Nurse call system

Abstract

To enable a medical worker who receives a notification to grasp urgency of a response according to a detected state of a patient.SOLUTION: An AI camera 1 analyzes a photographed image of an area on a patient's bed and its periphery, so as to detect a lying state of the patient lying on the bed, a leaving state after the patient leaves the bed, and an intermediate state between the lying state and the leaving state, and then, outputs a call signal when the leaving state is detected. At this time, the AI camera 1 sets different call types according to the detection state of the intermediate state. When the call signal is received, an NC master unit 4 performs a notification operation in different modes depending on the call types. Consequently, the modes of the notification operation are changed in accordance with the detection state of the lying state, the intermediate state and the leaving state, so that a medical worker who has received the notification confirms the state of the mode of the notification operation, and then, can grasp urgency of a response.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a nurse call system, and is particularly suitable for use in a nurse call system that calls a patient in accordance with the patient's condition detected by analyzing an image captured by a camera. [Background technology]

[0002] Conventionally, nurse call systems that analyze images captured by a camera and make a call based on the patient's condition are known (see, for example, Patent Documents 1 and 2). In the nurse call system described in Patent Document 1, when at least one of the following conditions is detected based on the image captured by a camera installed near the bed: the patient sits up or the patient leaves the bed, an attention state generation signal is output from a hallway light, and the nurse call master unit receives this attention state generation signal and issues an alert. Patent Document 1 describes that by outputting an attention state generation signal when the patient maintains an upright position for a certain period of time, it is possible to prevent erroneous operations such as mistaking a rolling over motion for a sitting up motion.

[0003] In the nurse call system described in Patent Document 2, the corridor light is equipped with a video analysis unit that detects when a patient leaves their bed based on the image captured by the camera. The video analysis unit sets a specific area within the camera's field of view to match the bed being imaged, and determines that the patient has left the bed when it detects the patient lying down in bed, sitting up with their upper body raised on the bed, and moving out of the specific area in that order. The nurse call master unit issues an alert when the video analysis unit determines that the patient has left the bed.

[0004] Here, the video analysis unit determines that bed exit has occurred if it detects a person moving out of the specific area within a predetermined time after detecting a sitting up state. Alternatively, the video analysis unit sets the predetermined time to an indefinite period, and once it detects a patient sitting up, it determines that bed exit has occurred if it detects a person moving out of the specific area regardless of the amount of time that has elapsed. Alternatively, the video analysis unit detects a state in which the patient is lying down in bed, detects a sitting up state from that state, and then determines that bed exit has occurred if it detects a person moving out of the specific area within a predetermined time. As in the third example, by using the detection of a sitting up movement from a lying down state as a prerequisite for determining that bed exit has occurred, erroneous detection, such as determining that a visitor sitting down on the bed is sitting up, is prevented.

[0005] Patent document 2 also describes that a bed exit may be determined to have occurred simply by detecting a person moving out of a specific area, or that the prerequisite for bed exit may be simply detecting a person lying down in bed, and that a bed exit may be determined to have occurred if a person moving out of the specific area from that position is detected. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Patent No. 5992296 [Patent Document 2] Patent No. 6454506 Summary of the Invention [Problem to be solved by the invention]

[0007] In the system described in Patent Document 1, if the system is configured to output a warning signal when the patient maintains an upright position for a certain period of time, if the patient moves quickly, the patient may leave the bed before the certain period of time has elapsed since waking up, and the patient may suddenly be detected as being out of bed without being detected as being upright. In such a situation, it is desirable for medical personnel to respond more quickly, but the system described in Patent Document 1 does not provide a mechanism to encourage this.

[0008] The system described in Patent Document 2 can also issue a warning when a patient suddenly gets out of bed, in addition to when a patient lying down in bed, getting up, and leaving a specific area are detected in sequence. However, the system described in Patent Document 2 does not provide a mechanism for prompting medical personnel to respond quickly when a patient suddenly gets out of bed.

