Cell diagnostic support program, cell diagnostic support method, and cell diagnostic support system

Abstract

To improve cell diagnostic support technology. [Solution] A cell diagnostic support program that causes an information processing device to perform operations including selecting at least one object of interest from a low-magnification image based on a predetermined analysis model, setting target coordinates on the at least one object of interest, setting a corresponding region of interest based on the target coordinates, acquiring a high-magnification image of the region of interest, and outputting the high-magnification image.

Description

Technical Field

[0001] The present disclosure relates to a cytodiagnosis support program, a cytodiagnosis support method, and a cytodiagnosis support system.

Background Art

[0002] Cytodiagnosis is a diagnostic method in which cells obtained as a specimen are observed under a microscope, and based on the morphological observation results, the benign and malignant nature of a tumor and the tumor name are estimated. Conventionally, various attempts have been made to improve the diagnostic accuracy and efficiency of cytodiagnosis.

[0003] For example, in the technique disclosed in Patent Document 1, after dividing a low-magnification image into a plurality of regions that are candidates for regions of interest, the low-magnification image is input into a neural network to select a region of interest, and the selected region of interest is enlarged to generate a high-magnification image. Further, in such a technique, the high-magnification image is input into a further neural network to analyze whether the high-magnification image has target features and generate a statistical result regarding the target features. According to such a technique, the analysis efficiency of a specimen can be improved.

[0004] Also, in the technique disclosed in Patent Document 2, there is disclosed an image processing apparatus including means for acquiring an image obtained by imaging a specimen observed using a microscope, and means for specifying a region to be magnified and observed based on an evaluation of a detection target included in the image and the magnification of the microscope when the image was taken. Here, in Patent Document 2, when specifying the region to be magnified and observed, it is specifically described that after dividing the image into a plurality of regions (in a specific example, a plurality of grid-like regions) in advance, each region is weighted to provide support information for specifying the region to be magnified and observed.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

[0006] The conventional diagnostic support technologies described above had room for improvement. Specifically, the technology disclosed in Patent Document 1 divides a low-magnification image into multiple candidate regions of interest, and then inputs these low-magnification images into a neural network to select a region of interest. Similarly, Patent Document 2 describes a method in which an image is divided into multiple regions in advance, and then support information is presented to determine the necessity of magnified observation for each region. As a result, during the division process, objects that would otherwise be preferable to be included as a single "region of interest" and used as the target of magnified observation are divided into other regions, potentially preventing sufficient improvement in diagnostic accuracy and efficiency. Thus, the conventional diagnostic support technologies described above had room for improvement. In view of these circumstances, the purpose of this disclosure is to improve cell diagnostic support technology. [Means for solving the problem]

[0007] A diagnostic support program according to one embodiment of this disclosure is: In an information processing device, Based on low-magnification images derived from microscopic images of multiple cell samples, at least one target of interest is selected based on a predetermined analytical model. Setting target coordinates on at least one object of interest, setting a corresponding region of interest based on the target coordinates, and acquiring a high-magnification image of the region of interest; Outputting the aforementioned high-magnification image, Perform an action that includes this.

[0008] A cell diagnostic support method according to one embodiment of this disclosure is: Based on low-magnification images derived from microscopic images of multiple cell samples, at least one target of interest is selected based on a predetermined analytical model. Setting target coordinates on at least one object of interest, setting a corresponding region of interest based on the target coordinates, and acquiring a high-magnification image of the region of interest, Outputting the aforementioned high-magnification image, Includes.

[0009] A cell diagnostic support system according to one embodiment of this disclosure is A diagnostic support system comprising an imaging unit that acquires microscopic images of multiple cell samples, and a control unit, wherein the control unit, A low-magnification image is generated based on the microscopic images of the aforementioned multiple cell samples. From the aforementioned low-magnification images, at least one object of interest is selected based on a predetermined analysis model. A target coordinate is set on at least one object of interest, a corresponding region of interest is set based on the target coordinate, and a high-magnification image of the region of interest is acquired. The device is characterized by outputting the acquired high-magnification image. [Effects of the Invention]

[0010] According to one embodiment of this disclosure, cell diagnostic support technology is improved. [Brief explanation of the drawing]

[0011] [Figure 1] This is a schematic diagram of an information processing system comprising an information processing device on which a cell diagnostic support program according to one embodiment of the present disclosure is executed. [Figure 2] This is a flowchart showing the flow of the cell diagnostic support method performed by the cell diagnostic support program. [Figure 3] This diagram illustrates the general procedure for stapling microscopic images of cell specimens. [Figure 4a] This figure shows an example of a region of interest set on a low-magnification image. [Figure 4b] This figure illustrates high-magnification images corresponding to each region of interest shown in Figure 4(a). [Modes for carrying out the invention]

[0012] Hereinafter, embodiments of the present disclosure will be described.

