KR20230050253A - Method for detecting pleurl effusion and the apparatus for therof - Google Patents

Abstract

The present disclosure relates to a method for detecting pleural effusion and an apparatus therefor. The method for detecting pleural effusion according to an embodiment of the present disclosure may include the steps of: obtaining a chest image generated corresponding to a continuous volume of the chest of a user; generating first image information by removing the heart and blood vessels surrounding the heart from the chest image; generating second image information by extracting a lung region including a lung and a pleura from the first image information; and detecting a site of pleural effusion between the lung and the pleura from the second image information.

Description

Method for detecting pleural effusion and device therefor

The technical idea of the present disclosure relates to a method for detecting pleural effusion and an apparatus therefor.

Pleural effusion is a disease in which fluid is exuded from the pleura of the lungs. Normally, a small amount of bodily fluid is constantly produced and entered into the pleura, but when a disease in which bodily fluid production increases or absorption of bodily fluid is reduced, excessive pleural fluid accumulates in the pleura, resulting in pleural effusion.

In order to relieve pleural effusion, pleural puncture is performed in which a needle is inserted into the pleural space to drain the pleural fluid to the outside of the body and a portion of the pleural fluid is drawn out with a syringe.

To this end, conventionally, after manually measuring the size of the pleural effusion site shown in the two-dimensional chest X-ray image or ultrasound image, the total amount was estimated based on this, but there was a problem in that it was difficult to accurately locate it.

In addition, attempts have been made to semi-automatically or automatically detect pleural effusion from radiographic images to solve this problem, but since the pleural fluid has a similar brightness value to other soft tissues around the lungs, it is difficult to determine the exact boundary.

A technical problem to be achieved by a method for detecting pleural effusion and an apparatus therefor according to the technical idea of the present disclosure is a method for detecting pleural effusion capable of accurately and quickly detecting a lung region from a chest image and detecting pleural effusion therefrom. and to provide the device.

Technical tasks to be achieved by a method for detecting pleural effusion and an apparatus therefor according to the technical idea of the present disclosure are not limited to the tasks mentioned above, and other tasks not mentioned can be clearly understood by those skilled in the art from the description below. There will be.

According to one aspect of the technical concept of the present disclosure, a method for detecting pleural effusion includes obtaining a chest image generated corresponding to a continuous volume of a user's chest; generating first image information by removing the heart and blood vessels surrounding the heart from the chest image; generating second image information by extracting a lung region including a lung and a pleura from the first image information; and detecting a site of pleural effusion between the lung and the pleura from the second image information.

According to an exemplary embodiment, the blood vessels around the heart may include an aorta, a vena cava, a pulmonary artery, and a pulmonary vein connected to the heart.

According to an exemplary embodiment, the generating of the second image information may include recognizing a chest and a diaphragm from the first image information; and extracting the lung region located inside the chest and the diaphragm from the first image information.

According to an exemplary embodiment, the generating of the second image information may include recognizing a chest and a diaphragm from the first image information; and removing the ribcage and the diaphragm from the first image information.

According to an exemplary embodiment, the method may further include calculating a volume of the pleural effusion site.

According to an exemplary embodiment, the method may further include generating third image information in which the pleural effusion site is labeled with a predetermined identification mark.

According to an exemplary embodiment, the chest image may be a 3D image or an image group including a plurality of 2D images.

According to one aspect of the technical concept of the present disclosure, an apparatus for detecting pleural effusion includes a memory for storing a program for detecting pleural effusion; and by executing the program, a chest image generated corresponding to the continuous volume of the user's chest is acquired, and first image information is generated by removing the heart and blood vessels around the heart from the chest image, and the first image information It may include at least one processor for generating second image information by extracting a lung region including a lung and a pleura, and detecting a region of pleural effusion between the lung and the pleura from the second image information.

According to one aspect of the technical idea of the present disclosure, a computer program stored in a recording medium may be provided to execute a method for detecting pleural effusion.

According to the apparatus for detecting pleural effusion and the apparatus for detecting pleural effusion according to embodiments according to the technical concept of the present disclosure, a lung region in a chest image can be quickly and accurately detected, and pleural effusion can be detected.