[0009] The present invention was made in consideration of such conventional technology, and aims to enable medical personnel who receive the notification to understand the urgency of the response depending on the detected state of the patient in a nurse call system that is configured to detect when a patient is lying down in bed, when they have left the bed, and when they are in an intermediate state between these and issue an alert. [Means for solving the problem]

[0010] In order to solve the above-mentioned problems, the present invention analyzes a captured image including the area on the patient's bed and its surrounding area to detect a prone state in which the patient lies down on the bed, an out-of-bed state in which the patient leaves the bed, and one or more intermediate states between the prone state and the out-of-bed state, and issues a warning when an out-of-bed state is detected. Here, the present invention issues a warning in different modes depending on the detection status of one or more intermediate states. [Effects of the Invention]

[0011] According to the present invention configured as described above, in a nurse call system configured to detect the prone state, intermediate state, and out-of-bed state and perform an alarm action, the manner of the alarm action changes depending on the detected state. Therefore, by checking the manner of the alarm action, medical personnel who receive the alarm can understand what state of the patient was detected that resulted in the alarm action being performed, and thereby understand the urgency of the response. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is a diagram showing an example of the overall configuration of a nurse call system according to an embodiment of the present invention. [Figure 2] FIG. 1 is a diagram showing an example of an AI camera installation. [Figure 3] FIG. 1 is a block diagram illustrating an example of the functional configuration of an AI camera according to the present embodiment. [Figure 4] FIG. 2 is a block diagram showing an example of the functional configuration of an NC master unit according to the present embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0013] An embodiment of the present invention will be described below with reference to the drawings. Fig. 1 is a diagram showing an example of the overall configuration of a nurse call system according to this embodiment. Note that, although the description here takes as an example a nurse call system installed in a hospital, the nurse call system of this embodiment is not limited to systems installed in hospitals. For example, it can also be applied to systems installed in nursing homes, etc.

[0014] As shown in FIG. 1, the nurse call system of this embodiment is configured with an AI camera 1, a camera controller 2, a nurse call server (hereinafter referred to as the NC server) 3, a nurse call master unit (hereinafter referred to as the NC master unit) 4, a control unit 5, a hallway light 6, a wall-mounted slave unit 7, and a handheld slave unit 8. The AI ​​camera 1 corresponds to a transmitting terminal installed near the patient's bed. The NC master unit 4 corresponds to a receiving terminal installed at a location away from the bed. Communication between the AI ​​camera 1 and the NC master unit 4 and between the NC server 3 and the control unit 5 is performed, for example, using HTTP (Hypertext Transfer Protocol). Communication between the control unit 5 and the handheld slave unit 8 is performed via a nurse call trunk line.

[0015] AI camera 1 is a camera with a built-in computer that captures images of the patient's bed and the surrounding area. For example, as shown in Figure 2, AI camera 1 is installed for each bed on the wall in the direction where the patient's head will be positioned when sleeping. The wall on which AI camera 1 is installed should preferably be higher than the bed, as shown by the arrow in Figure 2(a), and its horizontal position should be within the range of the bed width plus a specified width, as shown by the arrow in Figure 2(b).

[0016] As will be described in detail later, the AI ​​camera 1 detects a prone state in which the patient lies down on the bed, an out-of-bed state in which the patient leaves the bed, and one or more intermediate states between the prone state and the out-of-bed state, and outputs a call signal when it detects at least the out-of-bed state. In this embodiment, the AI ​​camera detects two intermediate states: an upright state in which the patient sits upright on the bed, and an edge-sitting state in which the patient sits at the edge of the bed with their feet outside the bed. An upright state refers, for example, to a state in which the patient's back is at an angle of 45 degrees or more relative to the bed surface. Hereinafter, the prone state, the upright state, the edge-sitting state, and the out-of-bed state are referred to as the patient's behavioral states.

[0017] The camera controller 2 receives a call signal from the AI ​​camera 1 and notifies the NC server 3. The NC server 3 notifies the NC master unit 4 of the call signal sent from the camera controller 2. The NC server 3 can be linked with other systems used in the hospital besides the nurse call system (for example, an electronic medical record system, a medical administrative system, an ordering system, etc.).

[0018] The NC master unit 4 includes at least one of a fixed type installed in, for example, a nurse's station and a portable type such as a smartphone carried by each nurse, and receives a call signal sent from the NC server 3 and performs a predetermined notification operation. This notification operation, for example, displays information about the called patient on a display and outputs a ring tone from a speaker. When the medical professional responds by going off-hook in response to this notification operation, a call path is established, allowing the medical professional and the patient to talk.

[0019] In addition, the AI ​​camera 1 may further be equipped with a speaker, and may be configured to output the voice of the medical professional received via the call path formed in response to the response operation on the NC master unit 4 as described above from the speaker.

[0020] The control unit 5 is connected between the NC server 3 and the corridor light 6, and controls the transmission and reception of calls and data. The corridor light 6 is installed outside near the entrance of each hospital room. The corridor light 6 is equipped with a display device that displays the name of the patient in the room and also displays the fact that a call has been made when the patient in the room makes a call using the wall-mounted handset 7 or handheld handset 8. The corridor light 6 also transfers the call signal sent from the wall-mounted handset 7 or handheld handset 8 to the NC server 3 via the control unit 5.

