JP7449551B2 - Image analysis device, image analysis method, and image analysis system - Google Patents

Description

The present invention relates to an image analysis device, an image analysis method, and an image analysis system for analyzing captured images of a patient undergoing peritoneal dialysis.

Conventionally, hemodialysis (HD) and peritoneal dialysis (PD) are known as dialysis treatment methods. Hemodialysis requires patients to visit the hospital each time, whereas peritoneal dialysis has the advantage of being able to be performed at home or at work.

When performing peritoneal dialysis, the patient or an assistant such as the patient's family must connect the dialysis device to a tube implanted in the patient's abdomen and operate the dialysis device according to prescribed procedures (for example, opening and closing clamps). There is. Patent Document 1 discloses a device that outputs operating procedures for a dialysis machine using audio and images when a patient performs peritoneal dialysis.

Japanese Patent Application Publication No. 2012-40445

If a patient or an assistant performs peritoneal dialysis using the wrong procedure, the patient may become infected. However, even if the device described in Patent Document 1 outputs a correct peritoneal dialysis procedure, the patient or the assistant does not necessarily perform peritoneal dialysis according to the procedure. Because peritoneal dialysis is performed at the patient's home or workplace, doctors and nurses should check the peritoneal dialysis procedures that the patient or assistant is actually performing and the patient's condition, such as whether he or she has an infectious disease. I can't. Therefore, there is a need for technology that improves the safety of peritoneal dialysis.

The present invention has been made in view of these points, and an object of the present invention is to improve the safety of peritoneal dialysis.

An image analysis device according to a first aspect of the present invention includes an image acquisition unit that acquires a captured image of a patient performing peritoneal dialysis using a dialysis machine or an assistant who assists the patient; a posture identification unit that identifies the posture of the patient or the assistant based on the posture; an analysis unit that analyzes the peritoneal dialysis procedure performed by the patient or the assistant based on the posture; and an analysis by the analysis unit. and an output unit that outputs the results.

The image analysis device further includes a component identification unit that identifies at least one of the position and state of a component included in the dialysis machine based on the captured image, and the analysis unit identifies the posture, the position and state of the component, and It is determined whether or not the procedure conforms to a predetermined standard based on at least one of the states, and the output unit outputs information indicating whether or not the procedure conforms to the predetermined standard as a result of the analysis. You can also output it as

The component specifying unit specifies at least one of the position and state of each of the plurality of components included in the dialysis device, and the analysis unit specifies at least one of the position and the state of each of the plurality of components. determining whether the procedure conforms to the predetermined standard by identifying the order in which the patient or the assistant operated the plurality of parts based on the order, and comparing the specified order with a predetermined reference order; may be determined.

The analysis unit calculates the operation time during which the patient or the assistant operated at least one of the plurality of parts in the specified order, and determines whether the procedure is smooth based on the calculated operation time. The output unit may output information indicating whether the procedure is smooth or not as the analysis result.

If the analysis unit determines that the procedure does not meet the predetermined criteria, the output unit may transmit the captured image to an external device of the image analysis device.

The analysis unit calculates the number of times the patient or the assistant touches an object other than the dialysis machine based on the posture, and evaluates the degree of hygiene of the procedure based on the calculated number of times, The output unit may output information indicating the degree of hygiene as the analysis result.

The image analysis device further includes a mask detection unit that detects a mask worn by the patient or the assistant based on the captured image, and the output unit detects that the mask detection unit detects the mask. Information indicating whether or not it is possible may be output.

The image acquisition section acquires the captured image including the outlet section of the patient's body to which the dialysis machine is connected, and the analysis section evaluates the degree of hygiene of the outlet section based on the captured image. However, the output unit may output information indicating the degree of hygiene as the analysis result.

The analysis unit inputs the captured image to a machine learning model that outputs information indicating the degree of hygiene of the exit section included in the captured image, thereby determining the level of hygiene of the exit section. Information indicating the level of hygiene of the person may be obtained.

The output unit may perform at least one of outputting a predetermined notification and outputting information for stopping the dialysis device when the degree of hygiene of the outlet portion satisfies a predetermined condition. good.

The image analysis method according to the second aspect of the present invention includes the steps of acquiring a captured image of a patient performing peritoneal dialysis using a dialysis machine or an assistant assisting the patient, which is executed by a processor; a step of identifying the posture of the patient or the assistant based on a captured image; a step of analyzing the peritoneal dialysis procedure performed by the patient or the assistant based on the posture; and a step of analyzing. outputting the analysis results.

An image analysis system according to a third aspect of the present invention includes an image acquisition device and an image analysis device capable of communicating with the image acquisition device, and the image acquisition device performs peritoneal dialysis using a dialysis machine. an image acquisition unit that acquires a captured image of a patient or an assistant assisting the patient; a posture identification unit that identifies the posture of the patient or the assistant based on the captured image; a transmitting unit that transmits information to the image analysis device, and the image analysis device includes a receiving unit that receives information indicating the posture from the image acquisition device; It has an analysis section that analyzes the procedure of the peritoneal dialysis that is being performed, and an output section that outputs the analysis result by the analysis section.

The output unit of the image analysis device transmits a request for the captured image to the image acquisition device when receiving a predetermined input from a user or when the analysis result satisfies a predetermined condition, and performs the image acquisition. The transmission unit of the device may transmit the captured image to the image analysis device on condition that the request is received from the image analysis device.

According to the present invention, it is possible to improve the safety of peritoneal dialysis.

FIG. 1 is a schematic diagram of an image analysis system according to a first embodiment. FIG. 1 is a block diagram of an image analysis system according to a first embodiment. FIG. 1 is a schematic diagram of a dialysis device used in peritoneal dialysis. FIG. 3 is a diagram showing a captured image acquired by an image acquisition unit. FIG. 3 is a diagram showing a captured image acquired by an image acquisition unit. FIG. 3 is a diagram showing a captured image acquired by an image acquisition unit. FIG. 2 is a sequence diagram of an image analysis method executed by the image analysis system according to the first embodiment. FIG. 2 is a block diagram of an image analysis system according to a second embodiment. FIG. 7 is a sequence diagram of an image analysis method executed by the image analysis system according to the second embodiment.

(First embodiment)
[Overview of image analysis system S]
FIG. 1 is a schematic diagram of an image analysis system S according to this embodiment. The image analysis system S includes an in-home device 1, a dialysis machine 2, an imaging device 3, a management server 4, and an information terminal 5. The image analysis system S may include other devices such as servers and terminals.

The in-home device 1, the dialysis machine 2, and the imaging device 3 are provided at a location such as a home or workplace where peritoneal dialysis is performed on a patient. The home device 1 is a computer that acquires a captured image captured by the imaging device 3 from the imaging device 3 and analyzes the captured image. That is, the home device 1 functions as an image analysis device. The home device 1 is also referred to as an edge device.