[0013] (Overview of Embodiment) Referring to FIG. 1, an overview of a cytodiagnosis support system including an information processing apparatus on which a cytodiagnosis support program according to an embodiment of the present disclosure is executed will be described. Typically, the cytodiagnosis support system 1 includes an information processing apparatus 10 on which the cytodiagnosis support program according to the present embodiment is executed, and an imaging unit 11. Further, as shown in FIG. 1, the information processing apparatus 10 may include an input unit 101, a control unit 102, a storage unit 103, and an output unit 104. The imaging unit 11 may include an imaging element 111 and an output unit 112. Details of each component will be described later.

[0014] First, an overview of the present embodiment will be described, and details will be described later. The cytodiagnosis support system 1 obtains microscopic images of a plurality of cell specimens by the imaging unit 11. The obtained plurality of microscopic images are captured by the information processing apparatus 10 through the output unit 112 of the imaging unit 11 and the input unit 101 of the information processing apparatus 10. Then, the control unit 102 of the information processing apparatus 10 generates a low-magnification image based on the captured microscopic images of the plurality of cell specimens, and at least one target of interest is selected from such a low-magnification image based on a predetermined analysis model. Further, in the control unit 102, target coordinates are set on at least one target of interest. The control unit 102 may also detect the size of the selected at least one target of interest from its state of existence or the like. Then, the control unit 102 sets a region of interest corresponding to the target of interest based on the set target coordinates. Further, the control unit 102 acquires a high-magnification image of the set region of interest and outputs the acquired high-magnification image.

[0015] In this way, by selecting a target of interest on a low-magnification image of a cell specimen, setting a region of interest based on the position (target coordinates) of the selected target of interest, and acquiring a high-magnification image of the region of interest, the diagnostic accuracy and diagnostic efficiency of cytodiagnosis can be improved. Specifically, by setting a region of interest based on the position of the target of interest, it is possible to output a high-magnification image that necessarily contains the selected target of interest, so it is considered that the diagnostic accuracy can be enhanced. Also, compared to the conventional case where a target of interest is detected from a pre-segmented region state and a plurality of regions containing a single target of interest are set as the regions of interest, by setting a region of interest based on the position of the target of interest, it is possible to suppress the output of high-magnification images for unnecessary parts, and it is considered that the diagnostic efficiency can be increased.

[0016] Next, each component of the cytodiagnosis support system 1 will be described in detail with reference to FIG. 1.

[0017] (Configuration of the cytodiagnosis support system) As described above, the cytodiagnosis support system 1 includes an information processing device 10 and an imaging unit 11. The information processing device 10 may include an input unit 101, a control unit 102, a storage unit 103, and an output unit 104. The imaging unit 11 may include an image sensor 111 and an output unit 112.

[0018] The imaging unit 11 can be a microscope equipped with an image sensor 111. The type of microscope is not particularly limited, and examples include an optical microscope and a fluorescence microscope. Although not shown, the imaging unit 11 may include a light source that irradiates illumination light during imaging. The specimen is not particularly limited, and for example, it can be cells collected from a subject applied to a slide glass or the like and subjected to processing such as staining as necessary. Then, the imaging unit 11 images the cell specimen with the image sensor 111 to acquire a plurality of microscope images. The imaging magnification can be arbitrarily set according to the type of specimen and the purpose of cytodiagnosis. The plurality of microscope images of the cell specimen acquired by the imaging unit 11 can be output from the output unit 112. The output unit 112 may include an interface for connecting to an external device.

[0019] Microscopic images of cell specimens output from the output unit 112 of the imaging unit 11 are input to the information processing device 10 via the input unit 101. The input unit 101 is not particularly limited and may include an interface for connecting to an external device. The microscopic images of cell specimens are then processed by the control unit 102. The microscopic images of cell specimens may optionally be stored in the storage unit 103 before or after processing in the control unit 102.

