KR102410545B1 - High-resolution skin clinical image management device - Google Patents

Abstract

The present invention relates to an apparatus for managing high-resolution skin clinical images, and more particularly, by improving the inconvenience of specifying a lesion site in a skin clinical image and taking and providing dermascopies every day, all microscopic images in the skin clinical image It simplifies the skin diagnosis process in which the conventional reading and imaging are separated by automatically generating, storing and managing high-resolution synthetic dermascopy images for the lesion site, and the patient according to the automatic transmission of skin medical images through digitalization of the skin images It relates to a high-resolution skin clinical image management device to enable the construction of separate big data and efficient history management of big data.

Description

High-resolution skin clinical image management device

The present invention relates to an apparatus for managing high-resolution skin clinical images, and more particularly, by improving the inconvenience of specifying a lesion site in a skin clinical image and taking and providing dermascopies every day, all microscopic images in the skin clinical image It simplifies the skin diagnosis process in which the conventional reading and imaging are separated by automatically generating, storing and managing high-resolution synthetic dermascopy images for the lesion site, and the patient according to the automatic transmission of skin medical images through digitalization of the skin images It relates to a high-resolution skin clinical image management device to enable the construction of separate big data and efficient history management of big data.

In recent clinical trials, the importance of using image data to evaluate the efficacy of new drugs is increasingly emphasized.

In particular, as the development of targeted anticancer drugs and immunotherapeutic agents is accelerated, drug efficacy evaluation using functional and molecular imaging biomarkers in addition to the existing drug efficacy evaluation based on tumor size has become very useful and important.

In addition, as the number of these imaging biomarkers increases and the amount of imaging data increases, the need for clinical trial medical image management in clinical trials or preclinical trials is increasing.

As described above, although the importance of image data in clinical trials is increasing, it is not possible to provide a system capable of integrating and managing image data and numerical data during clinical data processing.

Therefore, conventionally, video data is uploaded to a separate web page of the server or delivered by courier using CD/DVD. There were problems in that it was difficult to transmit and manage image data, and the possibility of loss and loss of image data was high.

In addition, there is a problem in that it is possible to read images (central review) only at a designated time and place, and there is a problem in that it is difficult to store image data obtained in a clinical trial.

That is, the traditional clinical method for skin diagnosis has a procedure in which a doctor reads a dermascope picture of a skin lesion site taken by a nurse, and if the skin lesion area is widely spread or has multiple skin diseases, the nurse There was a problem in that it was difficult to specify the lesion site for dermascope photography, so the diagnosis process was inefficiently lengthened.

Therefore, it is necessary to develop a technology to shorten the time taken for skin diagnosis, reduce unnecessary labor of nurses and doctors, and provide prompt and intelligent medical services to patients.

In particular, since the traditional clinical method for skin diagnosis is a procedure in which a doctor reads a dermoscopy picture of a skin lesion taken by a nurse, the clinical picture is used as a reference to record the location of the skin lesion, etc. was only used as

However, when the skin lesion site is widely spread or has multiple skin diseases, it is difficult for the nurse to specify the lesion site for dermoscopy, so the diagnosis process becomes inefficiently long.

Therefore, it is necessary to develop a technology for reducing the time required for skin diagnosis, reducing unnecessary labor of nurses and doctors, and providing prompt and intelligent medical services to patients.

On the other hand, in order to obtain clinical data, it is photographed through a DSR camera. In this case, the camera itself is heavy, and it was possible only by using a separate external lighting means.

That is, the inconvenience of portability, the operation problem due to the provision of a separate external lighting means, and the economic burden due to the provision of expensive equipment had to be provided.

Accordingly, there is a need for a new type of clinical imaging device that can be used more conveniently by users in consideration of portability and ease of use of the clinical imaging device.

In addition, there is a need for a technology capable of transmitting, storing and managing clinical images taken from a clinical imaging device to external devices in real time by enabling interworking with other devices.

Republic of Korea Patent Publication No. 10-2019-0118358

Therefore, the present invention has been proposed in view of the problems of the prior art as described above, and the first object of the present invention is to improve the inconvenience of specifying a lesion site in a skin clinical image and taking and providing dermascopies every day. Therefore, it is possible to automatically create, store, and manage high-resolution synthetic dermacoscopy images for all microlesion sites in skin clinical images.

A second object of the present invention is to simplify the skin diagnosis process in which the conventional reading and imaging are separated, and to establish big data for each patient and efficiently manage the history of big data according to the automatic transmission of skin medical images through digitization of skin images. is to do it.

A third object of the present invention is to provide a photographing device capable of selectively performing large-area photographing or local area photographing by enabling attachment and detachment between the first lighting case and the photographing device case.

A fourth object of the present invention is to provide a clinical camera capable of managing images taken using a smart device by mounting a smart device inside the imaging device case.