The dialysis device 2 is a device for injecting dialysate into the patient's body and discharging the dialysate from the patient's body in response to operations by the patient or an assistant assisting the patient. An assistant who assists a patient is a person other than the patient (for example, a patient's family, a caregiver, etc.) who performs work for performing peritoneal dialysis. The dialysis device 2 may also include a control device that electrically controls the injection of dialysate into the patient's body and the discharge of dialysate from the patient's body.

More specifically, the dialysis device 2 may be a device that performs CAPD (Continuous Ambulatory Peritoneal Dialysis) in which the dialysate in the patient's body is manually exchanged according to the operation of the patient or an assistant, and The device may also be a device that performs APD (Automated Peritoneal Dialysis) in which dialysate in a patient's body is automatically replaced by a control device. In both cases of CAPD and APD, in order to perform peritoneal dialysis on the patient, the patient or an assistant connects the dialysis device 2 to a tube implanted in the patient's abdomen, and turns the dialysis device 2 on according to the prescribed procedure. It is necessary to perform the second operation (for example, opening/closing a clamp or operating a switch).

In this embodiment, performing peritoneal dialysis by a patient or an assistant is a task for performing peritoneal dialysis on a patient, such as connecting the dialysis device 2 to a tube implanted in the patient's body, and injecting fluid. replacing the dialysis bag and drainage bag, operating the dialysis machine 2 itself, and managing the exit part of the patient's body where the tube is embedded (observation, cleaning, disinfection, application of chemicals, etc.) and other peritoneal dialysis-related tasks (record-keeping, etc.).

The imaging device 3 includes at least one imaging unit (camera) that images a patient performing peritoneal dialysis using the dialysis device 2 or an assistant who assists the patient. The imaging device 3 outputs to the in-home device 1 a captured image of a patient or an assistant undergoing peritoneal dialysis. The range imaged by the imaging device 3 may be the whole body of the patient or the assistant, or may be a part of the body of the patient or the assistant. The imaging device 3 may be integrated with the in-home device 1, or may exchange signals with the in-home device 1 by wire or wirelessly.

The management server 4 is a computer that records information output by the in-home device 1 and outputs information for the user to the information terminal 5. The user is, for example, a medical worker such as a doctor or a nurse. The management server 4 may be connected to multiple in-home devices 1 and multiple information terminals 5.

The information terminal 5 is a computer that receives user operations and displays information output by the management server 4. The information terminal 5 may be integrated with the management server 4, or may exchange signals with the management server 4 by wire or wirelessly. The information terminal 5 has an operation section such as a keyboard and a mouse that can accept operations by a user, and a display section such as a liquid crystal display that can display information.

An overview of the process by which the image analysis system S according to this embodiment analyzes images will be described below. First, the in-home device 1 uses the imaging device 3 to image a patient performing peritoneal dialysis using the dialysis device 2 or an assistant who assists the patient, and obtains a captured image.

The in-home device 1 identifies the posture of the patient or assistant undergoing peritoneal dialysis based on the acquired captured image. The in-home device 1 analyzes the peritoneal dialysis procedure being performed by the patient or the assistant based on the identified posture. For example, as will be described later, the in-home device 1 detects whether the peritoneal dialysis procedure is correct (i.e. (Whether or not the specified criteria are met) is determined. In addition, the in-home device 1 identifies the order in which parts such as clamps (fasteners) of the dialysis machine 2 are operated by the patient or the assistant based on the patient's or assistant's posture, and identifies the specified order. Based on whether the peritoneal dialysis procedure conforms to a predetermined order, it is determined whether the peritoneal dialysis procedure conforms to a predetermined standard.

The home device 1 transmits the analysis result using the captured image to the management server 4. Furthermore, when the in-home device 1 determines that the peritoneal dialysis procedure is incorrect, it transmits the captured image to the management server 4. The management server 4 causes the information terminal 5 to display the analysis results and captured images received from the in-home device 1.

In this way, the image analysis system S according to the present embodiment analyzes the peritoneal dialysis actually performed by the patient or the assistant based on the captured image of the patient undergoing peritoneal dialysis or the assistant assisting the patient. Analyze the procedure and output the analysis results to users such as medical personnel. Therefore, users such as medical professionals can provide appropriate guidance to patients or assistants based on the peritoneal dialysis procedures that patients or assistants are actually performing, and can improve the safety of peritoneal dialysis for patients. can be improved.

[Configuration of image analysis system S]
FIG. 2 is a block diagram of the image analysis system S according to this embodiment. In FIG. 2, arrows indicate main data flows, and there may be data flows that are not shown in FIG. In FIG. 2, each block shows the configuration of a functional unit rather than a hardware (device) unit. As such, the blocks shown in FIG. 2 may be implemented within a single device or may be implemented separately within multiple devices. Data may be exchanged between blocks via any means such as a data bus, a network, or a portable storage medium.

The indoor device 1 includes a control section 11 and a storage section 12. The control unit 11 includes an image acquisition unit 111 , a posture identification unit 112 , a mask detection unit 113 , a component identification unit 114 , an analysis unit 115 , and an output unit 116 . The storage unit 12 includes an image storage unit 121 , an analysis standard storage unit 122 , and an analysis result storage unit 123 .

The storage unit 12 is a storage medium including a ROM (Read Only Memory), a RAM (Random Access Memory), a hard disk drive, and the like. The storage unit 12 stores in advance a program to be executed by the control unit 11. The storage unit 12 may be provided outside the home device 1, and in that case, data may be exchanged with the control unit 11 via a network. The image storage unit 121 stores captured images captured by the imaging device 3. The analysis standard storage unit 122 stores information used for analysis, such as reference values for performing analysis and machine learning models. The analysis result storage unit 123 stores analysis results using captured images. The image storage section 121, the analysis standard storage section 122, and the analysis result storage section 123 may each be a storage area on the storage section 12, or may be a database configured on the storage section 12.

The control unit 11 is, for example, a processor such as a CPU (Central Processing Unit), and executes a program stored in the storage unit 12 to control the image acquisition unit 111, posture identification unit 112, mask detection unit 113, and component identification unit. 114, functions as an analysis section 115 and an output section 116. At least part of the functions of the control unit 11 may be performed by an electric circuit. Further, at least part of the functions of the control unit 11 may be realized by the control unit 11 executing a program executed via a network.

The management server 4 includes a control section 41 and a storage section 42. The control section 41 includes a receiving section 411, an analyzing section 412, and an output section 413. The storage unit 42 includes an image storage unit 421, an analysis standard storage unit 422, and an analysis result storage unit 423.

The storage unit 42 is a storage medium including ROM, RAM, hard disk drive, and the like. The storage unit 42 stores in advance a program to be executed by the control unit 41. The storage unit 42 may be provided outside the management server 4, and in that case, data may be exchanged with the control unit 41 via a network. The image storage unit 421 stores captured images captured by the imaging device 3. The analysis standard storage unit 422 stores information used for analysis, such as reference values for performing analysis and machine learning models. The analysis result storage unit 423 stores analysis results using captured images. The image storage section 421, the analysis standard storage section 422, and the analysis result storage section 423 may each be a storage area on the storage section 42, or may be a database configured on the storage section 42.

