WO2024181845A1 - Method and system for analyzing vaginal discharge of women of childbearing age and pregnant women - Google Patents

Abstract

The present invention relates to a method and system for analyzing vaginal discharge of women of childbearing age and pregnant women. According to the present invention, the vaginal discharge analysis system for women of childbearing age and pregnant women includes: an input unit for receiving basic information, symptom information, and image data of vaginal discharge from the subject according to a preset sequence; a preprocessing unit for preprocessing the data input by the input unit; an analysis unit that uses a pre-trained deep learning model to analyze the preprocessed data from the preprocessing unit and classify the color, amount, and viscosity of the vaginal discharge; and an output unit for outputting the analysis results of the vaginal discharge analyzed by the analysis unit.

Description

Method and system for analyzing vaginal secretions of women of childbearing age and pregnant women

The present invention relates to a method and system for analyzing vaginal secretions of women of childbearing age and pregnant women, and more specifically, to a method and system for analyzing vaginal secretions of women of childbearing age and pregnant women, including adolescents, which can provide information for appropriate health management according to the pathological condition of the vaginal secretions by analyzing the color, viscosity, amount, etc. of the vaginal secretions of women of childbearing age and pregnant women.

Vaginal discharge in women is a normal physiological phenomenon that plays an important role in preventing damage to the skin of the vagina and vulva due to external friction and maintaining an acidic environment inside the vagina to prevent the growth of external pathogens.

On the other hand, when vaginitis occurs, vaginal discharge differs from normal vaginal discharge in terms of amount, color, viscosity, odor, etc., so the state of the subject's vaginal health can be determined based on the state of the vaginal discharge.

Among these, in the case of red-colored vaginal discharge, bleeding findings can be suspected, and in the case of women of childbearing age, abnormal bleeding or normal menstruation can be determined based on the menstrual cycle, and in the case of pregnant women, the health status of the mother can be determined based on the pattern of bleeding before and after childbirth. Before childbirth, even if the bleeding is small, the need for treatment is strongly recommended, and after childbirth, it is possible to distinguish between normal and abnormal lochia based on the amount and color of bleeding.

However, in the case of the general public who lack medical knowledge, it is difficult to properly assess the condition of vaginal discharge in women of childbearing age or adolescents with poor access to medical care, so they often neglect it when it is severe or incur unnecessary expenses by going to the hospital even when there is no abnormality. In particular, in the case of mothers living in areas vulnerable to childbirth, there are cases where abnormal findings of vaginal discharge or bleeding are not properly screened, and the condition progresses to a severe state.

In addition, it is difficult to accurately determine the health of the vagina based on vaginal discharge because the normal vaginal discharge of a woman of childbearing age is different from that of a pregnant woman.

Therefore, before visiting a hospital and receiving specialist treatment, a tool is needed that can help women of childbearing age, including adolescents, postpartum and postpartum women, and pregnant women to easily analyze the condition of vaginal discharge and screen their own vaginal health, thereby enabling appropriate early treatment.

Thus, according to the present invention, the purpose is to provide a method and system for analyzing vaginal secretions of women of childbearing age and pregnant women, including adolescents, which can provide information for appropriate vaginal health management according to the pathological state of the vaginal secretions through analysis of the color, viscosity, amount, etc. of the vaginal secretions of women of childbearing age and pregnant women.

According to an embodiment of the present invention for achieving such technical tasks, a system for analyzing vaginal secretions of women of childbearing age and pregnant women includes an input unit for receiving basic information, symptom information, and image data of vaginal secretions of a subject according to a preset order, a preprocessing unit for preprocessing the data input by the input unit, an analysis unit for analyzing the data preprocessed in the preprocessing unit using a pre-learned deep learning model to classify the color, amount, and viscosity of the vaginal secretions, and an output unit for outputting the analysis results of the vaginal secretions analyzed by the analysis unit.

In addition, the method for analyzing vaginal discharge of women of childbearing age and pregnant women includes a step of receiving basic information, symptom information, and image data of vaginal discharge of the subject according to a preset order, a step of preprocessing the data input by the step of receiving the image data, a step of analyzing the preprocessed data by the step of preprocessing using a pre-trained deep learning model to classify the color, amount, and viscosity of the vaginal discharge, and a step of outputting the results analyzed by the step of analyzing the data.

According to the present invention, by providing feedback on mild vaginitis treatable through improvement of lifestyle habits and vaginitis requiring treatment to women of childbearing age and pregnant women who frequently have increased vaginal discharge and require improvement through management of vaginal discharge, it is possible to effectively manage the management of the subjects' vaginal discharge and thereby help promote women's health.