In order to achieve the object to be solved by the present invention, a high-resolution skin clinical image management apparatus according to an embodiment of the present invention,

The skin image management server acquires the lesion image present in the skin clinical image photographed from the camera 2000, generates a high-resolution lesion image for the lesion image using the high-resolution generation algorithm, and the dermascopy (3000) Acquire an actual high-resolution lesion image, but determine the authenticity of the generated high-resolution lesion image and the actual high-resolution lesion image using the authenticity classification algorithm. An artificial intelligence learning model generating means (6000) for generating

A lesion site extraction means 7000 for extracting a lesion site in a skin clinical image taken by a camera 2000 to generate a high-resolution synthetic dermatophyte image;

Using the generated high-resolution synthetic dermatophyte image generation learning model, high-resolution synthetic dermatography image is generated by performing high-resolution on the lesion site image extracted by the lesion site extraction means, and provided to the skin image management server 4000 It includes a high-resolution performing means (8000) for doing.

Through the high-resolution skin clinical image management device of the present invention, high-resolution synthesis of all micro-lesion sites in the skin clinical image is improved by specifying the lesion site in the skin clinical image, and improving the inconvenience of taking and providing dermascopies every day. The effect of simplifying the skin diagnosis procedure by providing a tweezers data acquisition method that actually takes a dermoscopy picture of a specific lesion site after reviewing a high-resolution synthetic dermatography image by automatically creating and managing the dermatoscope image will perform

In addition, it simplifies the skin diagnosis process in which the conventional reading and imaging are separated, and provides the effect of constructing big data for each patient and efficiently managing the history of big data according to the automatic transmission of skin medical images through digitalization of skin images. .

In addition, when skin medical image big data is secured with clinical images and synthetic dermacoscopy images for skin diseases, it is possible to provide scalability that can increase the accuracy of remote treatment for skin diseases through personal smart devices.

In addition, through the camera, it is possible to attach and detach between the first lighting case and the photographing device case so that large-area photographing or local area photographing can be selectively performed. It has the effect of performing large-area or local area imaging as needed.

In addition, by mounting a smart device inside the shooting device case of the camera to manage the images captured using the smart device, and automatically uploading the captured video image to the cloud or data storage to store and manage, It will provide the efficiency and scalability to check the clinical information of the subject to be observed anytime, anywhere.

1 is a conceptual block diagram of an apparatus for managing high-resolution skin clinical images according to an embodiment of the present invention.
2 is an exemplary diagram of an artificial intelligence reading model through high resolution of a large-area skin image of the apparatus for managing high-resolution skin clinical images according to an embodiment of the present invention.
3 is a block diagram of an artificial intelligence learning model generating means 6000 of a high-resolution skin clinical image management apparatus according to an embodiment of the present invention.
4 is an exemplary view of high-resolution and dermacopy composite photos of the apparatus for managing high-resolution skin clinical images according to an embodiment of the present invention.
5 is an exemplary diagram of a generator and authenticity separator of an apparatus for managing high-resolution skin clinical images according to an embodiment of the present invention.
6 is a schematic configuration diagram of an entire system to which a high-resolution skin clinical image management apparatus according to an embodiment of the present invention is applied.
7 is a block diagram of a skin image management server 4000 according to an embodiment of the present invention.
8 is an exemplary view of a fax page 4510 provided by the skin image management server 4000 in the entire system to which the high-resolution skin clinical image management apparatus according to an embodiment of the present invention is applied.
9 is an exemplary diagram of a body map provided by the skin image management server 4000 in the entire system to which the high-resolution skin clinical image management apparatus according to an embodiment of the present invention is applied.
10 is a perspective view of the camera 2000 within the entire system to which the apparatus for managing high-resolution skin clinical images according to an embodiment of the present invention is applied, FIG. 11 is a rear perspective view, FIG. 12 is a rear view, and FIG. 13 is a front view 14 is a side view, FIG. 15 is an exploded perspective view, FIG. 16 is an exemplary view showing a photographing device, and FIG. 17 is an exemplary photograph.

The following is merely illustrative of the principles of the invention. Therefore, those skilled in the art will be able to devise various devices that, although not explicitly described or shown herein, embody the principles of the present invention and are included within the spirit and scope of the present invention.

In addition, all conditional terms and examples listed herein are, in principle, expressly intended only for the purpose of understanding the concept of the present invention, and it should be understood that they are not limited to the specifically enumerated embodiments and states as such. do.

A high-resolution skin clinical image management apparatus according to an embodiment of the present invention,

The skin image management server acquires the lesion image present in the skin clinical image photographed from the camera 2000, generates a high-resolution lesion image for the lesion image using the high-resolution generation algorithm, and the dermascopy (3000) Acquire an actual high-resolution lesion image, but determine the authenticity of the generated high-resolution lesion image and the actual high-resolution lesion image using the authenticity classification algorithm. An artificial intelligence learning model generating means (6000) for generating

A lesion site extraction means 7000 for extracting a lesion site in a skin clinical image taken by a camera 2000 to generate a high-resolution synthetic dermatophyte image;

Using the generated high-resolution synthetic dermatophyte image generation learning model, high-resolution synthetic dermatography image is generated by performing high-resolution on the lesion site image extracted by the lesion site extraction means, and provided to the skin image management server 4000 It is characterized in that it is configured to include a high-resolution performing means (8000) for.

Hereinafter, an embodiment of the high-resolution skin clinical image management apparatus according to the present invention will be described in detail with reference to the drawings.

1 is a conceptual block diagram of an apparatus for managing high-resolution skin clinical images according to an embodiment of the present invention.