In addition, by determining the location and/or amount of pleural effusion, pleural puncture (discharge of pleural fluid through a tube) can be performed accurately and safely. In particular, the amount of pleural fluid to be discharged can be an indicator for accurately understanding the progress of treatment.

also, In the process of diagnosing and treating pneumonia, trauma, tuberculosis, cancer, etc. with images, the probability of misdiagnosis can be reduced by actively notifying when pleural effusion is accompanied.

In addition, when mild pleural effusion occurs, there may be little or no subjective symptoms, including pain, but the patient automatically recognizes and warns of pleural effusion during periodic imaging through health checkups, etc. for early examination and treatment. can be made possible.

In addition, by providing the location of pleural effusion expressed as a 3D image, it helps to determine the location of the puncture for effective drainage of pleural fluid, and reduces the number of radiographs to identify residual pleural fluid, thereby reducing concerns about radiation exposure of patients and It can reduce the burden of inspection cost.

Effects obtainable by the device for detecting pleural effusion and the device therefor according to the technical idea of the present disclosure are not limited to the effects mentioned above, and other effects not mentioned are in the technical field to which the present disclosure belongs from the description below. It will be clearly understood by those skilled in the art.

A brief description of each figure is provided in order to more fully understand the figures cited in this disclosure.
1 is a flowchart illustrating a method for detecting pleural effusion according to an embodiment of the present disclosure.
2 to 4 are exemplary views for explaining a process of detecting pleural effusion according to an embodiment of the present disclosure.
5 is a flowchart illustrating a process of generating an average shape of a target body part in a method for detecting pleural effusion according to an embodiment of the present disclosure.
6 is a flowchart illustrating a process of recognizing a body part by generating a shape of a target body part from image information in a method for detecting pleural effusion according to an embodiment of the present disclosure.
7 is a block diagram briefly illustrating the configuration of an apparatus for detecting pleural effusion according to an embodiment of the present disclosure.

Since the technical spirit of the present disclosure may be subject to various changes and may have various embodiments, specific embodiments are illustrated in the drawings and described in detail. However, this is not intended to limit the technical spirit of the present disclosure to specific embodiments, and should be understood to include all changes, equivalents, or substitutes included in the scope of the technical spirit of the present disclosure.

In describing the technical idea of the present disclosure, if it is determined that a detailed description of a related known technology may unnecessarily obscure the subject matter of the present disclosure, the detailed description will be omitted. In addition, numbers (eg, first, second, etc.) used in the description process of the present disclosure are only identifiers for distinguishing one component from another component.

In addition, in the present disclosure, when one component is referred to as "connected" or "connected" to another component, the one component may be directly connected or directly connected to the other component, but in particular Unless otherwise described, it should be understood that they may be connected or connected via another component in the middle.

In addition, terms such as "~ unit", "~ group", "~ character", and "~ module" described in the present disclosure mean a unit that processes at least one function or operation, which includes a processor, a micro Processor (Micro Processor), Micro Controller, CPU (Central Processing Unit), GPU (Graphics Processing Unit), APU (Accelerate Processor Unit), DSP (Digital Signal Processor), ASIC (Application Specific Integrated Circuit), FPGA (Field Programmable Gate Array) may be implemented by hardware or software or a combination of hardware and software.

In addition, it is intended to make it clear that the classification of components in the present disclosure is merely a classification for each main function in charge of each component. That is, two or more components to be described below may be combined into one component, or one component may be divided into two or more for each more subdivided function. In addition, each component to be described below may additionally perform some or all of the functions of other components in addition to its main function, and some of the main functions of each component may be performed by other components. Of course, it may be dedicated and performed by .

The method for detecting pleural effusion according to an embodiment of the present disclosure may be performed in a personal computer, a workstation, a computer device for a server, or the like, or a separate device for this purpose.

Also, the method may be performed on one or more computing devices. For example, at least one or more steps of a method according to an embodiment of the present disclosure may be performed by a client device and other steps may be performed by a server device. In this case, the client device and the server device may be connected through a network to transmit and receive calculation results. Alternatively, the method may be performed by distributed computing technology.

