membrane region extraction process will be described.

[First cell membrane region extraction process]
In the first cell membrane region extraction process, the cell membrane region extraction unit 103 sets a set of maximum value pixels from within a region having a luminance value higher than the threshold value for cells stored in the cell nucleus region extraction unit 102 in the cell membrane stained image. The region constituted by is extracted as a cell membrane region. As is clear from FIG. 4, in the cell membrane stained image, the luminance values of the cell membrane region and the cytoplasm region are different from each other, so that the luminance value change between the pixel located at the boundary and the periphery becomes a maximum value. Therefore, it is possible to extract the cell membrane region by extracting the maximum value pixel.

  More specifically, the cell membrane region extraction unit 103 applies, for example, an edge detection filter to a cell membrane stained image, and extracts a pixel having a luminance value that peaks in the filtered image as a maximum value pixel. good.

  Further, as is clear from FIG. 4, in the cell membrane stained image, the luminance values are different between the cell nucleus region and the cytoplasm region, and therefore, the luminance value change between the pixels located at the boundary and the surroundings becomes the maximum value. Therefore, when extracting the cell membrane region, the cell membrane region extraction unit 103 needs not to mistakenly extract a region constituted by a set of pixels located at the boundary between the cytoplasm region and the cell nucleus region. Therefore, the cell membrane region extraction unit 103 extracts, as a cell membrane region, a region constituted by a set of local maximum pixels located outside the cell nucleus region extracted by the cell nucleus region extraction unit 102. At this time, the cell membrane region extraction unit 103 may extract, as a cell membrane region, a region constituted by a set of maximum value pixels located at a predetermined distance or more away from the boundary of the cell nucleus region.

  In the first cell membrane region extraction process, since the cell membrane region is extracted on the condition that it is located outside the cell nucleus region, it is possible to prevent the boundary line of the cell nucleus region from being erroneously extracted as the cell membrane region. In addition, since the cell membrane region is extracted based on the change in the brightness value without using an absolute threshold value for the brightness value, the extraction can be performed with high accuracy even when the brightness is different for each cell. It is also effective for analysis of cell images in which the cell membrane of one cell does not have uniform brightness.

[Second cell membrane region extraction process]
In the second cell membrane region extraction process, the cell membrane region extraction unit 103 first sets a band-like region on the basis of a predetermined reference outside the boundary line of the cell nucleus region in the cell membrane stained image. FIG. 5 is a schematic diagram showing an outline of the band-like region. In FIG. 5, a region A2 represents a cell nucleus region, and A0 represents a strip region. As shown in the figure, the band-like region is separated from each point of the boundary line of the cell nucleus region by a distance (W1 + W2) from the boundary line L1 represented by a set of points separated by a distance W1 from each point of the cell nucleus region boundary line. It is defined as a region between the boundary line L2 represented by the set of points. Note that the above-described band-shaped region setting method is an example of a predetermined criterion, and another criterion may be applied as the predetermined criterion for setting the band-shaped region based on the boundary line of the cell nucleus region. In addition, the values of W1 and W2 are optimized according to the experimental environment by conducting a plurality of experiments in advance and taking the statistics of the distance between the region where the cell membrane is present and the cell nucleus in the cell to be detected. To be determined.

Then, the cell membrane region extraction unit 103 extracts, as a cell membrane region, a region constituted by a set of maximum value pixels located within the set band-like region.
Even in the second cell membrane region extraction process, the cell membrane region is extracted on the condition that it is located within the belt-like region set outside the cell nucleus region, so that the boundary line of the cell nucleus region is erroneously extracted as the cell membrane region. In addition to being able to prevent, it is possible to prevent the noise at a position too far from the cell nucleus region from being erroneously extracted as a cell membrane region.

  In addition, since the cell membrane region is extracted based on the change in the brightness value without using an absolute threshold value for the brightness value, the extraction can be performed with high accuracy even when the brightness is different for each cell. It is also effective for analysis of cell images in which the cell membrane of one cell does not have uniform brightness. Cell nucleus detection is possible with high accuracy, but is also effective when the cell membrane is not very clear.