As shown in Fig. 1, the present invention's high-resolution skin clinical image management apparatus includes an artificial intelligence learning model generating means 6000, a lesion site extraction means 7000, and a high-resolution performing means 8000. do.

That is, the high-resolution lesion for the lesion image is obtained by using the high-resolution generation algorithm by acquiring the lesion image present in the skin clinical image photographed from the camera 2000 in the skin image management server through the artificial intelligence learning model generating means 6000. An image 10 is created.

At this time, in order to perform the learning, an actual high-resolution lesion image taken from the dermacopy 3000 is acquired, and the generated high-resolution lesion image and the acquired real high-resolution lesion image are used to determine the authenticity of the lesion using an authenticity discrimination algorithm. Thus, when the accuracy is higher than the reference accuracy, the learning is terminated to generate a high-resolution synthetic dermacopy image generation learning model.

Specifically, as shown in FIG. 2, the artificial intelligence learning model generating means 6000 is,

a clinical skin image acquisition unit 6100 for acquiring an image of a lesion existing in a skin clinical image taken from the camera 2000;

a high-resolution lesion image generator 6200 for receiving the lesion image present in the acquired skin clinical image and generating a high-resolution lesion image for the lesion image using a high-resolution generation algorithm;

an actual high-resolution lesion image acquisition unit 6300 for acquiring an actual high-resolution lesion image taken from the dermacopy 3000;

Dermasscopy that receives the generated high-resolution lesion image and the actual high-resolution lesion image, determines the authenticity using the authenticity classification algorithm, and terminates the learning when the accuracy is greater than the reference accuracy. It is characterized in that it includes; AI learning model generation unit (6400) for image generation.

That is, the clinical skin image acquisition unit 6100 acquires the lesion image by extracting the lesion image existing in the skin clinical image photographed from the camera 2000 .

As shown in FIG. 2 , images of lesions present in a large-area skin clinical image are acquired.

At this time, the high-resolution lesion image generator 6200 receives the lesion image present in the acquired skin clinical image and uses the high-resolution generation algorithm to generate a high-resolution lesion image for the lesion image. .

That is, it is magnified to the same level as the image 10 taken by dermacopy.

For example, high resolution is performed using a neural network model, and the neural network model includes various models such as Recurrent Neural Networks (RNN), Feedforward Neural Network (FFNN), and Convolutional Neural Network (CNN).

In particular, a low-resolution skin clinical image is generated as a high-resolution image using the GAN model. For this, the GAN model must be trained.

Because learning proceeds with a generator neural network, it is defined as a high-resolution generation algorithm. It is an algorithm based on GAN (Generative Adversarial Networks), and a loss function that has better characteristics than MES (Manufacturing Execution System) is applied.

To be specific, the single image super resolution (SISR) technique that converts the extracted clinical picture of the microlesion into a dermoscopy image with high magnification is the key. It is characterized by constructing a high-resolution generation algorithm to create a low-resolution clinical picture through blurring, down-sampling, and noise addition, and to reconstruct a ground truth image from a low-resolution clinical picture.

In other words, as a traditional high-resolution approach, the interpolation-based method enlarges the image size but fails to reconstruct the fine details.

Since SRCNN was first proposed in 2014, deep learning-based methods have rapidly developed, and the performance of fine-grained reconstruction has been greatly improved.

However, the texture restoration is still not natural, so the PSNR is high, but the human eye's cognitive performance is unsatisfactory.

For natural high-resolution image restoration, a super resolution generative adversarial network (SRGAN), a generative adversarial network (GAN) family, has been proposed.

For example, in the generator, two sub-pixel convolution layers are used to increase the resolution of the low-resolution input image, and a loss function is used. The loss function calculates adversarial loss and content loss. will include

The adversarial loss is used to learn a discriminator network to distinguish between super-resolution images and natural images, and the content loss is used to learn perceptual similarity.

And, the real high-resolution lesion image acquisition unit 6300 performs a function for acquiring the actual high-resolution lesion image taken from the dermacopy (3000).

At this time, the AI learning model generating unit 6400 for generating the dermacopy image receives the generated high-resolution lesion image and the actual high-resolution lesion image, and determines the authenticity using the authenticity classification algorithm. By terminating the training, a high-resolution synthetic dermascopic image generation training model is created.

That is, as shown in FIG. 5 , in the discriminator neural network, the real high-resolution lesion image, which is the original HR image, and the high-resolution lesion image, which is the generated SR image, are compared, and the learning proceeds while determining whether authenticity cannot be distinguished. do.

In training, the loss function is calculated by removing the sigmoid function included at the end of the discriminator neural network, and the final loss value is obtained by adding the adversarial loss value calculated in the discriminator neural network and the content loss value generated in the independent VGG at a certain ratio. is to calculate

At this time, authenticity is determined using a discriminator neural network, which is an algorithm for distinguishing authenticity, and when the accuracy is greater than the reference accuracy, learning is terminated to generate a high-resolution synthetic dermascopy image generation learning model. The accuracy result of the learned model is 85 %, and when the set reference accuracy result value is 80%, since the accuracy result value of the trained model exceeds the set reference accuracy result value, the corresponding learning is terminated to generate a high-resolution synthetic dermacopy image generation learning model.