Throughout this specification, network function may be used interchangeably with neural network and/or neural network. Here, a neural network (neural network) may be composed of a set of interconnected computational units, which may be generally referred to as nodes, and these nodes may be referred to as neurons. A neural network is generally composed of a plurality of nodes. Nodes constituting a neural network may be interconnected by one or more links.

Some of the nodes constituting the neural network may form one layer based on distances from the first input node. For example, a set of nodes having a distance of n from the first input node may constitute n layers.

The neural network described in this specification may include a deep neural network (DNN) including a plurality of hidden layers in addition to an input layer and an output layer.

Hereinafter, embodiments of the present disclosure will be described in detail in turn.

1 is a flowchart illustrating a method for detecting pleural effusion according to an embodiment of the present disclosure.

In step S110 , the pleural effusion detection device may acquire a chest image generated corresponding to the continuous volume of the chest. The chest image may be, for example, an MR scan image, a CT scan image, an X-RAY scan image, or the like, generated by capturing the user's chest. For example, a chest image is a 3D image and may be divided into a plurality of 2D images. Also, for example, the chest image is an image group including a plurality of 2D images, and each 2D image may be a plurality of slices obtained by continuously photographing a cross section of the chest in one direction. By stacking these 2D images, it is possible to obtain 3D image information on the chest. The chest can include the area of the body that includes at least the diaphragm to the ribcage (chest cage).

In operation S120 , the apparatus for detecting pleural effusion may generate first image information by removing living organs adjacent to the lung region from the chest image. In this case, the living organ adjacent to the lung region may include the heart and blood vessels around the heart, and the blood vessels around the heart may include the aorta, vena cava, pulmonary artery, and pulmonary vein connected to the heart. That is, in operation S120 , the lung region may be separated from surrounding tissues, organs, and the like in the chest image by removing the heart adjacent to the lung region and blood vessels surrounding the heart. This is in consideration of the fact that when pleural effusion occurs, the pleural fluid has a similar shade and brightness to other living organs (eg, heart) adjacent to the lung area, so it is difficult to clearly determine the boundary. This is to first recognize and remove it from the chest image, so that the lung region remains in the first image information as a result.

In an embodiment, step S120 may include identifying a heart and blood vessels surrounding the heart from the chest image; and removing the heart and peripheral blood vessels from the chest image.

Depending on the embodiment, step S120 may be performed by inputting the chest image to the learned first network function. At this time, the first network function learns to extract the heart and/or blood vessels surrounding the heart in advance through learning data (eg, an image including a heart labeled by an expert, a blood vessel around the heart, etc.) may have been

In operation S130 , the apparatus for detecting pleural effusion may generate second image information by extracting a lung region from the first image information. Here, the lung region may include lungs and pleura surrounding the lungs (ie, visceral pleura and parietal pleura).

Depending on embodiments, the second image information may be performed by removing regions other than the lung region from the first image information or by separating the lung region from the first image information. In this case, the second image information may be generated by displaying only the lung region or by labeling the lung region with a predetermined identification mark in the chest image, the first image information, or image information processed therefrom. Also, the second image information may be, for example, data including coordinate information about the lung region.

In an embodiment, step S130 may include recognizing a chest and a diaphragm from the first image information; and extracting a lung region located inside the thorax and diaphragm from the first image information. That is, in the first image information, the inner regions of the thorax and diaphragm are first recognized, and in the corresponding region, biological tissues around the lungs (ie, the heart and blood vessels around the heart) are removed in step S120, so that only the lung regions are separated from the surroundings. Since it remains to be distinguished, it is possible to easily extract the lung region to generate second image information.

In an embodiment, step S130 may include recognizing a chest and a diaphragm from the first image information; and removing the chest and diaphragm from the first image information. By removing the chest and the diaphragm from the first image information, the second image information in which only the lung region remains may be generated. Additionally, step S130 may further include removing biological tissue unnecessary for detecting pleural effusion. In this case, the unnecessary living tissue may include muscle, membrane, and the like.