The control unit 41 is, for example, a processor such as a CPU, and functions as a reception unit 411, an analysis unit 412, and an output unit 413 by executing a program stored in the storage unit 42. At least part of the functions of the control unit 41 may be performed by an electric circuit. Furthermore, at least part of the functions of the control unit 41 may be executed by a program executed via a network.

The image analysis system S according to this embodiment is not limited to the specific configuration shown in FIG. 2. The in-home device 1 and the management server 4 are not each limited to one device, but may be configured by two or more physically separate devices connected by wire or wirelessly.

[Configuration of dialysis device 2]
FIG. 3 is a schematic diagram of a dialysis device 2 used in peritoneal dialysis. The dialysis device 2 includes a tube implanted in the patient's abdomen, a fluid injection bag B1 that holds the dialysate before it is injected into the patient's body, and a drain that holds the dialysate after it has been discharged from the patient's body. This is a device that connects the fluid bag B2 and controls the transport of dialysate between them.

The dialysis device 2 connects a tube embedded in the patient's abdomen, a liquid injection bag B1, and a drainage bag B2 through a pipe 21 (tube). The tube 21 includes a patient side connector 22 for connecting to a tube implanted in the patient's abdomen, a liquid injection side connector 23 for connecting to the liquid injection bag B1, and a liquid injection side connector 23 for connecting to the liquid drainage bag B2. It has a drain side connector 24. The tube 21 is branched into three directions by a branching portion 25 into a tube implanted in the patient's abdomen, a liquid injection bag B1, and a drainage bag B2.

A plurality of clamps 26 (fasteners) are provided in the middle of the pipe 21, which can be switched between an open state and a closed state according to a user's operation. The open clamp 26 allows the passage of dialysate through the tube 21 . Clamp 26 in the closed state prevents passage of dialysate through tube 21 . Since the shape of the clamp 26 changes between the open state and the closed state, the in-home device 1 can specify whether the clamp 26 is in the open state or the closed state based on the appearance in the captured image.

In the example of FIG. 3, the plurality of clamps 26 include a patient side clamp 26a, a liquid injection side clamp 26b, and a liquid drainage side clamp 26c. A patient-side clamp 26a is provided between the patient-side connector 22 and the branch portion 25. A liquid injection side clamp 26b is provided between the liquid injection side connector 23 and the branch portion 25. A drain side clamp 26c is provided between the drain side connector 24 and the branch portion 25. The plurality of clamps 26 is not limited to the example shown in FIG. 3, and may be of other numbers or may be provided at other positions.

Since the plurality of clamps 26 have mutually different colors or patterns, the in-home device 1 can identify each of the plurality of clamps 26 based on the appearance in the captured image. Further, by forming the plurality of clamps 26 in mutually different shapes, the in-home device 1 may be able to identify each of the plurality of clamps 26 based on the appearance in the captured image.

The dialysis device 2 is not limited to the configuration shown in FIG. 3, and may have other configurations. For example, the dialysis device 2 may include a control device that electrically controls the transport of dialysate. In this case, the control device is connected via tube 21 to a tube implanted in the patient's abdomen and electrically controls the infusion and drainage of dialysate. A plurality of clamps 26 are provided in the middle of the tube 21 that connects the control device and the tube implanted in the patient's abdomen, and the patient or an assistant can tighten the plurality of clamps 26 before and after starting control by the control device. Manipulate.

[Explanation of image analysis method]
Hereinafter, the image analysis method executed by the image analysis system S according to this embodiment will be described in detail. First, the in-home device 1 uses the imaging device 3 to start imaging the patient or the assistant. At this time, the in-home device 1 starts imaging when the patient or the assistant performs a predetermined operation (for example, pressing a start button provided on the in-home device 1 or the dialysis machine 2). Alternatively, the in-home device 1 may automatically start imaging when a predetermined condition (for example, the control device included in the dialysis machine 2 is turned on). In addition, when the patient or the assistant performs a predetermined operation (for example, pressing an end button provided on the in-home device 1 or the dialysis machine 2), or under a predetermined condition (for example, the control included in the dialysis machine 2), the in-home device 1 is activated. Imaging ends when the conditions (that the device is powered off) are satisfied. The in-home device 1 may collectively analyze the captured images from the start to the end of the imaging, or may sequentially analyze the captured images at predetermined time intervals after the imaging starts.

In the in-home device 1 , the image acquisition unit 111 acquires a captured image of a patient or an assistant performing peritoneal dialysis using the dialysis machine 2 captured by the imaging device 3 and stores it in the image storage unit 121 . . The captured image may be a moving image or a plurality of still images captured at predetermined time intervals. The captured image may be one image that captures one imaging range, or may be a set of multiple images that capture multiple different imaging ranges at the same time.

FIG. 4 is a diagram showing a captured image IM acquired by the image acquisition unit 111. FIG. 4 shows a region including a patient in the captured image IM. The posture identifying unit 112 identifies the patient's posture P based on the captured image acquired by the image acquiring unit 111. The posture P identified by the posture identifying section 112 may be the posture of the patient's whole body, or may be the position of a specific part of the patient. For example, the posture identifying unit 112 calculates the coordinates of a plurality of joint points of the human included in the captured image IM using the posture estimation model, and converts the set of coordinates of the calculated joint points into the patient's posture. Identify as P. The posture identifying unit 112 is not limited to the method shown here, and can use a known posture estimation method that can estimate the posture P of a human being. The posture specifying unit 112 outputs posture information indicating the specified posture P of the patient (for example, coordinates of a plurality of joint points of the patient). Additionally, the posture identifying unit 112 may identify the assistant's posture based on the captured image using a similar method.

The mask detection unit 113 detects the mask M worn by the patient or the assistant based on the captured image acquired by the image acquisition unit 111. For example, the mask detection unit 113 detects the mask M included in the captured image IM using an object recognition model. The mask detection unit 113 is not limited to the method shown here, and can use any known object detection method that can detect the mask M. The mask detection unit 113 outputs mask detection information indicating whether or not the mask M worn by the patient or the assistant has been detected (for example, the coordinates of the mask M). In this way, the mask detection unit 113 determines whether the patient or assistant is performing peritoneal dialysis in a sanitary condition by outputting information indicating whether the patient or assistant is wearing the mask M. can be made possible.

FIG. 5 is a diagram showing a captured image IM acquired by the image acquisition unit 111. FIG. 5 shows an area including a plurality of clamps 26 (patient-side clamp 26a, liquid injection-side clamp 26b, and drain-side clamp 26c) of the dialysis machine 2 in the captured image IM. The component identification unit 114 identifies the positions and states of each of the plurality of clamps 26, which are the plurality of components included in the dialysis machine 2, based on the captured image acquired by the image acquisition unit 111. For example, the component specifying unit 114 uses the object recognition model to specify the position of each of the plurality of clamps 26 included in the captured image IM, and whether each of the plurality of clamps 26 is in an open state or a closed state. to identify. As described above, based on the colors and shapes of the plurality of clamps 26, the positions and states of each of the plurality of clamps 26 can be specified from the captured image IM. The component identification unit 114 is not limited to the method shown here, and may use any known object detection method that can identify the state of the clamp 26. The component specifying unit 114 outputs component information indicating the position and state of each of the plurality of identified clamps 26. The component specifying unit 114 may output component information indicating the position and state of not only the clamp 26 but other components included in the dialysis machine 2.