[0021] The wall-mounted handset 7 is installed in the wall beside each bed in the hospital room. The wall-mounted handset 7 is equipped with a call button for the patient to call a medical professional, a microphone and speaker for the patient to use when talking with the medical professional, and a connection terminal for connecting the handheld handset 8. However, the configuration of the wall-mounted handset 7 is not limited to this. For example, it may not be equipped with a call button.

[0022] The handheld handset 8 is connected to the wall-mounted handset 7. The handheld handset 8 is equipped with a call button for the patient to call a medical professional, and a microphone and speaker for the patient to use when talking with the medical professional. However, the configuration of the handheld handset 8 is not limited to this. For example, it may be equipped without a microphone and speaker (a grip-type push button).

[0023] When a patient operates the call button on the wall-mounted handset 7 or handheld handset 8, a call signal is sent to the NC master unit 4 via the corridor light 6, the control unit 5, and the NC server 3. Hereinafter, the wall-mounted handset 7 and the handheld handset 8 will be collectively referred to as the nurse call handsets 7, 8.

[0024] FIG. 3 is a block diagram showing an example of the functional configuration of the AI ​​camera 1. As shown in FIG. 3, the AI ​​camera 1 of this embodiment has, as its functional configuration, a captured image acquisition unit 11, a behavioral state detection unit 12, and a call processing unit 13. These functional blocks 11 to 13 can be configured using any of hardware, a DSP (Digital Signal Processor), and software. For example, when configured using software, the functional blocks 11 to 13 are actually configured with a computer's CPU, RAM, ROM, etc., and are realized by the operation of a program stored in the RAM or ROM. Note that the program may also be stored in other storage media such as a hard disk or semiconductor memory.

[0025] The captured image acquisition unit 11 acquires captured images including the area on the patient's bed and the surrounding area. The AI ​​camera 1 constantly captures images of the area around the bed, and the captured images are constantly acquired by the captured image acquisition unit 11.

[0026] The behavioral state detection unit 12 detects whether the patient is in a prone state, a awake state, a sitting position on the edge of the bed, or an out-of-bed state by analyzing the photographed image acquired by the photographed image acquisition unit 11. The behavioral state detection unit 12 pre-sets a specific area within the photographed image to match the shape of the bed. The behavioral state detection unit 12 also extracts people appearing in the photographed image. The behavioral state detection unit 12 then detects whether the patient is in a prone state, a sitting state, a sitting position on the edge of the bed, or an out-of-bed state based on the posture of the extracted person and the relative positional relationship between the rectangular frame set within the person's range and the specific area.

[0027] Here, the behavioral state detection unit 12 detects the behavioral state of the patient at predetermined time intervals for the purpose of reducing the processing load, etc. Then, when the same behavioral state is detected multiple times in succession, it is determined that the patient is in that state. For example, the behavioral state detection unit 12 detects the behavioral state of the patient at one-second intervals, and when the same behavioral state is detected twice in succession (i.e., when the same behavioral state is detected for one second or more in succession), it is determined that the patient is in that state. In the following description, when "detecting the behavioral state" of the patient, it means detecting a confirmed behavioral state.

[0028] When the behavioral state detection unit 12 detects the behavioral state of the patient, it sets a state flag indicating the detected behavioral state. That is, when the behavioral state detection unit 12 detects that the patient is in a prone state, it sets a prone flag. When the behavioral state detection unit 12 detects that the patient is in an awake state, it sets a awake flag. When the behavioral state detection unit 12 detects that the patient is in an edge-sitting state, it sets an edge-sitting flag. When the behavioral state detection unit 12 detects that the patient is out of bed, it sets an out-of-bed flag.

[0029] These status flags are held until a call signal is output by the call processing unit 13, and are reset at the time the call signal is output. However, even when a call signal is not output, if a prone state is detected after a awake state is detected, the awake flag is reset. Also, if a sit-on-the-edge state is detected after a sit-on-the-edge state is detected, the sit-on-the-edge flag is reset. Also, if a prone state is detected after a sit-on-the-edge state is detected, the awake flag and the sit-on-the-edge flag are reset.

[0030] The behavioral state detection unit 12 can perform the above-described detection process of the patient's behavioral state using a machine-learned classification model. That is, multiple pieces of training data are prepared in which images of people or non-people objects on or around the bed are assigned correct labels—prone, awake, sitting on the edge of the bed, out of bed, or other states—and a classification model is machine-learned using this training data. The behavioral state detection unit 12 inputs the captured images acquired by the captured image acquisition unit 11 into the trained classification model, and obtains an output indicating a classification result—prone, awake, sitting on the edge of the bed, out of bed, or other states.