On the other hand, according to an additional aspect, the present invention is characterized in that it further comprises a polarization processing means (8500) for polarization processing by extracting the high-resolution synthetic dermacopy image generated by the high-resolution performing means.

As shown in FIG. 5, the polarization processing means 8500 is configured to extract the high-resolution synthetic dermacopy image generated by the high-resolution performing means and polarize it.

In this case, it is possible to obtain the same image as the dermascopic image in which spatial distortion does not occur, and at the same time, it is possible to extract the polarized synthetic dermascopic image.

On the other hand, according to another additional aspect, the present invention is characterized in that it further comprises a spatial distortion processing means (9000) for reducing spatial distortion by acquiring a high-resolution synthetic dermascopic image polarized by the polarization processing means. .

That is, since spatial distortion occurs in general dermascopes, a spatial distortion processing means is configured to reduce spatial distortion.

Through this, it is possible to obtain a high-resolution synthetic dermascopic image with reduced spatial distortion.

Then, the lesion site extraction means 7000 extracts the lesion site in the skin clinical image photographed by the camera 2000 in order to generate a high-resolution synthetic dermacopy image.

That is, it is for extracting the lesion site of the patient currently being treated, and by providing it to the high-resolution performing means 8000, it is possible to generate a high-resolution synthetic dermacopy image.

And, the high-resolution performing means 8000 performs high-resolution on the lesion site image extracted by the lesion site extracting means using the generated high-resolution synthetic dermascopy image generation learning model to obtain a high-resolution synthetic dermascopy image. It performs a function to be created and provided to the skin image management server 4000 .

That is, by providing the lesion site image extracted by the lesion site extraction means as an input value to the high-resolution synthetic dermascopy image generation learning model, the high-resolution synthetic dermascopy image is generated and provided to the skin image management server 4000 .

Therefore, the present invention improves the hassle of specifying the lesion site in the skin clinical image and providing daily dermacopy pictures, and automatically generates high-resolution synthetic dermatography images for all microlesion sites in the skin clinical image. By enabling storage and management, the tweezers data acquisition method that actually takes a dermoscopy picture of a specific lesion site after reviewing a high-resolution synthetic dermatography image is provided, thereby simplifying the skin diagnosis procedure.

6 is a schematic configuration diagram of an entire system to which an apparatus for managing high-resolution skin clinical images according to an embodiment of the present invention is applied.

As shown in FIG. 6, the entire system includes a high-resolution skin clinical image management device 10000, an auxiliary terminal 1000, a camera 2000, a derma camera 3000, and a skin image management server 4000. will be composed

More specifically, the auxiliary terminal 1000 obtains patient information and body map information from the skin image management server and displays them on the screen, and provides information on points checked on the body map displayed on the screen to the skin image management server to perform the function to

For example, as shown in FIG. 9 , the patient information and body map information provided from the skin image management server 4000 are displayed on the screen, and the practitioner sets the point 1100 to be treated on the body map. will become

Thereafter, information on points checked on the body map is provided to the skin image management server to be stored and managed.

When the camera 2000 obtains patient information and point information from the skin image management server, the skin clinical image of the patient corresponding to the point information is captured to generate skin clinical image information including the patient information, and the generated skin Clinical image information will be provided to the skin image management server.

For example, as shown in FIG. 9 , the entire abdominal region corresponding to the acquired point information is photographed, and skin clinical image information in which the photographed skin clinical image is matched with the patient information is generated and sent to the skin image management server. It provides storage and management.

The dermacopy 3000 takes a lesion image present in a skin clinical image taken from the camera 2000 for artificial intelligence learning, and provides the captured high-resolution lesion image to the skin image management server.

At this time, the skin image management server 4000 stores patient information and body map information, and stores the point information provided from the auxiliary terminal, the skin clinical image information provided from the camera, and the actual high-resolution lesion image provided from the dermacopy in the corresponding patient information field. It is matched and stored, and a pax page capable of displaying the skin clinical image information for each point of the patient and the high-resolution synthetic dermacopy image generated with respect to the lesion site image extracted by the lesion site extraction means is provided to the diagnostic terminal.

For example, the point information of the abdomen provided by the auxiliary terminal, the skin clinical image information of the abdomen provided by the camera, and the actual high-resolution lesion image provided by the dermascopy are matched to the corresponding patient information field and stored.

In addition, the PAX page is provided to the diagnostic terminal, and the skin clinical image information for each point of the patient and the actual high-resolution lesion image are displayed through the PAX page.

In general, in the case of dermatology, the skin surface is checked using a skin imaging means, for example, a magnifying glass. are suffering

In order to solve this problem, the problem of conventional misdiagnosis can be solved by establishing a photographing process, obtaining consistent information through this, and storing and managing it.

Hereinafter, the above-described components will be described in detail with reference to the drawings.

7 is a block diagram of a skin image management server 4000 according to an embodiment of the present invention.

As shown in FIG. 7 , the skin image management server 4000 includes a body map information providing unit 4100 , an image information obtaining unit 4200 , a skin image information storage processing unit 4300 , and a patient skin image information storing unit 4400 . ), is configured to include a fax page providing unit (4500).

Specifically, the body map information providing unit 4100 stores body map information and performs a function to provide the body map information to the auxiliary terminal.