[0020] The control unit 102 includes one or more processors, one or more programmable circuits, one or more dedicated circuits, or a combination thereof. The processor is a general-purpose processor such as a CPU (Central Processing Unit) or GPU (Graphics Processing Unit), or a dedicated processor specialized for a specific process, but is not limited to these. The programmable circuit is an FPGA (Field-Programmable Gate Array), but is not limited to this. The dedicated circuit is an ASIC (Application Specific Integrated Circuit), but is not limited to this. Hereinafter, processors, programmable circuits, and dedicated circuits will not be particularly distinguished and will also be referred to as "processors, etc." The control unit 102 can control the operation of the entire cell diagnostic support system 1.

[0021] The control unit 102 performs various processes based on microscopic images of multiple cell samples, as will be described later with reference to Figure 2. Specifically, it selects at least one object of interest from low-magnification images based on microscopic images of multiple cell samples, based on a predetermined analysis model. Furthermore, the control unit 102 sets target coordinates on at least one object of interest and sets a corresponding region of interest based on the target coordinates.

[0022] The storage unit 103 includes one or more memories. These memories are, for example, semiconductor memories, magnetic memories, or optical memories, but are not limited to these. Each memory included in the storage unit 103 may function as, for example, a main memory, an auxiliary memory, or a cache memory. The storage unit 103 stores any information used in the operation of the cell diagnostic support system 1. For example, the storage unit 103 may store system programs, application programs, and embedded software.

[0023] For example, the memory unit 103 stores a predetermined analysis model used by the control unit 102 when analyzing low-magnification images. Such a predetermined analysis model may be an AI (artificial intelligence) classification model, or a supervised classification model based on previously acquired training data without machine learning. Furthermore, these analysis models may be periodically updated via the input unit 101. The AI ​​classification model is trained, for example, by supervised learning. In one example, the AI ​​classification model is trained to detect an object of interest on an input low-magnification image by supervised learning using training data that includes a microscope image at approximately the same magnification as the low-magnification image and annotations for the object of interest on the image. However, the AI ​​learning model is not limited to this example, and may be trained by any machine learning algorithm, such as deep learning.

[0024] Furthermore, in one example, the focus of attention is on cells or groups of cells suspected of being lesions. The predetermined analytical model is, for example, an analytical model based on a machine learning model of cancer lesions. In addition, an appropriate model can be selected depending on the lesion to be diagnosed.

[0025] The output unit 104 includes one or more output devices for outputting information. These output devices may be, for example, displays. Alternatively, the output unit 104 may include an interface for connecting an external output device.

[0026] In this embodiment, for the sake of simplicity, the above-described components of the cell diagnostic support system 1 are described as being located in a single device. However, the components of the cell diagnostic support system 1 may be distributed across multiple devices that can communicate with each other. When the above-described components are distributed across multiple devices, the control unit 102 includes multiple processors, and each of the multiple devices is equipped with at least one processor. In one example, the imaging unit 11 may be configured as a single independent device (for example, a microscope).

[0027] (Cell diagnostic support program) The processing flow of the cell diagnostic support program according to this embodiment will be described with reference to Figures 2-4. The cell diagnostic support method according to this embodiment may also follow a similar flow.

[0028] Step S100: The imaging unit 11 starts imaging the cell sample using the image sensor 111.

[0029] Typically, a user performing a cytological diagnosis places the cell sample so that the imaging unit 11 can image it, and then initiates an operation to start imaging on the cytological diagnosis support system 1. In response to this operation, the imaging unit 11 begins imaging the placed cell sample.

[0030] Step S101: The control unit 102 acquires the output of the image sensor 111 from the imaging unit 11.

[0031] Specifically, the control unit 102 acquires a microscopic image of the cell sample output by the image sensor 111. The microscopic image of the cell sample acquired in this step may be a low-magnification image in order to capture the overall appearance of the cell sample.

[0032] Step S102: The control unit 102 generates a low-magnification image of the entire cell sample based on the multiple microscopic images of the cell sample output from the image sensor 111.