The identification of the posture P by the posture identification section 112, the detection of the mask M by the mask detection section 113, and the identification of the clamp 26 by the component identification section 114 may be performed in parallel or in any order. .

The analysis unit 115 analyzes the peritoneal dialysis procedure performed by the patient or the assistant based on the patient's posture P or the assistant's posture indicated by the posture information output by the posture identifying unit 112. The analysis of the peritoneal dialysis procedure by the analysis unit 115 includes analysis to determine whether the patient or assistant is engaging in unsanitary behavior during peritoneal dialysis, and whether the patient or assistant is operating the plurality of clamps 26 in the correct order. including analysis. Specifically, the analysis unit 115 first uses the time stamp of the captured image used by the posture identification unit 112 to identify the posture P, and the captured image used by the component identification unit 114 to identify the state of the clamp 26. By associating the time stamp with the time stamp, the times of the posture information outputted by the posture specifying section 112 and the parts information outputted by the component specifying section 114 are synchronized.

Next, when the posture information includes postures P of a plurality of people, the analysis unit 115 selects one of the postures P as the posture P of the patient. For example, the analysis unit 115 selects the patient's posture P based on past posture information. In this case, the analysis unit 115 selects the patient's posture P that is closest to the patient's posture P that was identified at the previous timing in the time series, among the plurality of human postures P that were identified at a certain timing. Select as P. As another method, the analysis unit 115 may select the posture P closest to the clamp 26 as the patient posture P from among the plurality of human postures P specified at a certain timing.

Next, the analysis unit 115 determines whether or not the dialysis device 2 is an unsanitary act performed by the patient or the assistant during peritoneal dialysis based on the patient's posture P or the assistant's posture indicated by the posture information output by the posture identifying unit 112. Calculate the number of times the object is touched. For example, the analysis unit 115 calculates the number of times the patient touches his or her own head. The head includes the patient's face and hair. At this time, the analysis unit 115 calculates the distance between the position of the hand and the position of the head in the patient's posture P, and when the distance is less than or equal to a predetermined value, the analysis unit 115 calculates the distance by adding 1 to the number of times the patient touches the head. Calculate the number of times the patient touched the head by adding . The analysis unit 115 may calculate the number of times the patient or an assistant touches not only the patient's head but also other objects that should not be touched during peritoneal dialysis. The analysis unit 115 then evaluates the degree of hygiene of the peritoneal dialysis procedure based on the calculated number of times.

For example, the analysis unit 115 calculates a lower score indicating the degree of hygiene as the number of times calculated is larger, and calculates a higher score indicating the degree of hygiene as the number of times calculated is smaller. The analysis unit 115 stores the number of times a patient or an assistant touches an object other than the dialysis machine 2 and information (score) indicating the degree of hygiene evaluated based on the number of times as an analysis result in the analysis result storage unit 123. Make me remember. In this way, the analysis unit 115 calculates the number of times a patient or an assistant touches an object other than the dialysis equipment 2, thereby allowing users such as medical workers to detect whether a patient or an assistant is engaging in unsanitary behavior during peritoneal dialysis. be able to understand the situation and provide appropriate guidance.

Next, the analysis unit 115 analyzes the patient's posture P or the assistant's posture indicated by the posture information output by the posture identifying unit 112, and the positions and states of each of the plurality of parts indicated by the component information output by the component identifying unit 114. Based on this, it is determined whether the patient or the assistant has operated the component. Below, the process performed by the analysis unit 115 will be explained using the clamp 26 as a component included in the dialysis device 2.

The analysis unit 115 determines at least one of the patient's posture P or the assistant's posture indicated by the posture information output by the posture specifying unit 112, and the position and state of each of the plurality of clamps 26 indicated by the component information output by the component specifying unit 114. Based on this, it is determined whether the patient or the assistant has operated the clamp 26. At this time, the analysis unit 115 calculates the distance between the hand position in the patient's posture P or the assistant's posture and the position of each of the plurality of clamps 26, and when the distance is less than a predetermined value, the Alternatively, it is determined that the assistant has touched the clamp 26 (that is, has operated the clamp 26).

When the analysis unit 115 determines that the patient or assistant has touched the clamp 26, the analysis unit 115 provides identification information of the clamp 26 touched by the patient or assistant (that is, which of the plurality of clamps 26 the patient or assistant touched). , the time when the patient or assistant touches the clamp 26 and the state of the clamp 26 at the time the patient or assistant touches it. The state of the clamp 26 is information indicating, for example, whether the clamp 26 is in an open state or a closed state. The analysis unit 115 identifies the order in which the patient or the assistant operated the plurality of clamps 26 by repeating the determination of the patient's operation of the clamps 26 in chronological order on the captured images IM from the start to the end of peritoneal dialysis. .

Alternatively, the analysis unit 115 may identify the order in which the patient or assistant operated the plurality of clamps 26 based on the posture of the patient or assistant and the position of the clamps 26, without using the state of the clamps 26. . In this case, the analysis unit 115 associates the identification information of the clamp 26 touched by the patient or assistant with the time when the patient or assistant touched the clamp 26, so that the patient or assistant can Identify the order in which 26 was operated. With such a configuration, even if the analysis unit 115 does not identify a change in the state of a component of the dialysis device 2, the analysis unit 115 can operate the component based on the relationship between the patient's or assistant's posture and the location of the component. The order can be estimated.

The analysis unit 115 compares the specified order with a predetermined standard order stored in the analysis standard storage unit 122 in advance. The standard order determines in what order and how the plurality of clamps 26 included in the dialysis device 2 are operated (for example, whether the clamps 26 are switched between open and closed states, or whether the clamps 26 are touched or not). This information indicates whether the The analysis unit 115 determines that the peritoneal dialysis procedure being performed by the patient or the assistant is incorrect when the specified order does not match the standard order (that is, when the predetermined standard is not met), If not, it is determined that the peritoneal dialysis procedure being performed by the patient or the assistant is correct. In this way, the analysis unit 115 analyzes the peritoneal dialysis procedure based on the operation order of the plurality of clamps 26, and thereby determines with high precision whether the peritoneal dialysis procedure is correct (that is, whether or not it meets a predetermined standard). ) can be determined.

