KR-20260065647-A - MEDICAL IMAGE PROCESSING device, MEDICAL IMAGE PROCESSING METHOD anD SPECIMEN CLASSIFYING METHOD

Abstract

According to the embodiments thereof, a medical image processing device, a medical image processing method, and a specimen classification method may include an input unit for receiving any medical image, a separation unit for separating syphilis bacteria from the any medical image using a separation model trained using a first image containing syphilis bacteria as training data, and a classification unit for classifying whether the syphilis bacteria separated from the separation unit are positive, weakly positive, or negative using a classification model trained using a second image containing syphilis bacteria annotated as positive, weakly positive, or negative as training data.

Inventors

• 천종기
• 김기나
• 윤석주
• 전상률

Assignees

• (재)씨젠의료재단
• 부산대학교 산학협력단

Dates

Publication Date

20260511

Application Date

20241101

Claims (18)

1. Input unit for receiving arbitrary medical images; A separation unit for separating syphilis bacteria from the arbitrary medical image using a separation model trained using a first image containing syphilis bacteria as training data; and A medical image processing device comprising: a classification unit that classifies whether the syphilis bacteria separated in the separation unit are positive, weakly positive, or negative using a classification model trained using a second image containing syphilis bacteria annotated as positive, weakly positive, or negative as training data.
2. In paragraph 1, The above classification unit is a medical image processing device that classifies negative syphilis bacteria among the syphilis bacteria separated from the above separation unit, and then classifies syphilis bacteria other than the negative syphilis bacteria as positive or weakly positive based on brightness information.
3. In paragraph 1, A medical image processing device further comprising a decision unit that determines an arbitrary medical image as positive when the ratio of positive syphilis bacteria among the syphilis bacteria classified by the above classification unit is greater than or equal to a preset ratio.
4. In paragraph 3, The above arbitrary medical image includes a plurality of patches, and The above classification unit classifies the syphilis bacteria contained in each of the plurality of patches as positive, weakly positive, or negative, and A medical image processing device in which the above-determining unit determines any medical image as positive when the ratio of patches containing positive syphilis bacteria among the plurality of patches is greater than or equal to the above-determined ratio.
5. In paragraph 1, The above-mentioned first image is a medical image processing device comprising a first stained image containing stained syphilis bacteria and a second stained image in which syphilis bacteria are separated and displayed.
6. In paragraph 1, The above separation unit is a medical image processing device that outputs an image in which only the syphilis bacteria are expressed in the above arbitrary medical image.
7. In paragraph 1, A medical image processing device having different brightness values for syphilis bacteria annotated as positive, weakly positive, or negative included in the second image above.
8. Input step for receiving an arbitrary medical image; A separation step of separating syphilis bacteria from an arbitrary medical image using a separation model that has been trained using a first image containing syphilis bacteria as training data; and A medical image processing method comprising: a classification step of classifying syphilis bacteria separated from the separation unit as positive, weakly positive, or negative using a classification model that has been trained using a second image containing syphilis bacteria annotated as positive, weakly positive, or negative in the classification unit as training data.
9. In Paragraph 9, A medical image processing method in which, in the above classification step, the classification unit classifies negative syphilis bacteria among the syphilis bacteria separated in the above separation step, and then classifies syphilis bacteria other than the negative syphilis bacteria as positive or weakly positive based on brightness information.
10. In paragraph 8, A medical image processing method further comprising a decision step in which, if the ratio of positive syphilis bacteria among the syphilis bacteria classified in the above classification step is greater than or equal to a preset ratio, a decision unit determines the arbitrary medical image as positive.
11. In Paragraph 10, The above arbitrary medical image includes a plurality of patches, and In the above classification step, the classification unit classifies the syphilis bacteria contained in each of the plurality of patches as positive, weakly positive, or negative, and A medical image processing method in which, in the above decision step, the decision unit determines the arbitrary medical image as positive when the ratio of patches containing positive syphilis bacteria among the plurality of patches is greater than or equal to the preset ratio.
12. In paragraph 8, A medical image processing method comprising a first stained image containing stained syphilis bacteria and a second stained image in which syphilis bacteria are separated and displayed.
13. In paragraph 8, A medical image processing method in which, in the above separation step, the separation unit outputs an image in which only the syphilis bacteria are expressed in the above arbitrary medical image.
14. In paragraph 8, A medical image processing method in which syphilis bacteria annotated as positive, weakly positive, or negative included in the second image above have different brightness values.
15. Step of preparing a sample from a specimen; A step of generating a digital image of the above sample; and A step of isolating syphilis bacteria from the digital image using a trained machine learning model, and classifying the isolated syphilis bacteria as positive, weakly positive, or negative; A sample classification method including
16. In paragraph 15, The machine learning model described above includes a first model for isolating syphilis bacteria from the digital image and a second model for classifying the isolated syphilis bacteria as positive, weakly positive, or negative. A sample classification method in which the first model is trained using a first image containing syphilis bacteria as training data, and the second model is trained using a second image containing syphilis bacteria annotated as positive, weakly positive, or negative as training data.
17. In paragraph 15, A sample classification method comprising the step of preparing a sample from the above specimen, which includes the step of binding an antibody to the syphilis bacteria using an antigen-antibody reaction.
18. A memory storing an arbitrary medical image, a separation model, and a classification model, wherein the separation model is a separation model that performs artificial neural network-based learning to separate syphilis bacteria from a medical image using a first image containing syphilis bacteria as training data, and the classification model is a classification model that performs artificial neural network-based learning to classify syphilis bacteria as positive, weakly positive, or negative using a second image containing syphilis bacteria annotated as positive, weakly positive, or negative as training data; and A computer device comprising a processor that, when a task to classify syphilis bacteria from an arbitrary medical image as positive, weakly positive, or negative is requested, executes a separation model based on artificial neural network learning stored in memory to separate syphilis bacteria from the arbitrary medical image, and executes a classification model based on artificial neural network learning stored in memory to classify syphilis bacteria as positive, weakly positive, or negative.