[0033] In one example, a low-magnification image of the entire cell sample is a concatenated image obtained by tiling (stacking) multiple microscopic images of the cell sample. Referring to Figure 3, the general procedure for stacking multiple microscopic images of cell samples is explained. As shown in Figure 3(a), when imaging the cell sample, the entire cell sample is imaged in a manner in which the imaging ranges partially overlap. Then, as shown in Figure 3(b), the captured microscopic images are image-processed to perform feature extraction processing such as binarization. Then, as shown in Figure 3(c), adjacent imaging ranges are tiled based on the features. Furthermore, as shown in Figure 3(d), a natural concatenated image can be obtained by performing known image processing such as shading correction in the overlapping areas between adjacent imaging ranges.

[0034] Here, when stitching microscope images, if the same object is captured in both adjacent imaging ranges, as in imaging ranges a1 to a3 in Figure 3(a), the adjacent imaging ranges can be stitched together by positioning them so that the object aligns. However, in Figure 3(a), imaging range a4 does not contain any object common to either the adjacent imaging ranges a2 or a3. Therefore, imaging range a4 cannot be positioned using the captured object as a guide, similar to the other imaging ranges a1 to a3. In such cases, positioning can be performed based on the number of overlapping pixels with adjacent images in nearby imaging ranges. For example, the number of overlapping pixels in at least two imaging ranges located above, below, to the left and right of imaging range a4 can be calculated, and the average value of the left / right overlapping pixel count and the average value of the up / down overlapping pixel count can be used as the number of overlapping pixels between imaging range a4 and its adjacent imaging ranges. Note that the number of imaging ranges from which the average value is taken is not limited to at least two located above, below, to the left and right of the imaging range, but can be any integer between 3 and 100.

[0035] Step S103: The control unit 102 selects at least one object of interest based on a predetermined analysis model.

[0036] In one example, the control unit 102 analyzes the low-magnification image obtained in step S102 using the predetermined analysis model described above to select a target of interest.

[0037] Step S104: The control unit 102 sets target coordinates on the object of interest and sets the corresponding region of interest based on the target coordinates.

[0038] An example of how to set the region of interest will be explained with reference to Figure 4(a). Figure 4(a) shows the state in which three regions of interest 401 to 403, each corresponding to an object of interest a to c, have been set on the low-magnification image 400 obtained in step S102. Region of interest 401 is a region set based on the target coordinates (x,y) of the object of interest. In one example, the target coordinates (x,y) are the center coordinates of object of interest a. Note that the location of the target coordinates (x,y) on the object of interest is not particularly limited, as long as the position of object of interest a can be identified. Similarly, regions of interest 402 and 403 are also set based on target coordinates, but in Figure 4(a), the corresponding target coordinates are not shown for these regions of interest 402 and 403 for clarity. Also, in Figure 4(a), each region of interest 401 to 403 is shown as being of approximately the same size, but this is not limited to the example. For example, the size of the region of interest may be determined based on the size of the object of interest.

[0039] The size of the region of interest preferably includes the entire object of interest. More preferably, the size of the region of interest should include not only the lesion cells that are the object of interest, but also cells other than lesion cells (cells other than the object of interest), in other words, cells that appear to be normal. The determination of whether a cell is a lesion cell or a normal cell may be performed according to an AI classification model, or according to a classification method other than an AI classification model. By setting the region of interest to include both lesion cells and cells that appear to be normal, when high-magnification images of the region of interest are acquired in a later step, lesion cells and cells that appear to be normal can be compared within the same field of view, thereby improving the ease of diagnosis when high-magnification images are presented to diagnosticians such as doctors as diagnostic support information. In particular, the relative size difference between normal cells and lesion cells can be intuitively grasped regardless of the magnification of the high-magnification image, further improving the ease of diagnosis.

[0040] Furthermore, when the region of interest includes not only diseased cells but also cells that appear to be normal, the control unit 102 can store information as associated information for the region of interest, indicating which parts of the region of interest are diseased cells and which parts are cells that appear to be normal.

[0041] Step S105: The control unit 102 acquires a high-magnification image of the region of interest.

[0042] Preferably, the control unit 102 acquires high-magnification images with different depths of focus from the imaging unit 11. The imaging unit 11 acquires high-magnification images of the region of interest by varying the depth of focus, preferably by two or more steps, more preferably by three or more steps, even more preferably by five or more steps, and preferably by 20 or fewer steps. If the number of images with different depths of focus is above the lower limit, the accuracy of cytological diagnosis can be further improved. Also, if the number of images with different depths of focus is below the upper limit, the efficiency of cytological diagnosis can be further improved. Note that a high-magnification image corresponding to one region of interest may be a single microscope image corresponding to one imaging range.