Depending on the embodiment, step S130 may be performed by inputting the first image information to the learned second network function. At this time, the second network function is trained on extraction of the chest, diaphragm, lung region, etc. in advance through learning data (eg, images including the chest, diaphragm, lung region, etc. labeled by experts, etc.) it could be

In step S140 , the pleural effusion detecting apparatus may detect a site of pleural effusion from the second image information. Pleural effusion is the abnormal accumulation of pleural fluid exuded from the pleura in the pleural cavity (the space between the lungs and the pleural cavity), and the pleural cavity is detected based on changes in the shading and shape in the lung area (particularly, the pleural cavity) in the second image information. The site of exudation can be detected.

Depending on the embodiment, step S140 may be performed by inputting the second image information to the learned third network function. In this case, the third network function may be one in which extraction of the pleural effusion site has been previously learned through learning data (eg, an image including pleural fluid in the lung area labeled by an expert, etc.). This is an example, and various image processing techniques (eg, active contour, etc.) may be applied to step S140 according to embodiments.

In embodiments, method 100 may further include calculating, by the device for detecting pleural effusion, a volume of a site of pleural effusion. Calculation of the volume can be performed in various ways using image information. For example, in the above step, the volume of the pleural effusion site may be calculated by calculating the area of the pleural effusion site in each 2D image of the second image information and summing the area. Also, for example, the above step may be performed by calculating a volume based on the number of pixels corresponding to the pleural effusion region and the volume per pixel in the second image information (eg, a 3D image). However, it is not limited thereto.

In an embodiment, the method 100 may further include generating third image information in which the pleural effusion detection apparatus labels the pleural effusion site with a predetermined identification mark. The third image information may be a labeling of a pleural effusion site with a predetermined identification mark. In this case, the third image information may include a plurality of 2D images and/or a 3D image generated by combining them.

In an embodiment, the third image information may be displayed by overlaying a pleural effusion region on an image of a lung region generated based on the second image information. The pleural effusion site and other areas can be distinguished by different colors, different transparencies, different indicators, and the like. However, this is illustrative, and according to embodiments, the third image information may be displayed by overlapping the pleural effusion region in the chest image and the first image information.

In an embodiment, the first to third network functions in the method 100 may be generated by the same or at least partially different learning models.

2 to 4 are exemplary views for explaining a process of detecting pleural effusion according to an embodiment of the present disclosure.

Referring to Figure 2, a chest image is shown. In the chest image, the thorax, diaphragm (labeled in yellow), heart (labeled in red), and peripheral blood vessels (labeled in red) may be recognized and separately labeled. In the chest image of FIG. 2 , first image information may be generated by removing red heart and peripheral blood vessels from the chest cavity.

Referring to FIG. 3 , second image information is shown. The second image information may be generated by extracting the lung region from the first image information. In this case, the lung region may include a lung and a pleura surrounding the lung.

Referring to FIG. 4 , third image information is shown. The third image information may be generated by detecting a pleural effusion site from the second image information and labeling it with a red color to overlap it. The location and amount of pleural effusion can be grasped more intuitively through the third image information.

The drawings shown in FIGS. 2 to 4 are exemplary, and various image information may be used according to an embodiment of the present disclosure.

In the method for detecting pleural effusion according to an embodiment of the present disclosure, from image information (eg, chest image, first image information, second image information, image information processed therefrom, etc.) to a body part, for example, It is required to extract at least one of the thorax, diaphragm, heart and blood vessels surrounding the heart. To this end, the method of detecting pleural effusion may perform a process of generating an average shape of a target body part from learning data, generating a shape of a target body part from user image information, and recognizing the corresponding body part.

In this regard, FIG. 5 is a flowchart illustrating a process of generating an average shape of a target body part in a method for detecting pleural effusion according to an embodiment of the present disclosure.

In operation S510 , the apparatus for detecting pleural effusion may acquire a plurality of first medical images generated corresponding to the continuous volume of the user's chest. For example, the first medical image may be received from an external database server or acquired through imaging from an imaging device (eg, X-ray imaging device, CT imaging device, etc.) connected to the device through wired or wireless communication. there is.