Description

Medical image processing device, medical image processing method and specimen classification method The present invention relates to a medical image processing device, a medical image processing method, and a specimen classification method. With the advancement of artificial intelligence technology, there is a growing number of attempts to apply machine learning to the medical field. In particular, the field of machine learning for recognizing and classifying medical imaging data is widely used in pathology diagnosis to determine the presence or absence of disease in patients. However, determining the presence or absence of disease in a patient using machine learning models requires a large amount of refined data. In other words, for a machine learning model to accurately determine the presence of disease, it is crucial to train the model sophisticatedly; however, such sophisticated training requires a large volume of well-classified or refined, high-quality data. This becomes a factor that hinders the efficiency of the disease determination process. FIG. 1 is a drawing showing an example of a medical image processing device according to embodiments of the present disclosure. FIG. 2 is a diagram illustrating data input to a medical image processing device according to embodiments of the present disclosure. FIG. 3 is a diagram showing an example of a learning process of a medical image processing device according to embodiments of the present disclosure. FIG. 4 is a drawing showing an example of a medical image processing device in operation according to embodiments of the present disclosure. FIG. 5 is a diagram showing an example of a learning process of a medical image processing device according to embodiments of the present disclosure. FIG. 6 is a drawing showing an example of a medical image processing device in operation according to embodiments of the present disclosure. FIG. 7 is a drawing showing an example of brightness information used in the operation of a medical image processing device according to embodiments of the present disclosure. FIG. 8 is a drawing showing another example of a medical image processing device according to embodiments of the present disclosure. FIG. 9 is a drawing showing an example of a medical image processing device in operation according to embodiments of the present disclosure. FIG. 10 is a diagram illustrating an example of a medical image processing method according to embodiments of the present disclosure. FIG. 11 is a drawing showing another example of a medical image processing method according to embodiments of the present disclosure. FIG. 12 is a diagram showing an example of a sample classification method according to embodiments of the present disclosure. FIG. 13 is a configuration diagram of a computing system according to embodiments of the present disclosure. FIG. 14 is a configuration diagram of a client-server computer system according to embodiments of the present disclosure. Hereinafter, some embodiments of the present disclosure will be described in detail with reference to the exemplary drawings. In assigning reference numerals to the components of each drawing, the same components may have the same reference numeral as much as possible, even if they are shown in different drawings. Furthermore, in describing the embodiments, if it is determined that a detailed description of related known components or functions may obscure the essence of the technical concept, such detailed description may be omitted. Where terms such as "comprising," "having," or "consisting of" are used in this specification, other parts may be added unless "only" is used. Where a component is expressed in the singular, it may include a plural unless otherwise specified. Additionally, terms such as first, second, A, B, (a), (b), etc., may be used to describe the components of the present disclosure. These terms are used merely to distinguish the components from other components, and the nature, order, sequence, or number of the components are not limited by such terms. In describing the positional relationship of components, where it is stated that two or more components are "connected," "combined," or "joined," it should be understood that while the two or more components may be directly "connected," "combined," or "joined," they may also be "connected," "combined," or "joined" with other components "intervened." Here, the other components may be included in one or more of the two or more components that are "connected," "combined," or "joined" with one another. In describing the temporal flow relationship regarding components, methods of operation, or methods of production, for example, when the temporal or sequential relationship is described using "after," "following," "next," or "before," it may include cases where the relationship is not continuous unless "immediately" or "directly" is used. Meanwhile, where numerical values or corresponding information regarding a compone