That is, by providing the body map as shown in FIG. 9 to the auxiliary terminal, the diagnostician checks the suspected lesion region on the body map displayed on the screen of the auxiliary terminal.

For example, in the case of an X-ray that is generally used in departments other than dermatology, the location of the lesion is not necessarily indicated.

Because most of the shots are taken in a large area, a specialist can immediately determine whether it is the hand, leg, body, or head of the human body.

However, since the photographed image used in dermatology uses a skin image enlarged through a magnifying glass, the location of the photographing must be provided so that a specialist can see it and identify it immediately. In order to confirm the location, the specialist calls the photographer or receives a separate message from the photographer and then goes through the process of checking.

Therefore, in the case of a large dermatology clinic, the waiting time for treatment was long, which made patients reluctant to go to the dermatology clinic.

And, the image information acquisition unit 4200 performs a function for acquiring the point information provided from the auxiliary terminal, the skin clinical image information provided from the camera, and the skin derma image information provided from the derma camera.

For example, it is to acquire point information on a lesion located in the abdominal region, skin clinical image information on the abdominal region including the lesion, and an actual high-resolution lesion image provided by dermascopy.

Then, the skin image information storage processing unit 4300 matches the point information, the skin clinical image information, and the actual high-resolution lesion image acquired in the image information acquisition unit to the corresponding patient information field, and stores it in the patient skin image information storage unit. .

For example, patient detailed information including patient unique identification information is stored for each patient. The detailed patient information includes not only patient ID, but also patient name, date of birth, gender, and the like.

And, the photographed skin clinical image information and skin derma image information, photographing date information, unique identification information of the camera 2000 and the dermacopy (3000), unique identification information of the treating doctor, doctor opinion information, photographed lesion location information , characterized in that it includes information about the illumination applied when photographing the lesion.

Information such as #id_01#kimsun#800112#female#abc-jpeg#202103031000#camera_id_01#doctor_id_01#opinion-txt#face#uv will be saved.

That is, point information, skin clinical image information, and actual high-resolution lesion images are stored in the patient skin image information storage unit 4400 for each patient information field.

The pax page providing unit 4500 displays the skin clinical image information for each point of the patient and a high-resolution synthetic dermacopy image for the lesion site, and diagnoses the pax page 4510 for registering or modifying a patient and searching for a specific patient. The function provided to the own terminal 5000 is performed.

That is, as shown in FIG. 8, the fax page 4510 is

a patient information retrieval layer 4510A for displaying a retrieval button for retrieving patient information;

a patient information display layer 4510B for displaying the retrieved patient information;

a patient information update layer 4510C for displaying a patient addition or patient modification button;

a history image information layer 4510D for displaying the searched history image information of the patient;

a body map location information layer 4510E for displaying body map point information corresponding to the searched history image of the patient;

a skin clinical image information layer 4510F for displaying skin clinical image information including the expressed body map points;

It includes; a synthetic skin derma image information layer (4510G) for displaying a high-resolution synthetic dermacopy image for the lesion site corresponding to the expressed body map point.

Specifically, the patient information search layer 4510A displays a search button for searching patient information, and when a patient name is written on the layer and the search button is selected, the patient information corresponding to the patient name is retrieved. Gathering will provide it to the screen.

That is, the retrieved patient information is displayed through the patient information display layer 4510B.

The patient information update layer 4510C is a layer for expressing a patient add or patient edit button. If patient information does not exist, patient information is added, and when patient information exists but detailed patient information is changed, patient information will be able to edit.

The history image information layer 4510D is a layer for displaying the searched history image information of the patient, and provides history image information that the patient has treated so far so that the diagnostician can refer to it.

The body map location information layer 4510E is a layer for displaying body map point information corresponding to the searched history image of the patient, and intuitively confirms the location on the body map for a specific lesion (point) that the diagnostician has currently photographed. This allows for a comprehensive judgment.

The skin clinical image information layer 4510F is a layer for expressing skin clinical image information including the expressed body map points, and for example, provides overall clinical image information about the abdomen on the body map. This allows for a comprehensive judgment.

The skin composite derma image information layer 4510G is a layer for expressing a high-resolution synthetic dermacopy image for the lesion site corresponding to the expressed body map point, so that the presence or absence of a lesion can be precisely checked through the high-resolution synthetic image image. will do

Therefore, during dermatology treatment, the diagnostician can determine the body map position, clinical image, synthetic derma image, etc. while comprehensively checking the effect.

On the other hand, the skin image management server 4000,

At least one of PACS (Picture Archiving and Communication System), EMR (Electronic Medical Record), PHR (Personal Health Record), RIS (Radiology Information System), CRM (Customer Relationship Management), or characterized in that it is a cloud server that can be connected thereto do.

Through this, the specialists can access the above-described server and check the history information of the patient in the past.

On the other hand, as shown in FIGS. 10 to 17 , the photographing device 2000 described in the present invention includes a first lighting case 100 , a photographing apparatus case 200 , a photographing apparatus 300 , and a control case 400 . ), and the control unit 500 is included.

Specifically, the first lighting case 100 is characterized in that a central opening 110 of a certain size is formed, and a plurality of lighting modules 120 are installed at regular intervals. The first lighting case 100 is characterized in that it has a circular or polygonal shape.