[0043] Step S106: The control unit 102 outputs a high-magnification image of the region of interest via the output unit 112.

[0044] Specifically, the control unit 102 outputs a high-magnification image via the display of the output unit 112. The magnification of the high-magnification image can be arbitrarily set according to the size of the region of interest and the type of disease targeted for cytological diagnosis. For example, the magnification of the high-magnification image is preferably 10 times or more than that of the low-magnification image, more preferably 40 times or more, and may be, for example, 1000 times or less.

[0045] Here, high-magnification images can be presented in a manner that allows them to be referenced simultaneously with the overall low-magnification image, as shown in Figures 4(a) and (b) arranged vertically. For each region of interest 401 to 403 selected on the low-magnification image 400 in Figure 4(a), high-magnification images 401' to 403' can be presented as shown in Figure 4(b). Note that in the illustrated embodiment, the relative relationship between the magnification of the low-magnification image and the magnification of the high-magnification image is merely an example. By outputting high-magnification images of regions of interest in this manner, diagnosticians such as physicians can visually grasp not only the lesion cells but also their relative relationship with surrounding cells, thereby further improving the accuracy and efficiency of cytological diagnosis.

[0046] Furthermore, as shown in Figures 4(a) and 4(b), the control unit 102 can output high-magnification images in a state where they can be simultaneously referenced with low-magnification images, and can also present diagnostic support information obtained from the high-magnification images. Such diagnostic support information includes the type of lesion cells, the stage if the lesion cells are cancerous, the location of discovery, the number of lesion cells in each region of interest, and the ratio of lesion cells to the total number of cells in each region of interest. This diagnostic support information is not particularly limited and can be provided based on an AI diagnostic model constructed for the lesion to be diagnosed.

[0047] As described above, the cell diagnostic support program according to this embodiment first selects at least one target of interest from low-magnification images based on microscopic images of multiple cell samples, based on a predetermined analysis model. The program then sets target coordinates on at least one target of interest, sets the position and size of the corresponding region of interest based on the target coordinates and the size of at least one target of interest, and acquires a high-magnification image of the region of interest. Furthermore, the program outputs the high-magnification image obtained as described above.

[0048] This configuration allows for the output of high-magnification images based on the position and size of the selected object of interest, thereby improving cell diagnostic support technology.

[0049] While this disclosure has been described based on the drawings and embodiments, it should be noted that those skilled in the art may make various modifications and alterations based on this disclosure. Therefore, it should be noted that these modifications and alterations are within the scope of this disclosure. For example, the functions, etc., included in each component or step can be rearranged in a logically consistent manner, and multiple components or steps can be combined into one or divided into two.

[0050] For example, in the embodiment described above, it was described as a preferred configuration in step S106 to output a high-magnification image and simultaneously present diagnostic support information obtained from the high-magnification image. However, it is also possible to provide the diagnostician with a high-magnification image in the manner shown in Figures 4(a) and 4(b), without accompanying diagnostic support information obtained from the high-magnification image. This is because an experienced diagnostician may be able to intuitively determine whether further analysis is necessary and make a diagnosis based on the high-magnification image.

[0051] Furthermore, in the embodiment described above, in step S102, when the control unit 102 tiles multiple adjacent microscope images of cell samples, a technique was described in which the position determination of adjacent images without overlapping objects is performed based on the average number of overlapping pixels between other adjacent images that do have overlapping objects. However, the position determination of adjacent images without overlapping objects may also be performed by machine learning based on image features derived from sources other than the object, such as image grayscale.

[0052] Furthermore, in the embodiments described above, it was explained that in step S105, if there are multiple microscope images for a single region of interest, the acquired multiple high-magnification microscope images can be linked together. In such cases, that is, when the size of the region of interest is large, the field of view may be adjusted so that particularly important areas within the region of interest are captured within a single imaging range when acquiring high-magnification microscope images. This operation can be performed automatically or manually.

[0053] Furthermore, in the embodiments described above, when acquiring a high-magnification image corresponding to a region of interest in step S105, it was explained that the high-magnification image corresponding to one region of interest may be a single microscope image corresponding to one imaging range. However, the high-magnification image may be an image based on multiple microscope images corresponding to multiple imaging ranges, for example, when the magnification range of the imaging unit 11 is large, or when the object of interest covers a wide area. If there are multiple microscope images for a single region of interest, the acquired multiple high-magnification microscope images may be concatenated according to a known method, as explained in step S102 with reference to Figure 3, to provide a single high-magnification image for each region of interest.