In particular, in the case of postpartum lochia of the mother among the vaginal discharge, it is possible to determine the condition after childbirth through input information such as the date of delivery, and by selecting the postpartum lochia of the mother after childbirth based on the pattern of postpartum bleeding that occurs at this time, it is possible to track the health condition of the mother.

That is, before childbirth, it is strongly recommended to seek medical attention even if the bleeding is small, and after childbirth, it is possible to distinguish between normal and abnormal lochia based on the amount and color of bleeding.

FIG. 1 is a diagram illustrating the configuration of a system for analyzing vaginal secretions of women of childbearing age and pregnant women according to one embodiment of the present invention.

Figure 2 is a flow chart of a method for analyzing vaginal secretions of women of childbearing age and pregnant women according to one embodiment of the present invention.

Figure 3 is a flowchart of a process in which an input unit receives basic information of a subject according to one embodiment of the present invention.

Hereinafter, a preferred embodiment of the present invention will be described in detail with reference to the attached drawings. In this process, the thickness of lines and the size of components illustrated in the drawings may be exaggerated for the sake of clarity and convenience of explanation.

In addition, the terms described below are terms defined in consideration of their functions in the present invention, and may vary depending on the intent or custom of the subject or operator. Therefore, the definitions of these terms should be made based on the contents throughout this specification.

FIG. 1 is a diagram illustrating the configuration of a system for analyzing vaginal secretions of women of childbearing age and pregnant women according to one embodiment of the present invention.

As illustrated in Figure 1, the vaginal secretion analysis system (100) of women of childbearing age and pregnant women includes an input unit (110), a preprocessing unit (120), an analysis unit (130), and an output unit (140).

The input unit (110) receives the subject's basic information, symptom information, and image data of vaginal discharge according to a preset order.

At this time, the basic information of the subject being entered includes the subject's identification information (name, ID, etc.), date of birth, and pregnancy status.

In addition, the input unit (110) can additionally receive basic information of the subject depending on whether the subject is pregnant, and can classify the subject based on the received basic information.

At this time, if the subject is pregnant, the input unit (110) additionally receives input from the subject regarding the number of weeks of pregnancy, whether or not she has undergone a cervical cancer test, and whether or not she has been vaccinated against cervical cancer.

In addition, the subject can input the expected date of delivery or the subject's last menstruation, or the date of delivery if the subject has already given birth, instead of the number of weeks of pregnancy, and the input unit (110) can automatically calculate the expected date of delivery or the number of days elapsed after giving birth based on the input last menstrual date.

On the other hand, if the subject is not pregnant, the input unit (110) additionally receives input from the subject regarding menstrual history, whether or not sexual intercourse occurred immediately before the examination, whether or not a cervical cancer test was performed, whether or not a cervical cancer vaccination was administered, and whether or not bleeding other than menstrual bleeding occurred.

At this time, menstrual history may include whether the subject has had menstruation, the menstrual cycle, and the regularity of the menstrual cycle.

Additionally, the subject's symptom information includes whether the subject's vaginal discharge has an odor and whether the subject experiences itching of the vulva.

At this time, if there is an odor in the vaginal discharge, the type of odor can be classified into preset types (e.g., fishy odor, yeast odor, rotten meat odor), and the subject can input which category of odor the odor of the vaginal discharge is closest to.

In addition, the image data of the vaginal discharge may be an image taken of the corresponding area of the panty liner stained with the subject's vaginal discharge, and at this time, a panty liner of a preset color (e.g., white) may be used to facilitate identification of the color, amount, and viscosity of the vaginal discharge through the image.

The preprocessing unit (120) preprocesses data received through the input unit (110).

At this time, the preprocessing unit (120) performs preprocessing such as image resizing, image pixel normalization, and white balance correction (based on the color of the panty liner) on the image data of the input vaginal discharge.

The analysis unit (130) analyzes the data preprocessed in the preprocessing unit (120) using a pre-learned deep learning model.

At this time, the analysis unit (130) analyzes the image data of the vaginal discharge using a deep learning model to classify the color, amount, and viscosity of the vaginal discharge.

For example, the deep learning model is trained to classify the color of vaginal discharge into a predetermined number of colors (e.g., clear, white, yellow, green, pink, scarlet, and brown), the amount of vaginal discharge into a plurality of levels (e.g., levels 0 to 4), and the viscosity of vaginal discharge into a plurality of levels (e.g., levels 0 to 2).

At this time, if the color of the vaginal discharge is classified as a color that includes red colors such as pink, scarlet, or brown, the subject is judged to be in a state where there is a possibility of bleeding.

In addition, the analysis unit (130) can analyze the image classification result classified by the analysis unit (130) and the symptom information of the subject input by the input unit (110) according to the classification result of the subject classified by the input unit (110) to determine the quality of health of the subject.