In an embodiment, the first medical image may be a computed tomography (CT) image generated by performing tomography scans of body parts including the chest of a plurality of subjects. That is, the first medical image is composed of a plurality of 2D slice images (i.e., image group) generated by continuously photographing body parts including the chest of a plurality of subjects in one direction through a computed tomography method, or , it may be a 3D image generated based on this. However, it is not limited thereto.

In operation S520, the apparatus for detecting pleural effusion may standardize the plurality of first medical images into a predetermined format suitable for extracting an average shape of a target body part. For example, the device may perform normalization by extracting at least a partial area of the first medical image or by enlarging, reducing, and/or rotating the same. Here, the target body part may include at least one of a thorax, a diaphragm, a heart, and blood vessels surrounding the heart.

In an embodiment, step S520 may be performed based on at least one of a living organ and at least one bone part surrounding the target body part. For example, when the target body part is the diaphragm, the apparatus for detecting pleural effusion extracts (cuts) a partial region of the image so that only the heart, liver, and stomach, which are peripheral organs of the diaphragm, are included in each of the first medical images; The first medical images may be normalized in a form corresponding to each other by enlarging, reducing, and/or rotating the regions extracted from each first medical image. In addition, for example, the apparatus for detecting pleural effusion may enlarge, reduce, and/or rotate the first medical images based on bone regions (eg, the spine and ribs, etc.), so that the first medical images correspond to each other. can be normalized in the form However, this is exemplary, and according to embodiments, the device may perform normalization based on various reference points.

Through step S520, living organs and bone parts included in the plurality of first medical images may be transformed into sizes and angles corresponding to each other and disposed on adjacent 3D spatial coordinates.

In operation S530 , the apparatus for detecting pleural effusion may extract an average shape corresponding to the target body part and pixel gradients around the average shape from the plurality of normalized first medical images.

In an embodiment, the extraction of the average shape of the target body part may be performed through a previously learned fourth network function. In this case, the fourth network function may be learning about extraction of the target body part in advance through learning data (eg, CT images including target body parts labeled through experts, examinations, etc.) .

Also, according to embodiments, the extraction of the average shape of the target body part may be performed through image processing based on changes in pixel values. However, this is an example, and is not limited thereto.

Meanwhile, although not shown, the apparatus for detecting pleural effusion may store information about the average shape of the extracted target body part. The average shape of the target body part can be expressed as a surface with curvature in a 3-dimensional space, and the information on the average shape of the target body part includes information on coordinates in the 3-dimensional space constituting the surface and the It may contain information about the gradient of pixels around the average shape.

In addition, although steps S520 and S530 are described as being performed in a 3D coordinate space based on the first 3D medical image, this is exemplary, and according to an embodiment, steps S520 and S530 may be performed on the first medical image. Each of the plurality of 2D images (ie, slice images) constituting the image may be performed. In this case, in step S530, a 3D average shape of the target body part may be generated by extracting the average shape of the target body part from each 2D image and stacking them.

The method 500 described above may be performed simultaneously or separately for various target body parts. In this case, the target body part may include at least one of a thorax, a diaphragm, a heart, and blood vessels surrounding the heart. For example, at least steps S530 and S540 may be performed for the thorax, diaphragm, heart, and blood vessels around the heart, respectively, and information on the average shape of the target body part is also provided for the ribcage, diaphragm, heart, and blood vessels around the heart, respectively. can be stored However, it is not limited thereto.

6 is a flowchart illustrating a process of recognizing a body part by generating a shape of a target body part from image information in a method for detecting pleural effusion according to an embodiment of the present disclosure.

The method 600 may be performed in steps S120 and/or S130 of the method 100 using the information about the average shape of the target body part stored in the method 500 . For example, the method 600 may be used to extract at least one of a living organ adjacent to the lung region, ie, a heart and blood vessels surrounding the heart, in step S120 of the method 100 . Also, for example, method 600 may be used to extract lung regions in step S130 of method 100 . However, it is not limited thereto, and the method 600 may be applied to extract various body parts according to an embodiment of the present disclosure.

Specifically, in step S610 , the apparatus for detecting pleural effusion may obtain a second medical image generated corresponding to the continuous volume of the body part including the chest. Here, the second medical image is an input image for which detection of the target body part is requested, and may be received from an external database server or captured and input from a photographing device connected to the device through wired or wireless communication. there is.