As shown in FIGS. 10 to 17 , a ring-shaped first lighting case having a central opening 110 of a certain size is formed, and a plurality of lighting modules 120 are installed inside the first lighting case. Arranged to form a circle-shaped lighting layer.

In the example of the present invention, although the first lighting case is formed in a ring shape, it is obvious that it can be configured in various polygonal shapes such as a triangle, a square, and a rhombus in addition to the ring shape.

And, the photographing device case 200 is formed across the central opening formed in the first lighting case, has an opening 210 at a position corresponding to the camera of the photographing device, and can mount the photographing device in the interior space. structure will be provided.

As shown in FIGS. 10 to 17 , the photographing device case 200 is formed to cross the central opening formed in the first lighting case, and the front of the photographing device case has an opening at a position corresponding to the camera of the photographing device. The 210 is formed so that the camera is exposed to the outside so that an image can be taken through the camera.

In addition, the photographing device case 200 provides an internal space so that the photographing apparatus 300 can be mounted thereon.

For example, an internal space may be formed by combining the observer-side first housing unit and the observation target-side second housing unit, and the imaging device will be mounted in the formed internal space.

In addition, the control case 400 is formed on the lower side of the first lighting case 100, and provides a structure in which a control unit is installed therein.

10 to 17, a control case having a certain size is connected to the lower side of the first lighting case, and an internal space is provided to form the control unit 500 in the corresponding internal space.

At this time, the bottom surface of the control case is configured to have a certain flatness, so that when the ring-shaped first lighting case is positioned in a general horizontal place, it is not inclined.

In addition, the control unit 500 is formed inside the control case 400 to perform a function for controlling the operation of a plurality of lighting modules and the photographing device.

For example, when a button-on signal is provided from the photographing button unit 750, a lighting signal is provided to a plurality of lighting modules to turn on the lighting module, and an operation signal is provided to a photographing device to perform photographing in a state in which the lighting is turned on. controlling what you can do.

On the other hand, according to an additional aspect, the present invention camera 2000,

It is formed around the opening 210 formed in the photographing device case 200, and a second lighting case 600 in which a plurality of lighting modules 610 are installed and configured at regular intervals; characterized in that it is configured to further include .

10 to 17 , in the second lighting case 600 formed around the opening 210 formed in the photographing device case 200 , the photographing device case 200 is connected to the first lighting case 100 . In the combined state, it functions as an auxiliary light, and when the photographing device case 200 is separated from the first lighting case 100, it functions as a main light for close-up photography.

A plurality of lighting modules 120 formed in the first lighting case 100 because the photographing is performed with the photographing device case 200 coupled to the first lighting case 100 when photographing a large area in a wide range. is the main lighting, and when necessary (eg, when the amount of light is insufficient), a plurality of lighting modules 610 formed in the photographing device case 200 are used as auxiliary lighting.

On the other hand, when taking a close-up photographing of a local area, since the photographing is performed with the photographing device case 200 separated from the first lighting case 100 , a plurality of lighting modules 610 formed in the photographing device case 200 . is used as an auxiliary light.

For example, if epithelial cells in the mouth are photographed, it corresponds to a local area, so in this case, close-up photography can be performed by separating the imaging device case 200 from the first illumination case 100 .

In addition, the plurality of lighting modules 120 and the plurality of lighting modules 610 described above may adopt a light source that provides various wavelength bands such as LED, UV, IR, etc., and is limited to any one wavelength band. not. Accordingly, by controlling the wavelength for each of the various observation objects, a clinical image image capable of accurately examining the state of the object to be observed is obtained.

On the other hand, according to another additional aspect, the camera 2000 applied to the present invention,

It is characterized in that it comprises at least one first handle portion 700 installed in the first lighting case 100 and a photographing button portion 750 formed on one side of the first handle portion 700 . .

10 to 17, at least one first handle portion 700 is installed on one side of the first lighting case 100, and in the example, two handle portions are located on the left and right sides, respectively. is forming

At this time, the first handle unit 700 is configured to install the shooting button unit 750 on one side, when the shooting button unit 750 is pressed by the user, the shooting button unit 750 transmits a button-on signal to the control unit 500 ) is characterized in that it is provided.

On the other hand, according to another additional aspect, the present invention camera 2000,

It characterized in that it is configured to further include; at least one second handle portion 800 is configured to be installed on either side of the photographing apparatus case 200.

10 to 17, the second handle portion 800 is formed on the left and right sides of the photographing device case 200, respectively, and after the user grips the second handle portion using the second handle portion, photographing is performed. will be able to do

In addition, as another embodiment, in the case of the second handle, since it is configured in the photographing apparatus case 200 , the photographing apparatus case can serve as a handle in a state separated from the first lighting case 100 .

On the other hand, according to another additional aspect, the present invention camera 2000,

In the case of configuring the rechargeable power supply unit 450 in the control case 400, the pedestal 900 having a power supply interface 950 for providing external power to the rechargeable power supply unit is configured to further include; .

In order to supply power to the control unit formed inside the control case and the lighting modules formed in the first lighting case 100 and the photographing device case 200, a power connector to which commercial power is connected may be configured, but it is inconvenient to carry and move This is to configure the rechargeable power supply inside the control case because it can provide performance.