[0031] When the behavioral state detection unit 12 detects that the patient has left the bed, the call processing unit 13 transmits a call signal to the NC master unit 4 via the camera controller 2 and the NC server 3. This call signal includes a terminal ID and call type information that identify the AI ​​camera 1. The call signals transmitted from the nurse call slave units 7 and 8 include a slave ID and call type information that identify the nurse call slave units 7 and 8.

[0032] In this embodiment, the call processing unit 13 changes the call type depending on the detection status of one or more intermediate states by the behavioral state detection unit 12. For example, the call processing unit 13 changes the call type when the behavioral state detection unit 12 detects a prone state, and then successively detects intermediate states (a slumped state and a sitting-on-the-edge state) and an out-of-bed state (hereinafter referred to as a first detection status), and when the behavioral state detection unit 12 detects an out-of-bed state without detecting at least one of the intermediate states (hereinafter referred to as a second detection status). Whether the detection status is the first or second detection status can be determined by checking whether the slumped state flag and the sitting-on-the-edge flag are set when the out-of-bed flag is set.

[0033] The first detection situation is a situation in which the patient is detected as getting out of bed after the sitting state and the edge-sitting state have been detected consecutively for a predetermined time or more (after each state has been confirmed). Therefore, this first detection situation is considered to be a situation in which the patient behaves normally from the lying down state to the getting out of bed state, and the degree of urgency is not considered to be high, so the call type is set to "general call."

[0034] On the other hand, the second detection situation is a situation in which the out-of-bed state is detected after at least one of the sitting state and the edge-sitting state has transitioned within a predetermined time. Therefore, this second detection situation is considered to be a situation in which the patient has moved from the prone state to the out-of-bed state more quickly than usual, and since this is considered to be a high level of urgency, the call type is set to "emergency call."

[0035] Fig. 4 is a block diagram showing an example of the functional configuration of the NC master unit 4. As shown in Fig. 4, the NC master unit 4 of this embodiment has, as its functional configuration, a call receiving unit 41 and a notification processing unit 42. These functional blocks 41 to 42 can be configured using any of hardware, DSP, and software. For example, when configured using software, the functional blocks 41 to 42 are actually configured with a computer's CPU, RAM, ROM, etc., and are realized by the operation of a program stored in the RAM or ROM. Note that the program may also be stored in other storage media such as a hard disk or semiconductor memory.

[0036] The call reception unit 41 receives call signals transmitted from the AI ​​camera 1 and the nurse call handsets 7 and 8. When the call reception unit 41 receives a call signal (i.e., when the behavioral state detection unit 12 of the AI ​​camera 1 detects that the patient has left bed, or when the call button of the nurse call handsets 7 and 8 is operated), the notification processing unit 42 performs the predetermined notification operation described above in response to the reception of the call signal transmitted thereby.

[0037] For example, the notification processor 42 identifies the patient registered in association with the ID transmitted by the call signal (the terminal ID of the AI ​​camera 1, the handset ID of the nurse call handset 7, 8) by referencing a patient information management database (not shown). The notification processor 42 then reads part of the patient information (e.g., name, room number, bed number, etc.) of the identified patient from the patient information management database, generates a pop-up screen for the patient information, and displays it on the display. The notification processor 42 also sounds a call tone from the speaker to notify the medical staff.

[0038] At this time, the notification processor 42 changes the mode of the notification operation depending on the call type indicated by the call signal. For example, the notification processor 42 changes the priority of the notification depending on the call type. That is, when a general call and an emergency call overlap, the notification processor 42 prioritizes the notification of the emergency call. One mode of prioritizing the notification is to set the display order of emergency calls higher than general calls when patient information is displayed as a pop-up on the display of the NC master unit 4. Alternatively, the pop-up display may be configured to make emergency calls more noticeable than general calls in terms of window size, shape, color, blinking, etc. As another example, when a medical professional responds to the notification operation, the call status may be set to prioritize emergency calls over general calls.

[0039] As described above, in this embodiment, the notification processor 42 performs a different notification operation depending on the intermediate state detected by the behavioral state detector 12 of the AI ​​camera 1. That is, the notification processor 42 changes the mode of the notification operation depending on whether the situation pattern of the patient's behavioral state detected by the behavioral state detector 12 is the first detection situation or the second detection situation. Specifically, when the situation pattern of the behavioral state is the second detection situation, the notification operation is performed with a higher priority than a general call in accordance with the fact that the call type indicated by the call signal is an "emergency call."