In an embodiment, the second medical image may be a computed tomography (CT) image similar to the first medical image, and may be composed of a plurality of 2D slice images (ie, image group) or a 3D image generated based on the 2D slice images. can be

In an embodiment, the second medical image may be a chest image acquired in step S110 of the method 100, first image information generated in step S120, and/or an image processed therefrom. If the method 600 follows the method 100, step S620 may be performed directly without separately performing step S610. However, it is not limited thereto.

In step S620, the pleural effusion detection apparatus may transform the second medical image to correspond to the first medical image. That is, the apparatus for detecting pleural effusion extracts, enlarges, or extracts at least a partial region of the second medical image to correspond to the first medical image (ie, the normalized first medical image) used to extract the average shape of the target body part. It may be reduced and/or rotated, and accordingly, the second medical image may be deformed suitable for extraction of the target body part. In this case, the deformed second medical image may be mapped on the same 3D coordinate space to correspond to the first medical image used to extract the average shape of the target body part. However, it is not limited thereto.

Similar to step S520 of the method 500, step S620 may be performed based on at least one of a living organ (heart, liver, stomach, etc.) and at least one bone part (chest, etc.) surrounding the target body part.

In operation S630 , the apparatus for detecting pleural effusion may match and arrange the average shape of the target body part to the second medical image based on the information about the average shape of the target body part. That is, based on the coordinates of the average shape of the target body part, the average shape of the target body part may be arranged at a predetermined position of the deformed second medical image. In this case, the average shape of the target body part may be extracted (or generated) from a plurality of first medical images by the method 500 of FIG. 5 .

Depending on the embodiment, the apparatus for detecting pleural effusion may arrange the average shape of the target body part on a plurality of 2D images (ie, slice images) constituting the second medical image. In this case, step S640 described in detail below may be performed for each 2D image.

In step S640, the apparatus for detecting pleural effusion converts the coordinates of at least a portion of the average shape of the target body part disposed in the second medical image, thereby converting the target body part included in the second medical image from the average shape of the target body part. shape can be created.

In an embodiment, in step S640, the apparatus for detecting pleural effusion converts the coordinates of at least a portion of the average shape of the target body part in a vertical or horizontal direction on a three-dimensional coordinate space, and correspondingly converts the coordinates around the average shape of the target body part. The output value of the loss function for the pixel gradient can be measured. Subsequently, the apparatus for detecting pleural effusion may generate the shape of the target body part included in the second medical image by optimizing the coordinates of the average shape of the target body part based on the output value of the loss function.

That is, the loss function outputs a lower value as the position and shape of the target body part included in the second medical image and the average shape of the target body part become similar to each other, and a repetitive process of reducing the output value of the loss function ( The shape of the target body part may be generated by optimizing the coordinates of the average shape of the target body part through coordinate transformation -> measuring the output value of the loss function). To this end, stochastic gradient descent or the like may be applied, but is not limited thereto.

The shape of the target body part generated through the above steps may be a target body part identified and recognized in the method 100 . Depending on the embodiment, coordinate information of the shape of the target body part may be stored as data or may be labeled with a predetermined identification mark and displayed on the second medical image and image information processed therefrom.

For example, the apparatus for detecting pleural effusion may inversely transform the shape of the target body part to correspond to the original second medical image, and display (eg, render) together with the second medical image. That is, the apparatus for detecting pleural effusion performs an inverse transformation on the shape of the target body part in the same format as the second medical image before performing step S620, and then converts the shape of the target body part based on the transformed coordinates of the second medical image. It can be displayed on a three-dimensional space together with at least some of them. However, it is not limited thereto.

In an embodiment, generation of the shape of the target body part may be performed through a previously learned fifth network function. In this case, the fifth network function may be learning about extraction of the target body part in advance through learning data (eg, CT images including target body parts labeled through experts, examinations, etc.) .

7 is a block diagram briefly illustrating the configuration of an apparatus for detecting pleural effusion according to an embodiment of the present disclosure.