<Modification>
In the third cell nucleus region extraction process, the first cell membrane region extraction process, and the second cell membrane region extraction process, the cell nucleus region extraction unit 102 or the cell membrane region extraction unit 103 replaces the maximum value pixel with a luminance value with the periphery. A pixel whose change exceeds a threshold value may be detected, and a region surrounded by such a pixel may be extracted as a cell nucleus region or a cell membrane region. The threshold in this case is set in advance for the cell nucleus region extraction unit 102 or the cell membrane region extraction unit 103.

  Further, in the cell membrane region extraction process, the cell membrane region extraction unit 103 detects all the pixels that satisfy the predetermined condition with reference to this pixel after detecting the maximum value pixel or the pixel whose luminance value change with the surroundings exceeds the threshold value, It may be detected as a pixel constituting the cell membrane region, and a cell membrane region including all the detected pixels may be extracted. FIG. 6 is a diagram showing an example of a cell membrane region detected by such processing. The upper part of FIG. 6 represents a part of the cell membrane image in the cell membrane stained image and a part of the pixels in the cell membrane stained image (64 pixels a1 to h8). The lower part of FIG. 6 represents the luminance values of the pixels h1 to h8.

  For example, when the pixel of h2 becomes a maximum value pixel or a pixel whose luminance value change with the surroundings exceeds the threshold, the pixel satisfying a predetermined condition with the pixel of h2 as a reference is regarded as a pixel of the cell membrane region by the cell membrane region extraction unit 103 Detected. The predetermined condition is appropriately set so that the cell membrane to be detected can be accurately detected according to the experimental environment. For example, it may be set as the predetermined condition that the pixel adjacent to either the reference pixel or the pixel satisfying the predetermined condition has a luminance value larger than the luminance value of the reference pixel. In the case of this condition, the h3 and h4 pixels are detected by the cell membrane region extraction unit 103 as pixels of the cell membrane region after h2 as a reference in FIG. By performing such processing, the cell membrane region can be accurately extracted when the cell membrane has a thickness (width).

[Second Embodiment]
FIG. 7 is a schematic block diagram showing a functional configuration of a cell image analysis apparatus 1b which is a second embodiment of the cell image analysis apparatus. The cell image analysis apparatus 1b includes an image input unit 101, a cell nucleus region extraction unit 102, a cell membrane region extraction unit 103, and a correction unit 104. The cell image analysis apparatus 1b included in the cell image analysis apparatus 1b may be configured using an information processing apparatus such as a personal computer or a workstation, or may be configured as a dedicated apparatus incorporated in a microscope or the like.
Since the configuration of the image input unit 101, the cell nucleus region extraction unit 102, and the cell membrane region extraction unit 103 is the same as each configuration of the cell image analysis apparatus 1a of the first embodiment, description thereof is omitted.

  The correction unit 104 detects the thickness of each cell membrane region of the portion in contact with the adjacent cell, and cuts the two cell membrane regions into the portion in contact based on the detected thickness. FIG. 8 is a diagram illustrating an outline of processing of the correction unit 104. When there is an adjacent cell, the correction unit 104 recognizes the thickness based on the cell membrane region of the portion where each adjacent cell is not in contact. In FIG. 8, W11 represents the thickness of the part of the cell membrane not touching the left cell, and W12 represents the thickness of the part of the cell membrane not touching the right cell. In addition, the correction unit 104 detects the thickness W13 of the contacting part of the cell membrane region. Then, the correcting unit 104 cuts the thickness W13 according to the ratio between W11 and W12. For example, the correction unit 104 assigns a thickness of W13 × W11 / (W11 + W12) to the cell membrane region of the left cell, assigns a thickness of W13 × W12 / (W11 + W12) to the cell membrane region of the right cell, The cell membrane region is cut according to the thickness.

  According to the cell image analyzing apparatus 1b of the second embodiment configured as described above, the cell membrane region of the contacted portion due to the presence of adjacent cells is separated and recognized as the cell membrane region of each cell. Is possible.