The output unit (140) outputs the analysis results of the vaginal secretions analyzed by the analysis unit (130).

For example, if the analysis result from the analysis unit (130) is that the color of the vaginal discharge is yellow, the amount is at level 2, and the viscosity is at level 1, the output unit (140) outputs “the vaginal discharge is yellow, the amount is appropriate, and it has the appearance of a sticky runny nose.”

In addition, the output unit (140) can output the subject's vaginal health status determined by the analysis unit (130) and an appropriate response based on the subject's vaginal health status.

At this time, the subject's quality of health can be classified into a preset number of stages (e.g., stage 1 to stage 5) according to preset conditions.

Below, a method for analyzing vaginal secretions of women of childbearing age and pregnant women is described with reference to Fig. 2.

Figure 2 is a flow chart of a method for analyzing vaginal secretions of women of childbearing age and pregnant women according to one embodiment of the present invention.

The input unit (110) receives the subject's basic information, symptom information, and image data of vaginal discharge according to a preset order (S210).

At this time, the basic information of the subject being input may include the subject's identification information, date of birth, and pregnancy status, and depending on whether the subject is pregnant, additional basic information of the subject may be input, and the subject may be classified based on the input basic information.

At this time, the order in which the subject's basic information is entered is specifically explained with reference to Figure 3.

Figure 3 is a flowchart of a process in which an input unit receives basic information of a subject according to one embodiment of the present invention.

First, the input unit (110) receives the subject's identification information and the subject's date of birth (S211).

At this time, the subject's identification information may be the subject's real name or a nickname and ID arbitrarily set by the subject.

After entering the subject's identification information and date of birth, the input unit (110) receives input on whether the subject is pregnant (S212).

At this time, if the subject has entered a pregnancy status, the input unit (110) receives the subject's pregnancy weeks (S213).

At this time, the input unit (110) can receive the expected date of birth or the date of the subject's last menstrual period instead of the number of weeks of pregnancy.

Additionally, the input unit (110) can receive the date of delivery in case of delivery.

Additionally, the input unit (110) can automatically calculate the predicted date of delivery or the number of days elapsed after delivery based on the input last menstrual date.

After entering the number of weeks of pregnancy, the input unit (110) can enter whether a cervical cancer test has been performed and, if so, whether cervical cancer has been detected (S214).

After receiving input on whether or not to have a cervical cancer test and whether or not to have cervical cancer, the input unit (110) receives input on whether or not to have been vaccinated against cervical cancer (S215).

On the other hand, if the subject enters that she is not pregnant, the input unit (110) first receives input on whether the subject has had her first period, her menstrual cycle, and whether the menstrual cycle is regular (S216).

Next, the input unit (110) receives input on whether the subject has had sexual intercourse immediately before the test (S217).

After inputting whether or not you have had sexual intercourse, the input unit (110) can input whether or not you have had a cervical cancer test and, if so, whether or not you have had cervical cancer (S218).

After receiving input on whether or not to have a cervical cancer test and whether or not to have cervical cancer, the input unit (110) receives input on whether or not to have been vaccinated against cervical cancer (S219).

In addition, after receiving basic information of the subject according to the process in Fig. 3, the input unit (110) can classify the subject according to the result of the input.

After the subject's basic information is entered, the input unit (110) receives the subject's symptom information and image data of vaginal discharge.

At this time, symptom information includes whether the subject's vaginal discharge smells and whether the subject experiences itching of the vulva.

At this time, if there is an odor in the vaginal discharge, the type of odor can be classified into preset types, and the subject can input which category of odor the odor of the vaginal discharge feels is closest to.

In addition, the image data of the vaginal discharge may be an image taken of the corresponding area of the panty liner stained with the subject's vaginal discharge, and at this time, a panty liner of a preset color (e.g., white) may be used to facilitate identification of the color, amount, and viscosity of the vaginal discharge through the image, or an image may be used that has undergone preprocessing, such as white balance correction, based on the color of the panty liner.

Next, data input through the input unit (110) is preprocessed through the preprocessing unit (120) (S220).

After preprocessing is performed, the analysis unit (130) analyzes the data preprocessed in the preprocessing unit (120) using the pre-learned deep learning model (S230).

At this time, the deep learning model can be trained in the following order.

First, collect data to train the model.

At this time, the data collected refers to a set of image data of vaginal discharge image data for a preset quantity and an image data set after a preset time has elapsed from the vaginal discharge image data.

Additionally, the image data is a file in which a rectangle is cropped on a panty liner and the color, amount, and viscosity of each image are labeled.

After collecting data, data analysis is conducted using the Exploratory Data Analysis (EDA) technique, which is an analysis method that flexibly explores the raw data and creates a statistical model based on information obtained from the characteristics and structure of the data.

Next, the analyzed data is preprocessed.