The communication unit 710 may receive input data (such as a chest CT image) for detecting pleural effusion. The communication unit 710 may include a wired/wireless communication unit. When the communication unit 710 includes a wired communication unit, the communication unit 710 may include a local area network (LAN), a wide area network (WAN), a value added network (VAN), and a mobile communication network ( mobile radio communication network), a satellite communication network, and one or more components that enable communication through a mutual combination thereof. In addition, when the communication unit 710 includes a wireless communication unit, the communication unit 710 transmits and receives data or signals wirelessly using cellular communication, a wireless LAN (eg, Wi-Fi), and the like. can In one embodiment, the communication unit 710 may transmit/receive data or signals with an external device or an external server under the control of the processor 740 .

The input unit 720 may receive various user commands through external manipulation. To this end, the input unit 720 may include or connect one or more input devices. For example, the input unit 720 may be connected to an interface for various inputs such as a keypad and a mouse to receive user commands. To this end, the input unit 720 may include an interface such as a thunderbolt as well as a USB port. In addition, the input unit 720 may receive an external user command by including or combining various input devices such as a touch screen and buttons.

The memory 730 may store programs for operation of the processor 740 and may temporarily or permanently store input/output data. The memory 730 may include a flash memory 730 type, a hard disk type, a multimedia card micro type, or a card type memory 730 (for example, SD or XD memory ( 730), etc.), RAM, SRAM, ROM, EEPROM, PROM, magnetic memory 730, magnetic disk, and optical disk.

In addition, the memory 730 may store various network functions and algorithms, and may store various data, programs (one or more instructions), applications, software, commands, codes, etc. for driving and controlling the device.

The processor 740 may control the overall operation of the device. Processor 740 may execute one or more programs stored in memory 730 . The processor 740 may mean a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor 740 on which methods according to the technical idea of the present disclosure are performed.

In an embodiment, the processor 740 generates first image information by acquiring a chest image generated corresponding to the continuous volume of the user's chest and removing a heart adjacent to the lung region and blood vessels around the heart from the chest image; , The lung region including the lung and the pleura is extracted from the first image information to generate the second image information, and the pleural effusion site between the lung and the pleura can be detected from the second image information. Here, the chest image may include a 3D image or a plurality of 2D images. Blood vessels around the heart may also include the aorta, vena cava, pulmonary artery, and pulmonary vein.

In an embodiment, the processor 740 may generate second image information by recognizing the chest and the diaphragm from the first image information and extracting lung regions located inside the chest and the diaphragm from the first image information.

In an embodiment, the processor 740 may recognize the chest and the diaphragm from the first image information, and may generate the second image information by removing the chest and the diaphragm from the first image information.

In an embodiment, the processor 740 may calculate the volume of the pleural effusion site.

In an embodiment, the processor 740 may generate third image information in which the pleural effusion site is labeled with a predetermined identification mark.

In an embodiment, the processor 740 acquires a plurality of first medical images generated corresponding to the continuous volume of the body part including the chest, and extracts an average shape of the target body part from the plurality of first medical images. and receiving a second medical image generated corresponding to the continuous volume of the body part including the chest, and displaying the target body part in the second medical image based on the information about the average shape of the target body part. The shape of the target body part may be created in the second medical image by matching and arranging the average shape and converting coordinates of at least a part of the average shape of the target body part. Here, the information about the average shape of the target body part may include coordinates of the average shape of the target body part and information about a pixel gradient around the average shape of the target body part.

In an embodiment, the processor 740 measures an output value of a loss function related to a pixel gradient around the average shape of the target body part in response to the coordinate transformation of the average shape of the target body part, and based on the output value of the loss function, the processor 740 measures the target body part. By optimizing the coordinates of the average shape of the body part, the shape of the target body part can be generated. Here, the target body part may include at least one of a thorax, a diaphragm, a heart, and blood vessels surrounding the heart. However, it is not limited thereto.

In an embodiment, the processor 740 may generate the shape of the target body part by optimizing the coordinates of the average shape of the target body part so that the output value of the loss function is minimized.

In an embodiment, the processor 740 may normalize the plurality of first medical images based on at least one of a peripheral organ of the at least one target body part and at least one bone part.