<Modification>
FIG. 9 is a diagram showing an example of a cell image when a plurality of cells are adjacent to each other to form one lump. In this case, the correction unit 104 first detects the line LL that forms the outline of the lump. Next, the correcting unit 104 detects a cell having the longest length of the portion in contact with the line LL that forms the outline among the cells included in the lump. Next, the correcting unit 104 repeatedly performs the process of dividing the cell membrane region with other adjacent cells for the detected cells. The correction unit 104 recognizes this cell as a cell that is not included in the lump when the separation process of the detected cell from all the other adjacent cells is completed, and detects the line LL that forms the outline of the lump again. . Then, the correcting unit 104 detects the cell having the longest length in the portion in contact with the line LL that forms the outline among the cells included in the lump, and repeats the above processing.

  By configuring the correction unit 104 in this way, it becomes possible to perform the separation process more efficiently and accurately than in the case of performing the separation process from a cell or the like located inside the lump. That is, by performing the separation process on the cells in contact with the outer shell, it becomes possible to accurately detect the thickness of the cell membrane region (W11 and W12 in FIG. 8) for this cell, and the cell membrane region of the contacted portion can be determined. It is possible to accurately perform the separation process.

  You may make it implement | achieve the function of the cell image analyzer 1a, 1b in embodiment mentioned above with a computer. In that case, a program for realizing each function may be recorded on a computer-readable recording medium, and the program recorded on the recording medium may be read into a computer system and executed. Here, the “computer system” includes an OS and hardware such as peripheral devices. The “computer-readable recording medium” refers to a portable medium such as a flexible disk, a magneto-optical disk, a ROM, and a CD-ROM, and a storage device such as a hard disk built in the computer system. Furthermore, the “computer-readable recording medium” dynamically holds a program for a short time like a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line. In this case, a volatile memory inside a computer system serving as a server or a client in that case may be included and a program that holds a program for a certain period of time. The program may be a program for realizing a part of the functions described above, and may be a program capable of realizing the functions described above in combination with a program already recorded in a computer system.

  The embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to this embodiment, and includes designs and the like that do not depart from the gist of the present invention.

DESCRIPTION OF SYMBOLS 1 ... Cell image analyzer, 101 ... Image input part, 102 ... Cell nucleus area | region extraction part, 103 ... Cell membrane area | region extraction part, 104 ... Correction part

Claims (6)

An image input unit for inputting a cell image obtained by imaging cells stained with a fluorescent substance;
A cell nucleus region extraction unit for extracting a cell nucleus region from the input cell image;
A cell membrane region extraction unit that extracts a cell membrane region located outside the cell nucleus region based on the luminance value of the cell image;
A cell image analysis apparatus comprising:   The cell membrane region extraction unit extracts a cell membrane region located outside the cell nucleus region, the region including a set of pixels satisfying a predetermined condition in which a change in luminance value with respect to a periphery in the cell image is set in advance. The cell image analysis apparatus according to claim 1, wherein:   The cell membrane region extraction unit is a region that exists in a belt-like region that is set based on a predetermined reference based on a boundary line of the cell nucleus region, and a pixel that satisfies a predetermined condition in which a luminance value change with the surroundings is set in advance The cell image analysis apparatus according to claim 1, wherein a region constituted by the set of cells is extracted as a cell membrane region.   When cell membrane regions of cells adjacent to each other are detected by the cell membrane region extraction unit, the thicknesses of the cell membrane regions that are not in contact with each cell are detected, and the portions that are in contact with each other according to the detected thickness The cell image analysis apparatus according to claim 1, further comprising a correction unit that cuts the cell membrane regions into respective cell membrane regions. An image input step in which the cell image analyzer inputs a cell image in which cells stained with a fluorescent material are imaged;
A cell nucleus region extraction step in which the cell image analyzer extracts a cell nucleus region from the input cell image;
A cell membrane region extraction step in which the cell image analysis device extracts a cell membrane region located outside the cell nucleus region based on a luminance value of the cell image;
A cell image analysis method comprising: An image input step for inputting a cell image in which cells stained with a fluorescent material are imaged;
A cell nucleus region extraction step for extracting a cell nucleus region from the input cell image;
A cell membrane region extraction step for extracting a cell membrane region located outside the cell nucleus region based on the luminance value of the cell image;
Is a computer program for causing an information processing apparatus to execute.

Images