The analysis unit 115 specifies the order in which the patient or the assistant operates not only the clamp 26 but also other parts of the dialysis machine 2 (for example, the infusion bag B1 and the drainage bag B2), and performs the operations in the specified order. Based on this, it may be determined whether the peritoneal dialysis procedure meets predetermined criteria. In this case, the analysis unit 115 determines the order in which the patient or assistant operated the parts of the dialysis machine 2 based on the patient's or assistant's posture and at least one of the positions and states of the parts of the dialysis machine 2. Identify.

The analysis unit 115 analyzes the order in which the patient or the assistant operated the plurality of parts (clamps 26) and the determination as to whether or not the peritoneal dialysis procedure based on the order conforms to a predetermined standard. The analysis results are stored in the analysis result storage unit 123.

FIGS. 6A and 6B are diagrams showing captured images IM acquired by the image acquisition unit 111. FIGS. 6(a) and 6(b) represent a region of the patient's body including the outlet to which the dialysis machine 2 is connected in the captured image IM. The outlet portion is a portion in the abdomen of the patient's body in which a tube (catheter) connected to the dialysis device 2 is embedded in advance. By connecting the outlet tube to the tube 21 of the dialysis device 2, the inside of the patient's body (intraperitoneal cavity) and the tube 21 of the dialysis device 2 communicate with each other, making it possible to inject and discharge dialysate.

If the area around the exit is unsanitary, bacterial infection may occur. If peritoneal dialysis is continued with an infection occurring at the exit site, bacteria and the like may invade the patient's body, leading to the occurrence of more serious infections such as peritonitis. FIG. 6(a) shows a state in which no infection has occurred at the outlet, and FIG. 6(b) shows a state in which infection has occurred at the outlet. As shown in FIGS. 6A and 6B, there is a correlation between the appearance of the exit section included in the captured image IM and the degree of hygiene of the exit section. Therefore, the analysis unit 115 evaluates the degree of hygiene of the exit section based on the captured image IM.

The analysis standard storage unit 122 stores in advance information indicating a machine learning model that inputs an image and outputs information (score) indicating the degree of hygiene of the exit section included in the image. The machine learning model is generated by learning in advance an image and a score included in the image indicating the degree of hygiene of the exit section using a known machine learning method. The degree of hygiene score is a value determined depending on, for example, the size of swelling, the size of eschar, the size of redness, the intensity of pain, and the type of exudate. In this embodiment, the score is determined such that the higher the score, the worse the sanitary condition. Scores are not limited to the definitions shown here.

The analysis unit 115 obtains a score indicating the degree of hygiene of the exit section by inputting the captured image IM into a machine learning model stored in advance in the analysis standard storage unit 122. The analysis unit 115 causes the analysis result storage unit 123 to store the obtained score indicating the degree of hygiene of the exit section as an analysis result. Although it takes a lot of effort for a user such as a medical worker to judge the level of hygiene at the exit every time from the captured image IM, the analysis unit 115 automatically determines the level of hygiene at the exit by using a machine learning model. can be evaluated.

In this embodiment, the analysis unit 115 uses a captured image IM captured by the imaging device 3 installed in the house, but an image captured using an imaging unit of a communication terminal (such as a smartphone) owned by a patient or an assistant. IM may also be used. In this case, the patient or the assistant images the exit portion using the imaging unit of the communication terminal at any one of the start, middle, and end of peritoneal dialysis. The image acquisition unit 111 of the in-home device 1 acquires a captured image IM including the exit portion captured by the communication terminal through communication. The analysis unit 115 then uses the captured image IM acquired from the communication terminal to acquire a score indicating the degree of hygiene at the exit.

Furthermore, the analysis unit 115 may evaluate the degree of hygiene based on the environment around the patient and the dialysis machine 2 included in the captured image IM. For example, the analysis unit 115 uses a known object detection method to detect a predetermined object (food or garbage) that causes unsanitary conditions around the patient and the dialysis machine 2 in the captured image IM. Then, the analysis unit 115 calculates information (score) indicating the degree of hygiene based on the number and type of detected objects in addition to the state of the exit included in the captured image IM. Thereby, the analysis unit 115 can evaluate the degree of hygiene in consideration of the environment around the patient and the dialysis machine 2.

Furthermore, the analysis unit 115 may evaluate the level of hygiene based on at least one of the patient's biological information, the patient's medication information, and the drainage information of the dialysate discharged from the patient's body. In this case, the analysis unit 115 acquires biological information such as body temperature, blood pressure, and pulse rate from a sensor provided on the patient's body. The analysis unit 115 also acquires patient medication information (medication name, etc.) stored in the storage unit 12 in advance. The analysis unit 115 also acquires drainage information such as the color and components of the dialysate after it has been discharged from the patient's body, from a sensor provided in the dialysis device 2 .

The analysis unit 115 then calculates information (score) indicating the level of hygiene based on the acquired biological information, medication information, and drainage information in addition to the state of the outlet included in the captured image IM. . Specifically, the analysis unit 115 adds a predetermined value to the score when the biological information indicates that the patient's condition is different from that of a normal patient (for example, the patient has a fever). This is because if the patient is not in a normal condition, there is a risk that the patient may have an infection.

Further, the analysis unit 115 adds a predetermined value to the score when the medication information indicates administration of a predetermined drug (for example, a prescription drug for diabetes). This is because patients with diabetes and the like have a high risk of infection. In addition, the analysis unit 115 adds a predetermined value to the score when the drainage information indicates that the state of the drainage fluid is different from the normal drainage fluid state (for example, the drainage fluid is cloudy). This is because if the drainage is not normal, the patient may have an infection.

In this way, the analysis unit 115 calculates a score indicating the degree of hygiene based on biological information, medication information, and drainage information in addition to the captured image IM including the exit portion, thereby achieving more accurate hygiene. Can evaluate the degree.

The output unit 116 outputs the posture information output by the posture identification unit 112, the mask detection information outputted by the mask detection unit 113, the component information output by the component identification unit 114, and the analysis result by the analysis unit 115 to the management server. Output by sending to 4. The analysis results include the number of times the patient or assistant touched an object other than the dialysis device 2, information indicating the degree of hygiene evaluated based on the number of times, and the order in which the patient or assistant operated multiple parts (clamps 26). , a determination of whether the peritoneal dialysis procedure based on the sequence meets predetermined criteria, and a score indicating the degree of outlet hygiene. Only part of the analysis results may be sent to the management server 4. The output unit 116 is not limited to sending the analysis results to the management server 4, but may output the analysis results by other methods such as recording data in a storage device, printing on paper using a printer, or outputting audio using a speaker. It's okay.

Furthermore, the output unit 116 determines whether or not a peritoneal dialysis procedure based on the order in which a patient or an assistant operates a plurality of parts (clamps 26) conforms to a predetermined standard. (that the procedure is incorrect), the captured image IM is transmitted to the management server 4. As a result, the output unit 116 transmits the captured image to the outside of the in-home device 1 only when it is determined that there is an abnormality in the peritoneal dialysis procedure, so the amount of data for communication performed by the in-home device 1 can be reduced. can.