More specifically, the image data is resized to fit the system, the pixels of the image data are normalized, and the white balance is corrected based on the color of the panty liner.

Additionally, during the preprocessing process, data can be increased by flipping the image data up/down, left/right, rotating it, etc., as needed.

Finally, the vaginal discharge data is trained through multi-classification and the model performance is evaluated through the F1 score.

Additionally, the analysis unit (130) analyzes image data of vaginal discharge to classify the color, amount, and viscosity of the vaginal discharge.

For example, the deep learning model is trained to classify the color of vaginal discharge into a predetermined number of colors (e.g., clear, white, yellow, green, pink, scarlet, and brown), the amount of vaginal discharge into a plurality of levels (e.g., levels 0 to 4), and the viscosity of vaginal discharge into a plurality of levels (e.g., levels 0 to 2).

At this time, if the color of the vaginal discharge is classified as a color that includes red colors such as pink, scarlet, or brown, the subject is judged to be in a state where there is a possibility of bleeding.

In addition, the analysis unit (130) can analyze the image classification result classified by the analysis unit (130) and the symptom information of the subject input by the input unit (110) according to the classification result of the subject classified by the input unit (110) to determine the quality of health of the subject.

After completing data analysis using the learned deep learning model, the analysis results are output (S240).

For example, if the analysis result from the analysis unit (130) is that the color of the vaginal discharge is yellow, the amount is at level 2, and the viscosity is at level 1, the output unit (140) outputs “the vaginal discharge is yellow, the amount is appropriate, and it has the appearance of a sticky runny nose.”

In addition, the output unit (140) can output the subject's vaginal health status determined by the analysis unit (130) and an appropriate response based on the subject's vaginal health status.

In this way, according to the present invention, by providing feedback on mild vaginitis that can be treated through improvement of lifestyle habits and vaginitis that requires treatment to women of childbearing age and pregnant women who frequently have increased vaginal discharge and need improvement through vaginal discharge management, it is possible to effectively manage the vaginal discharge of the subjects and thereby help promote women's health.

In addition, in the case of mothers, by identifying the degree of infection or bleeding in postpartum lochia, one of the vaginal discharges before and after childbirth, it is possible to effectively manage the subject's postpartum lochia, thereby helping to promote women's postpartum health.

Although the present invention has been described with reference to the embodiments shown in the drawings, these are merely exemplary, and those skilled in the art will understand that various modifications and equivalent other embodiments are possible from this. Therefore, the true technical protection scope of the present invention should be determined by the technical idea of the following patent claims.

Claims (8)

An input section for receiving the subject's basic information, symptom information, and image data of vaginal discharge according to a preset order; A preprocessing unit that preprocesses data entered by the above input unit; An analysis unit that analyzes the data preprocessed in the above preprocessing unit using a pre-learned deep learning model to classify the color, amount, and viscosity of vaginal discharge; and A system for analyzing vaginal secretions of women of childbearing age and pregnant women, comprising an output unit for outputting the analysis results of vaginal secretions analyzed by the above analysis unit. In paragraph 1, The above basic information is, A system for analyzing vaginal secretions of women of childbearing age and pregnant women, including the subject's identification information, date of birth, and pregnancy status. In the second paragraph, The above input section, A vaginal secretion analysis system for women of childbearing age and pregnant women that additionally inputs information about the subject based on whether the subject is pregnant or has given birth, and classifies the subject based on the basic information input. In paragraph 1, The above symptom information is, A system for analyzing vaginal discharge of women of childbearing age and pregnant women, including the presence of odor in the subject's vaginal discharge and itching of the vulva. In paragraph 4, A vaginal secretion analysis system for women of childbearing age and pregnant women, which inputs the odor of the vaginal secretion as a corresponding odor type among preset odor types when the above vaginal secretion has an odor. In paragraph 1, The above analysis section, A vaginal secretion analysis system for women of childbearing age and pregnant women, which analyzes the image classification results and the symptom information of the subject entered in the input section by classification group of the subject according to basic information in the input section above to determine the vaginal health status of the subject. In paragraph 1, The above output section, A vaginal secretion analysis system for women of childbearing age and pregnant women that outputs the subject's vaginal health status determined by the above analysis unit and a predefined response plan based on the subject's vaginal health status. A step of receiving the subject's basic information, symptom information, and image data of vaginal discharge according to a preset order; A step of preprocessing data input by the step of inputting the image data; A step of analyzing the preprocessed data in the preprocessing step using the pre-learned deep learning model to classify the color, amount, and viscosity of the vaginal discharge; and A method for analyzing vaginal secretions of women of childbearing age and pregnant women, comprising a step of outputting the analysis results of vaginal secretions analyzed by the step of analyzing the above data.

Images