[0054] Furthermore, in the embodiments described above, when outputting a high-magnification image of the region of interest in step S106, a method was described in which the low-magnification image and the high-magnification image are presented side by side on a single screen. However, the invention is not limited to this method, and for example, the high-magnification image of the region of interest can also be presented superimposed on the low-magnification image. The presentation position of the high-magnification image depends on the number and size of the objects of interest to be displayed, but it is preferable that it is mapped to some extent based on the position where the objects of interest are detected.

[0055] Furthermore, for example, in the above embodiments, a program that performs all or part of the computer's functions or processes can be recorded on a computer-readable recording medium. The computer-readable recording medium includes non-temporary computer-readable media, such as magnetic recording devices, optical discs, magneto-optical recording media, or semiconductor memory. The program can be distributed, for example, by selling, transferring, or lending portable recording media such as DVDs (Digital Versatile Discs) or CD-ROMs (Compact Disc Read Only Memory) on which the program is recorded. Alternatively, the program can be distributed by storing it in the storage of any server and transmitting it from any server to other computers. The program can also be provided as a program product. Embodiments of this disclosure may include a system, a program, and a storage medium on which the program is recorded (for example, an optical disc, magneto-optical disc, CD-ROM, CD-R, CD-RW, magnetic tape, hard disk, or memory card).

[0056] The implementation form of the program is not limited to application programs such as object code compiled by a compiler or program code executed by an interpreter, but may also be in the form of a program module embedded in an operating system. Furthermore, the program may or may not be configured so that all processing is performed only on the CPU on the control board. The program may be configured so that some or all of its processing is performed by another processing unit implemented on an expansion board or expansion unit attached to the board, as needed. [Explanation of symbols]

[0057] 1. Cell diagnostic support system 10 Information Processing Devices 11 Imaging Unit 111 Image sensor 112 Output section 101 Input Section 102 Control Unit 103 Storage section 104 Output section a1~a4 Imaging area a-c Points of interest 400 Low-magnification image 401-403 Areas of Interest 401'~403' High-magnification images

Claims

1. In an information processing device, Based on low-magnification images derived from microscopic images of multiple cell samples, at least one target of interest is selected based on a predetermined analytical model. Setting target coordinates on at least one object of interest, setting a corresponding region of interest based on the target coordinates, and acquiring a high-magnification image of the region of interest; Outputting the aforementioned high-magnification image, A cell diagnostic support program that performs actions including those mentioned above.

2. The cell diagnostic support program according to claim 1, wherein the low-magnification image is a stitched image of multiple microscope images.

3. The cell diagnostic support program according to claim 1, wherein, in setting the region of interest, the size of the region of interest is determined based on the size of at least one object of interest.

4. The aforementioned objects of interest are cells or groups of cells suspected of being lesions, The size of the at least one region of interest is determined to include cells other than the at least one object of interest, The cell diagnostic support program according to claim 1.

5. The aforementioned information processing device, The cell diagnostic support program according to claim 1, which performs an operation to acquire multiple images with different depths of focus for at least one region of interest.

6. The aforementioned information processing device, A cell diagnostic support program according to any one of claims 1 to 5, which causes the program to output an image in which the high-magnification image and the low-magnification image are presented side by side.

7. Based on low-magnification images derived from microscopic images of multiple cell samples, at least one target of interest is selected based on a predetermined analytical model. Setting target coordinates on at least one object of interest, setting a corresponding region of interest based on the target coordinates, and acquiring a high-magnification image of the region of interest; Outputting the aforementioned high-magnification image, A method for supporting cell diagnostics, including the above.

8. A diagnostic support system comprising an imaging unit that acquires microscopic images of multiple cell samples, and a control unit, wherein the control unit, A low-magnification image is generated based on the microscopic images of the aforementioned multiple cell samples. From the aforementioned low-magnification images, at least one object of interest is selected based on a predetermined analysis model. A target coordinate is set on at least one object of interest, a corresponding region of interest is set based on the target coordinate, and a high-magnification image of the region of interest is acquired. The acquired high-magnification image is output. A cell diagnostic support system that includes the following.

Images