For example, when the battery is configured as a rechargeable power supply unit, the pedestal 900 provided with a power supply interface for supplying external power to the battery is configured.

That is, since the power supply interface is configured on the pedestal, if the charging port for supplying power to the battery is contacted here, it is possible to easily charge the battery when power is insufficient.

On the other hand, according to another additional aspect, the camera 2000 applied to the present invention,

By forming a first magnetic force unit 1100 around the central opening 110 of the first lighting case 100 and forming second magnetic force units 1200 at both ends of the photographing apparatus case 200, the first It is characterized in that the first lighting case 100 and the photographing apparatus case 200 are detachably coupled to each other by the magnetic force of the magnetic force part and the second magnetic force part.

As shown in FIGS. 10 to 17 , a first magnetic force unit is formed on the inner side of the first lighting case 100 and around the central opening 110 , and a second magnetic force unit is formed at both ends of the photographing device case.

The first magnetic force portion is formed around the central opening 110 in a cylindrical shape, it characterized in that a part is exposed to the outside. That is, the magnetic part is configured in a cylindrical shape, and when the second magnetic part is brought into contact with the exposed part by exposing a portion, for example, 0.01 mm to 0.1 mm from the outer surface, it can be coupled by a strong magnetic force.

And, the first magnetic force part and the second magnetic force part are characterized in that the neodymium magnet (neodymium magnet).

In other words, neodymium magnets have the strongest magnetic force (25~50MGOe) among magnets currently used on earth, have good workability, and are cheaper than samarium-cobalt or alnico magnets, but they rust easily and the surface is coated with nickel. It is used after plating, and the temperature coefficient is low, so it has a disadvantage that the magnetism is easily weakened by heat.

As described above, by forming the first magnetic part and the second magnetic part in the first lighting case and the photographing device case, respectively, it is possible to attach and detach between the first lighting case and the photographing device case, so that a large area photographing or a local area photographing can be selectively performed. to enable it to be done.

For example, when photographing a large area, the photographing device case is combined with the first lighting case and used as an integrated body. It is to enable imaging of a local area using only the

On the other hand, according to another additional aspect, in the camera 2000 applied to the present invention, a first connection port 1300 is formed around the first magnetic force part, and a second connection port ( 1310) is formed.

Therefore, when the photographing device case 200 is coupled to the first lighting case 100, the first connection port 1300 and the second connection port 1310 are coupled, and the first lighting case 100 and the photographing device case 200 is electrically connected.

When the first lighting case 100 and the photographing device case 200 are electrically connected by the combination of the first connection port 1300 and the second connection port 1310 , the rechargeable power supply unit configured in the control case 400 ( 450) is provided to the photographing device case 200 through the first lighting case 100, and for lighting the plurality of lighting modules 610 of the second lighting case 600 formed in the photographing device case 200 power will be applied.

Meanwhile, according to another additional aspect, a rechargeable power supply unit may be configured inside the photographing device case 200 .

That is, when the photographing device case 200 is separated from the first lighting case 100 and used independently, power cannot be supplied from the rechargeable power supply unit 450 formed in the control case 400, and thus the photographing device case is independently used. As described above, to supply power to the plurality of lighting modules 610 of the second lighting case 600 formed in the 200, a rechargeable power supply unit is configured inside the photographing device case 200.

Meanwhile, the photographing device 300 described in the present invention is a smart device including a display screen. That is, if the imaging device is configured as a smart device with a display screen, the user can take pictures while checking the clinical imaging site through the display screen. It provides the advantage of being able to manage, and the smart device may be a smart phone.

Meanwhile, according to another additional aspect, the camera 2000 applied to the present invention is characterized in that it automatically uploads the captured clinical image image to the cloud or data storage.

Through this, without the need to perform an inconvenient secondary operation of downloading the captured clinical image images to a separate storage device and storing them in an external terminal, simply press the shooting button to capture and store the clinical image of the observed subject currently being observed by the observer, or Clinical image information captured by the terminal can be transmitted in real time.

According to the present invention, it is possible to improve the inconvenience of specifying the lesion site in the skin clinical image and providing daily dermatophyte photos, thereby automatically generating a high-resolution synthetic dermatography image for all microlesion sites in the skin clinical image. By enabling storage and management, the tweezers data acquisition method that actually takes a dermoscopy picture of a specific lesion site after reviewing the high-resolution synthetic dermatography image is provided, thereby simplifying the skin diagnosis procedure.

In the above, preferred embodiments of the present invention have been illustrated and described, but the present invention is not limited to the specific embodiments described above, and it is common in the technical field to which the present invention pertains without departing from the gist of the present invention as claimed in the claims. Various modifications may be made by those having the knowledge of, of course, and these modifications should not be individually understood from the technical spirit or perspective of the present invention.