Further, the output unit 116 may output a notification to at least one of the patient and the user when the score indicating the degree of hygiene of the exit section satisfies a predetermined condition stored in advance in the analysis criteria storage unit 122. The condition for outputting the notification is, for example, that the score is greater than or equal to a predetermined value. The notification includes the fact that the sanitary conditions at the exit section are poor, and is sent to a communication terminal (such as a smartphone) owned by the patient or an assistant, or to the information terminal 5 used by the user, and is output by screen display or audio. Further, when the score indicating the degree of hygiene of the outlet part satisfies a predetermined condition, and when the dialysis device 2 includes a control device that electrically controls the transport of dialysate, the output section 116 Information (signal) for stopping 2 may be output. The condition for stopping the dialysis machine 2 is, for example, that the score is greater than or equal to a predetermined value. Thereby, the output unit 116 can provide feedback to the patient or user based on the score indicating the degree of hygiene of the exit section.

The conditions for outputting the notification and the conditions for stopping the dialysis machine 2 may be different from each other. As a result, the output unit 116 outputs a notification when the level of hygiene satisfies the first condition, and forces the dialysis machine 2 to operate when the level of hygiene satisfies the second condition different from the first condition. can be stopped.

In the management server 4, the receiving unit 411 receives the posture information, mask detection information, parts information, and analysis results transmitted from the in-home device 1, and stores them in the analysis result storage unit 423. Further, the receiving unit 411 receives the captured image IM transmitted by the in-home device 1, and stores it in the image storage unit 421.

The analysis unit 412 acquires the posture information, mask detection information, parts information, and analysis results transmitted by the in-home device 1. The analysis unit 412 then performs additional analysis on the acquired posture information, mask detection information, parts information, and analysis results. For example, the analysis unit 412 calculates the operation time of each of the plurality of steps included in the order in which the patient or the assistant operated the plurality of parts (clamps 26) of the dialysis machine 2, which is included in the analysis result. Specifically, the analysis unit 412 analyzes the order in which the patient or the assistant operates the plurality of clamps 26 after the patient or the assistant starts operating at least one of the plurality of clamps 26 (for example, after the patient or the assistant starts operating the plurality of clamps 26). The time from when the clamp 26 is touched until the end of the operation of the clamp 26 (for example, until the clamp 26 is released) is calculated as the operation time of one step. The analysis unit 412 calculates each of the plurality of steps included in the order in which the plurality of clamps 26 are operated by the patient or the assistant by repeating the calculation of the operation time for the captured images IM from the start to the end of peritoneal dialysis in chronological order. Calculate the operation time.

The analysis unit 412 determines whether the peritoneal dialysis procedure is smooth based on the calculated operation time of each of the plurality of steps, and outputs information indicating whether the peritoneal dialysis procedure is smooth or not as an analysis result. For example, the analysis unit 412 determines that the peritoneal dialysis procedure is not smooth if the operation time of at least one of the calculated steps is longer than a predetermined reference time, and if not, the analysis unit 412 determines that the peritoneal dialysis procedure is not smooth. It is judged to be smooth. By calculating the operation time of each of the plurality of steps in this way, the analysis unit 412 allows a user such as a medical worker to understand which step the patient or assistant is confused about, and can provide appropriate guidance. . The analysis unit 412 is not limited to the analysis shown here, and may perform other analyses.

In the image analysis system S, it is assumed that the home device 1 (edge device) has a low computing power because it is provided in each home, and the management server 4 has a higher computing power than the home device 1. In the case of such a configuration, the load on the home device 1 can be suppressed by the analysis unit 412 of the management server 4 performing additional analysis that is not performed on the home device 1 . On the other hand, the management server 4 may omit the analysis by the analysis unit 412 and use the analysis result sent by the home device 1 as is.

The output unit 413 outputs the analysis results transmitted by the in-home device 1 and the analysis results output by the analysis unit 412 by transmitting them to the information terminal 5. The information terminal 5 outputs the analysis results sent by the management server 4 through screen display or audio. Further, the output unit 413 may transmit the mask detection information transmitted by the in-home device 1 to the information terminal 5. Further, the output unit 413 may transmit the captured image IM transmitted by the in-home device 1 to the information terminal 5.

[Sequence of image analysis method]
FIG. 7 is a sequence diagram of an image analysis method executed by the image analysis system S according to this embodiment. First, the in-home device 1 uses the imaging device 3 to start imaging the patient or the assistant. The image acquisition unit 111 acquires an image captured by the imaging device 3 of a patient performing peritoneal dialysis using the dialysis machine 2 or an assistant assisting the patient (S11).

The posture identifying unit 112 identifies the patient's posture P or the assistant's posture based on the captured image acquired by the image acquiring unit 111 (S12). The mask detection unit 113 detects the mask M worn by the patient or the assistant based on the captured image acquired by the image acquisition unit 111 (S13). The component identifying unit 114 identifies the positions and states of each of the plurality of components (clamps 26) based on the captured image acquired by the image acquiring unit 111 (S14).

The analysis unit 115 calculates the number of times the patient or the assistant touches an object other than the dialysis machine 2 based on the patient's posture P or the assistant's posture indicated by the posture information output by the posture identifying unit 112 (S15). . The analysis unit 115 determines at least one of the patient's posture P or the assistant's posture indicated by the posture information output by the posture identifying unit 112, and the position and state of each of the plurality of components indicated by the component information output by the component identifying unit 114. Based on this, it is determined whether the patient or the assistant has operated the part. The analysis unit 115 identifies the order in which the patient or the assistant operated the plurality of parts by repeating the determination of the patient's manipulation of the parts in chronological order (S16).

The analysis unit 115 calculates information (score) indicating the degree of hygiene based on the calculated number of times. In addition, the analysis unit 115 determines that the peritoneal dialysis procedure performed by the patient or the assistant is incorrect if the specified order does not match the standard order (that is, does not meet the predetermined criteria). However, if not, it is determined that the peritoneal dialysis procedure being performed by the patient or the assistant is correct (S17).

The analysis unit 115 evaluates the degree of hygiene of the exit section based on the captured image IM (S18). The output unit 116 outputs the posture information output by the posture identification unit 112, the mask detection information outputted by the mask detection unit 113, the component information output by the component identification unit 114, and the analysis result by the analysis unit 115 to the management server. 4 (S19). At this time, the output unit 116 determines whether or not the peritoneal dialysis procedure conforms to a predetermined standard based on the order in which the patient or the assistant operated the plurality of parts. Incorrect), the captured image IM may be transmitted to the management server 4.

In the management server 4, the receiving unit 411 receives the posture information, mask detection information, parts information, and analysis results transmitted by the in-home device 1. The analysis unit 412 calculates the operation time of each of the plurality of steps included in the order in which the patient or the assistant operated the plurality of parts, which is included in the analysis result (S20). The analysis unit 412 outputs the calculated operation time of each of the plurality of steps as an analysis result.

The output unit 413 outputs the analysis results transmitted by the in-home device 1 and the analysis results output by the analysis unit 412 by transmitting them to the information terminal 5 (S21). The information terminal 5 outputs the analysis results sent by the management server 4 through screen display or audio.