[0040] According to the nurse call system of this embodiment configured as described above, when a patient's state of getting out of bed is detected and an alarm action is performed, the manner of the alarm action changes depending on the detection status of the intermediate state.Therefore, by checking the manner of the alarm action, medical personnel who receive the alarm can understand what state of the patient was detected as a result of which the alarm action was performed, and thereby understand the urgency of the response.

[0041] In the above embodiment, the alarm processor 42 changes the mode of the alarm operation depending on whether the first detection situation is a first detection situation in which the behavioral state detector 12 detects a prone state and then sequentially detects the sit-up state, edge-sitting state, and out-of-bed state, or a second detection situation in which the out-of-bed state is detected without detecting at least one of the sit-up state and edge-sitting state. However, the present invention is not limited to this. For example, the alarm processor 42 may change the mode of the alarm operation depending on whether the behavioral state detector 12 detects a prone state and then sequentially detects the sit-up state, edge-sitting state, and out-of-bed state, whether the sit-up state or edge-sitting state and out-of-bed state are detected, or whether the out-of-bed state is detected without detecting any intermediate states. In this case, the call processor 13 transmits a call signal by setting one of three call types: a general call, a first emergency call, or a second emergency call.

[0042] In the above embodiment, the behavioral state detection unit 12 detects two intermediate states, a waking state and a sitting-on-the-edge state, but the present invention is not limited to this. For example, the behavioral state detection unit 12 may detect only one of the waking state and the sitting-on-the-edge state as the intermediate state.

[0043] In the above embodiment, an example in which a notification action is performed when a bed-out state is detected has been described, but the present invention is not limited to this. For example, a notification action may be performed when an intermediate state is detected in addition to when a bed-out state is detected. Here, the user may be able to set whether or not a notification action is performed when an intermediate state is detected.

[0044] Furthermore, the above-described embodiments are merely examples of specific embodiments for carrying out the present invention, and the technical scope of the present invention should not be construed as being limited thereby. In other words, the present invention can be carried out in various forms without departing from its spirit or main characteristics. [Explanation of symbols]

[0045] 1 AI camera (transmitting device) 2 Camera Controller 3 Nurse call server (NC server) 4 Nurse call base unit (NC base unit) (receiving terminal) 11. Image acquisition unit 12 Behavioral state detection unit 13 Call Processing Unit 41 Call Reception Unit 42 Notification processing section

Claims

1. an image acquisition unit for acquiring an image including an area on the patient's bed and its surroundings; a behavioral state detection unit that analyzes the photographed images acquired by the photographed image acquisition unit to detect, as the behavioral state of the patient, a prone state in which the patient lies down on the bed, an out-of-bed state in which the patient leaves the bed, and one or more intermediate states between the prone state and the out-of-bed state; a notification processing unit that performs a notification operation only when the bed getting out state is detected by the behavior state detection unit, The notification processing unit performs the notification operation with a call type set to a general call when the one or more intermediate states and the getting-out state are successively detected by the behavioral state detection unit from the state in which the lying-down state has been detected by the behavioral state detection unit, and performs the notification operation with a call type set to an emergency call when the getting-out state is detected without detecting at least one of the one or more intermediate states. A nurse call system characterized by:

2. The intermediate state includes two states: a sitting state in which the patient sits upright on the bed, and a sitting state in which the patient sits on the edge of the bed with his / her feet outside the bed, The notification processing unit performs the notification operation with the call type set to the general call when the behavioral state detection unit sequentially detects the sitting state, the edge-sitting state, and the getting-out state from the state in which the lying-down state has been detected by the behavioral state detection unit; performs the notification operation with the call type set to the first emergency call when the behavioral state detection unit sequentially detects either the sitting state or the edge-sitting state and the getting-out state; and performs the notification operation with the call type set to the second emergency call when the intermediate state is not detected at all and only the getting-out state is detected.

2. The nurse call system according to claim 1.

3. The notification processing unit changes the priority of notification according to the call type set according to the detection status of the one or more intermediate states by the behavioral state detection unit.

3. The nurse call system according to claim 1 or 2.

4. The nurse call system includes a transmitting terminal installed near the bed and a receiving terminal installed at a position away from the bed, The transmitting terminal is The captured image acquisition unit; The behavioral state detection unit; a call processing unit that transmits a call signal to the receiving terminal only when the getting out of bed state is detected by the behavior state detection unit, the call processing unit changes the call type in accordance with a detection status of the one or more intermediate states by the behavioral state detection unit; the receiving terminal includes the notification processing unit, The notification processing unit changes the mode of the notification operation in accordance with the call type indicated by the call signal transmitted from the transmitting terminal.

4. The nurse call system according to claim 1, wherein the nurse call system is a call center.

Images