In an embodiment, the processor 740 may perform normalization by extracting, enlarging, reducing, or rotating at least a partial region of the first medical image based on at least one of a surrounding organ and a bone region of the target body part.

In an embodiment, the processor 740 may render and display the shape of the target body part together with the second medical image.

A method for detecting pleural effusion according to an embodiment may be implemented in the form of program instructions that can be executed by various computer means and recorded on a computer readable medium. The computer readable medium may include program instructions, data files, data structures, etc. alone or in combination. Program commands recorded on the medium may be specially designed and configured for the present disclosure, or may be known and usable to those skilled in computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic media such as floptical disks. - includes hardware devices specially configured to store and execute program instructions, such as magneto-optical media, and ROM, RAM, flash memory, and the like. Examples of program instructions include high-level language codes that can be executed by a computer using an interpreter, as well as machine language codes such as those produced by a compiler.

In addition, the method for detecting pleural effusion according to the disclosed embodiments may be provided by being included in a computer program product. Computer program products may be traded between sellers and buyers as commodities.

A computer program product may include a S/W program and a computer-readable storage medium in which the S/W program is stored. For example, a computer program product may include a product in the form of a S/W program (eg, a downloadable app) that is distributed electronically through a manufacturer of an electronic device or an electronic marketplace (eg, Google Play Store, App Store). there is. For electronic distribution, at least a part of the S/W program may be stored in a storage medium or temporarily generated. In this case, the storage medium may be a storage medium of a manufacturer's server, an electronic market server, or a relay server temporarily storing SW programs.

A computer program product may include a storage medium of a server or a storage medium of a client device in a system composed of a server and a client device. Alternatively, if there is a third device (eg, a smart phone) that is communicatively connected to the server or the client device, the computer program product may include a storage medium of the third device. Alternatively, the computer program product may include a S/W program itself transmitted from the server to the client device or the third device or from the third device to the client device.

In this case, one of the server, the client device and the third device may execute the computer program product to perform the method according to the disclosed embodiments. Alternatively, two or more of the server, the client device, and the third device may execute the computer program product to implement the method according to the disclosed embodiments in a distributed manner.

For example, a server (eg, a cloud server or an artificial intelligence server) may execute a computer program product stored in the server to control a client device communicatively connected to the server to perform a method according to the disclosed embodiments.

Although the embodiments have been described in detail above, the scope of the present disclosure is not limited thereto, and various modifications and improvements of those skilled in the art using the basic concepts of the present disclosure defined in the following claims are also within the scope of the present disclosure. belongs to

Claims (9)

As a method for detecting pleural effusion,
obtaining a chest image generated corresponding to the continuous volume of the user's chest;
generating first image information by removing the heart and blood vessels surrounding the heart from the chest image;
generating second image information by extracting a lung region including a lung and a pleura from the first image information; and
Detecting a site of pleural effusion between the lung and the pleura from the second image information
Including, method. According to claim 1,
The blood vessels around the heart include an aorta, a vena cava, a pulmonary artery, and a pulmonary vein connected to the heart. According to claim 1,
The generating of the second image information may include recognizing a chest and a diaphragm from the first image information; and extracting the lung region located inside the thorax and the diaphragm from the first image information. According to claim 1,
The generating of the second image information may include recognizing a chest and a diaphragm from the first image information; and removing the ribcage and the diaphragm from the first image information. According to claim 1,
The method further comprising calculating the volume of the pleural effusion site. According to claim 1,
The method further comprising generating third image information in which the pleural effusion site is labeled with a predetermined identification mark. According to claim 1,
The chest image is an image group including a 3D image or a plurality of 2D images. In the device for calculating pleural effusion,
a memory storing a program for detection of pleural effusion; and
By executing the program, a chest image generated corresponding to the continuous volume of the user's chest is acquired, and first image information is generated by removing the heart and blood vessels around the heart from the chest image, and from the first image information An apparatus comprising: at least one processor for generating second image information by extracting a lung region including a lung and a pleura, and detecting a region of pleural effusion between the lung and the pleura from the second image information. A computer program stored in a recording medium to execute the method of any one of claims 1 to 7.

Images