[Effects of embodiment]
The image analysis system S according to the present embodiment identifies the patient's posture or the assistant's posture from a captured image of a patient undergoing peritoneal dialysis or an assistant who assists the patient, and based on the identified posture, Alternatively, the peritoneal dialysis procedure actually performed by an assistant is analyzed, and the analysis results are output to a user such as a medical worker. Therefore, users such as medical professionals can provide appropriate guidance to patients based on the peritoneal dialysis procedures actually performed by patients or assistants, and can improve the safety of peritoneal dialysis for patients. .

(Second embodiment)
In the first embodiment, the in-home device 1 mainly performs the analysis, whereas in the present embodiment, the management server 4 mainly performs the analysis. Hereinafter, a configuration different from the first embodiment will be explained.

FIG. 8 is a block diagram of the image analysis system S according to this embodiment. The in-home device 1 according to this embodiment does not have the analysis section 115, unlike the configuration shown in FIG. Instead, the analysis unit 412 of the management server 4 performs the analysis performed by the analysis unit 115 in the first embodiment. That is, in this embodiment, the home device 1 functions as an image acquisition device, and the management server 4 functions as an image analysis device.

Specifically, in the home device 1, the output unit 116 outputs the posture information outputted by the posture identification unit 112, the mask detection information outputted by the mask detection unit 113, and the parts information outputted by the component identification unit 114. It is output by transmitting it to the management server 4 as information used for analysis. Unlike the first embodiment, the output unit 116 does not transmit the analysis results. The output unit 116 may transmit the posture information output by the posture identifying unit 112 and the component information output by the component identifying unit 114 after synchronizing the times.

In the management server 4, the receiving unit 411 receives the posture information, mask detection information, and parts information transmitted by the in-home device 1 as information used for analysis. Similar to the analysis unit 115 of the first embodiment, the analysis unit 412 calculates the number of times the patient or assistant touches an object other than the dialysis machine 2 based on the posture information transmitted by the in-home device 1, and Evaluate the degree of hygiene of the peritoneal dialysis procedure based on the number of times performed. The analysis unit 412 causes the analysis result storage unit 423 to store, as an analysis result, the number of times a patient or an assistant touches an object other than the dialysis machine 2 and information indicating the degree of hygiene evaluated based on the number of times.

Further, similar to the analysis unit 115 of the first embodiment, the analysis unit 412 analyzes whether the patient or the assistant has operated a plurality of parts (clamps 26) based on the posture information and parts information transmitted by the in-home device 1. The order is specified, and based on the specified order, it is determined whether the peritoneal dialysis procedure is correct (that is, whether or not it meets a predetermined standard). The analysis unit 412 stores the order in which a patient or an assistant operated a plurality of parts and the determination as to whether the peritoneal dialysis procedure based on the order conforms to a predetermined standard as an analysis result in an analysis result storage unit. 423.

Furthermore, similarly to the first embodiment, the analysis unit 412 calculates the operation time of each of the plurality of steps included in the order in which the patient or the assistant operated the plurality of parts. The analysis unit 412 stores the calculated operation time of each of the plurality of steps in the analysis result storage unit 423 as an analysis result.

The output unit 413 outputs the analysis result by the analysis unit 412 by transmitting it to the information terminal 5. The information terminal 5 outputs the analysis results sent by the management server 4 through screen display or audio. Further, the output unit 413 may transmit the mask detection information transmitted by the in-home device 1 to the information terminal 5. In this embodiment, since the in-home device 1 does not perform analysis, the computing power of the in-home device 1 can be configured to be low, and the cost for introducing the in-home device 1 into the patient's home can be reduced.

The analysis unit 412 may evaluate the degree of hygiene of the exit section based on the captured image IM. In this case, the output unit 116 of the home device 1 transmits the captured image IM to the management server 4 in addition to the posture information, mask detection information, and parts information. In the management server 4, the receiving unit 411 receives the captured image IM transmitted by the in-home device 1, and stores it in the image storage unit 421. Similar to the analysis unit 115 of the first embodiment, the analysis unit 412 evaluates the degree of hygiene of the exit portion based on the captured image IM. The analysis unit 412 causes the analysis result storage unit 423 to store the obtained score indicating the degree of hygiene of the exit section as an analysis result. Thereby, the management server 4 instead of the in-home device 1 can evaluate the degree of hygiene at the exit section.

The information terminal 5 may receive an input for requesting a captured image IM from a user such as a medical worker who has referred to the analysis result. When the information terminal 5 receives an input from the user to request the captured image IM, the output unit 413 transmits the request for the captured image IM to the in-home device 1 . In addition, the output unit 413 can determine whether or not the peritoneal dialysis procedure conforms to a predetermined standard based on the order in which the patient or assistant operated a plurality of parts, based on a predetermined condition (for example, the peritoneal dialysis procedure Incorrect), a request for the captured image IM may be automatically transmitted to the in-home device 1.

In the in-home device 1, the output unit 116, on the condition that the request for the captured image IM is received from the management server 4, outputs the next peritoneal dialysis (that is, the first peritoneal dialysis after the output unit 116 receives the request for the captured image IM). A captured image IM captured during dialysis is transmitted to the management server 4. In the management server 4, the receiving unit 411 receives the captured image IM transmitted by the in-home device 1, and stores it in the image storage unit 421. The output unit 413 transmits the captured image IM transmitted by the in-home device 1 to the information terminal 5. With such a configuration, even if the home device 1 has not recorded the captured image IM, it can transmit the next captured image IM to the management server 4 in response to a user's request, and the user can record the captured image IM. It can be made verifiable.

[Sequence of image analysis method]
FIG. 9 is a sequence diagram of an image analysis method executed by the image analysis system S according to this embodiment. Steps S11 to S14 are similar to those in FIG. The output unit 116 transmits the posture information outputted by the posture identification unit 112, the mask detection information outputted by the mask detection unit 113, and the parts information outputted by the component identification unit 114 to the management server 4 as information used for analysis. The data is output by doing so (S15').

In the management server 4, the receiving unit 411 receives the posture information, mask detection information, and parts information transmitted by the in-home device 1 as information used for analysis. The analysis unit 412 calculates the number of times the patient or assistant touches an object other than the dialysis device 2 based on the posture information transmitted by the in-home device 1, and determines the level of hygiene of the peritoneal dialysis procedure based on the calculated number of times. is evaluated (S16'). The analysis unit 412 identifies the order in which the patient or the assistant operated the plurality of parts (clamps 26) of the dialysis machine 2 based on the posture information and parts information transmitted by the in-home device 1 (S17').

The analysis unit 412 calculates information (score) indicating the degree of hygiene based on the calculated number of times. In addition, the analysis unit 412 determines that the peritoneal dialysis procedure performed by the patient or assistant is incorrect if the specified order does not match the standard order, and if not, the analysis unit 412 determines that the peritoneal dialysis procedure performed by the patient or assistant is incorrect. It is determined that the peritoneal dialysis procedure being performed is correct (S18').