1000: auxiliary terminal
2000: camera
3000 : Derma camera
4000: skin image management server
5000: diagnostic terminal
6000: artificial intelligence learning model creation means
7000: lesion site extraction means
8000: means for performing high resolution

Claims (15)

In the management apparatus of high-resolution skin clinical image,
Auxiliary terminal 1000 for acquiring patient information and body map information from the skin image management server 400 and displaying it on the screen, and providing the checked point information on the body map displayed on the screen to the skin image management server;
When patient information and point information are acquired from the skin image management server 400, a skin clinical image of a patient corresponding to the point information is taken to generate skin clinical image information including patient information, and the generated skin clinical image information A camera (2000) for providing as a skin image management server,
For artificial intelligence learning, a dermacopy (3000) for photographing a lesion image existing in a skin clinical image taken from a camera (2000), and providing an actual high-resolution lesion image taken as a skin image management server;
The patient information and body map information are stored, and the point information provided by the auxiliary terminal, the skin clinical image information provided by the camera, and the actual high-resolution lesion image provided by the dermacopy are matched to the corresponding patient information field and stored, and the patient's skin by point A skin image management server 4000 for providing a PAX page capable of displaying clinical image information and a high-resolution synthetic dermacopy image generated with respect to the lesion site image extracted by the lesion site extraction means to the diagnostic terminal 5000; ,
A high-resolution lesion image is generated for the lesion image by using a high-resolution generation algorithm by acquiring an image of a lesion existing in a skin clinical image taken from the camera 2000, and an actual high-resolution lesion image taken from the dermacopy (3000) Artificial intelligence that generates a high-resolution synthetic dermascopy image generation learning model by determining whether the generated high-resolution lesion image and the actual high-resolution lesion image are authentic using the authenticity classification algorithm, and when the accuracy is greater than the reference accuracy, the learning is terminated Learning model generating means (6000) and,
A lesion site extraction means 7000 for extracting a lesion site in a skin clinical image taken by a camera 2000 to generate a high-resolution synthetic dermatophyte image;
By using the high-resolution synthetic dermacopy image generation learning model generated by the artificial intelligence learning model generating means 6000, high-resolution synthetic dermatography images are generated by performing high-resolution on the lesion site images extracted by the lesion site extraction means. A high-resolution skin clinical image management apparatus, characterized in that it comprises a high-resolution performing means (8000) for providing to the skin image management server (4000).
The method of claim 1,
High-resolution skin clinical image management apparatus, characterized in that it further comprises a polarization processing means (8500) for polarization processing by extracting the high-resolution synthetic dermacopy image generated by the high-resolution performing means.
3. The method of claim 2,
High-resolution skin clinical image management apparatus, characterized in that it further comprises a spatial distortion processing means (9000) for reducing spatial distortion by obtaining a high-resolution synthetic dermascopic image polarized by the polarization processing means.
The method of claim 1,
The artificial intelligence learning model generating means 6000,
a clinical skin image acquisition unit 6100 for acquiring an image of a lesion existing in a skin clinical image taken from the camera 2000;
a high-resolution lesion image generator 6200 for receiving the lesion image present in the acquired skin clinical image and generating a high-resolution lesion image for the lesion image using a high-resolution generation algorithm;
an actual high-resolution lesion image acquisition unit 6300 for acquiring an actual high-resolution lesion image taken from the dermacopy 3000;
Dermasscopy that receives the generated high-resolution lesion image and the actual high-resolution lesion image, determines the authenticity using the authenticity classification algorithm, and terminates the learning when the accuracy is greater than the reference accuracy. High-resolution skin clinical image management apparatus comprising a; AI learning model generation unit (6400) for image generation.
delete The method of claim 1,
The skin image management server 4000,
At least one of PACS (Picture Archiving and Communication System), EMR (Electronic Medical Record), PHR (Personal Health Record), RIS (Radiology Information System), CRM (Customer Relationship Management), or characterized in that it is a cloud server that can be connected thereto A high-resolution skin clinical image management device.
The method of claim 1,
The skin image management server 4000,
a body map information providing unit 4100 for storing body map information and providing the body map information to an auxiliary terminal;
an image information acquisition unit 4200 for acquiring the point information provided from the auxiliary terminal, the skin clinical image information provided from the camera, and the actual high-resolution lesion image provided from the dermacopy;
a skin image information storage processing unit 4300 for matching the point information obtained in the image information acquisition unit, the skin clinical image information, and the actual high-resolution lesion image to the corresponding patient information field, and storing and processing in the patient skin image information storage unit;
a patient skin image information storage unit 4400 that stores point information, skin clinical image information, and actual high-resolution lesion images for each patient information field;
Pax, which displays skin clinical image information for each point of the patient and a high-resolution synthetic dermacopy image of the lesion site, and provides a pax page 4510 that can register or modify a patient, and search for a specific patient as a diagnostic terminal 5000 Page providing unit (4500); High-resolution skin clinical image management apparatus comprising a.
8. The method of claim 7,
The pax page 4510,
a patient information retrieval layer 4510A for displaying a retrieval button for retrieving patient information;
a patient information display layer 4510B for displaying the retrieved patient information;
a patient information update layer 4510C for displaying a patient add or patient edit button;
a history image information layer 4510D for displaying the searched history image information of the patient;
a body map location information layer 4510E for displaying body map point information corresponding to the searched history image of the patient;
a skin clinical image information layer 4510F for displaying skin clinical image information including the expressed body map points;
A high-resolution skin clinical image management apparatus comprising a; skin synthetic derma image information layer (4510G) for displaying a high-resolution synthetic dermacopy image for a lesion site corresponding to the expressed body map point.
delete delete delete delete delete delete delete

Images