The analysis unit 412 evaluates the level of hygiene at the exit based on the captured image IM transmitted by the in-home device 1 (S19'). The analysis unit 412 calculates the operation time of each of the plurality of steps included in the order in which the patient or the assistant operated the plurality of parts (S20').

The output unit 413 outputs the analysis result by the analysis unit 412 by transmitting it to the information terminal 5 (S21'). The information terminal 5 outputs the analysis results sent by the management server 4 through screen display or audio.

[Effects of embodiment]
According to the image analysis system S according to the present embodiment, similarly to the first embodiment, a user such as a medical worker is able to respond to a patient based on the peritoneal dialysis procedure actually performed by the patient or an assistant. Appropriate guidance can be given to patients, improving the safety of peritoneal dialysis for patients. Furthermore, since the management server 4 rather than the home device 1 mainly performs the analysis, the cost for introducing the home device 1 can be reduced.

Although the present invention has been described above using the embodiments, the technical scope of the present invention is not limited to the scope described in the above embodiments, and various modifications and changes can be made within the scope of the gist. be. For example, all or part of the device can be functionally or physically distributed and integrated into arbitrary units. In addition, new embodiments created by arbitrary combinations of multiple embodiments are also included in the embodiments of the present invention. The effects of the new embodiment resulting from the combination have the effects of the original embodiment.

The processors of the in-home device 1 and the management server 4 are responsible for each step (process) included in the image analysis method shown in FIGS. 7 and 9. That is, the processors of the home device 1 and the management server 4 read a program for executing the image analysis method shown in FIGS. 7 and 9 from the storage unit, and execute the program to control each part of the image analysis system S. By doing so, the image analysis method shown in FIGS. 7 and 9 is executed. Some of the steps included in the image analysis method shown in FIGS. 7 and 9 may be omitted, the order of the steps may be changed, or a plurality of steps may be performed in parallel.

S Image analysis system 1 In-home device 11 Control unit 111 Image acquisition unit 112 Posture identification unit 113 Mask detection unit 114 Component identification unit 115 Analysis unit 116 Output unit 4 Management server 41 Control unit 411 Receiving unit 412 Analysis unit 413 Output unit

Claims (11)

an image acquisition unit that acquires a captured image of a patient performing peritoneal dialysis using a dialysis machine or an assistant who assists the patient;
a posture identifying unit that identifies the posture of the patient or the assistant based on the captured image;
a component identification unit that identifies, based on the captured image, whether each of the plurality of components included in the dialysis device is in an open state that allows the dialysate to pass through or a closed state that does not allow the dialysate to pass ;
specifying the order in which the patient or the assistant operated the plurality of parts based on the posture and whether each of the plurality of parts is in the open state or the closed state , and the specified order an analysis unit that determines whether or not the reference order matches the reference order stored in advance in the storage unit;
an output unit that outputs an analysis result including information indicating whether the order matches the reference order;
An image analysis device having: The analysis unit calculates the operation time during which the patient or the assistant operated at least one of the plurality of parts in each of the plurality of steps of the peritoneal dialysis included in the specified order, and the calculated operation time. Determining whether each of the plurality of steps was smoothly performed based on the
The image analysis device according to claim 1 , wherein the output unit outputs the analysis result including information indicating whether each of the plurality of steps has been smoothly performed. The image analysis device according to claim 1 or 2 , wherein when the analysis section determines that the order does not match the reference order, the output section transmits the captured image to the outside of the image analysis device. The analysis unit calculates the number of times the patient or the assistant touches an object other than the dialysis machine based on the posture, and determines the degree of hygiene of the peritoneal dialysis procedure based on the calculated number of times. Evaluate,
The image analysis device according to any one of claims 1 to 3 , wherein the output unit outputs information indicating the degree of hygiene as the analysis result. further comprising a mask detection unit that detects a mask worn by the patient or the assistant based on the captured image,
The image analysis device according to any one of claims 1 to 4 , wherein the output unit outputs information indicating whether the mask detection unit has detected the mask. The image acquisition unit acquires the captured image including an exit portion of the patient's body to which the dialysis machine is connected,
The analysis section evaluates the degree of hygiene of the exit section based on the captured image,
The image analysis device according to any one of claims 1 to 5 , wherein the output unit outputs information indicating the degree of hygiene as the analysis result. The analysis unit inputs the captured image to a machine learning model that outputs information indicating the degree of hygiene of the exit section included in the captured image, thereby determining the level of hygiene of the exit section. 7. The image analysis device according to claim 6 , wherein the image analysis device acquires information indicating the degree of hygiene of the object. The output unit performs at least one of outputting a predetermined notification and outputting information for stopping the dialysis device when the degree of hygiene of the outlet portion satisfies a predetermined condition. The image analysis device according to item 6 or 7 . The processor executes
A step of identifying the posture of a patient or an assistant assisting the patient based on a captured image, the captured image being an image of the patient or the assistant performing peritoneal dialysis using a dialysis machine. A certain step and
identifying whether each of the plurality of components included in the dialysis device is in an open state that allows the dialysate to pass through or a closed state that does not allow the dialysate to pass, based on the captured image;
identifying the order in which the patient or the assistant operated the plurality of parts based on the posture and whether each of the plurality of parts is in the open state or the closed state ;
determining whether or not the specified order matches a reference order stored in advance in a storage unit;
outputting an analysis result including information indicating whether the order matches the reference order;
An image analysis method comprising: comprising an image acquisition device and an image analysis device capable of communicating with the image acquisition device,
The image acquisition device includes:
an image acquisition unit that acquires a captured image of a patient performing peritoneal dialysis using a dialysis machine or an assistant who assists the patient;
a posture identifying unit that identifies the posture of the patient or the assistant based on the captured image;
a component identification unit that identifies, based on the captured image, whether each of the plurality of components included in the dialysis device is in an open state that allows the dialysate to pass through or a closed state that does not allow the dialysate to pass ;
a transmitter that transmits information indicating the posture and information indicating whether each of the plurality of components is in the open state or the closed state to the image analysis device;
has
The image analysis device includes:
a receiving unit that receives information indicating the posture and information indicating whether each of the plurality of components is in the open state or the closed state from the image acquisition device;
specifying the order in which the patient or the assistant operated the plurality of parts based on the posture and whether each of the plurality of parts is in the open state or the closed state , and the specified order an analysis unit that determines whether or not the reference order matches the reference order stored in advance in the storage unit;
an output unit that outputs an analysis result including information indicating whether the order matches the reference order;
An image analysis system with The output unit of the image analysis device transmits a request for the captured image to the image acquisition device when a predetermined input is received from a user or when the analysis result satisfies a predetermined condition;
The image analysis system according to claim 10 , wherein the transmission unit of the image acquisition device transmits the captured image to the image analysis device on the condition that the request is received from the image analysis device.

Images