JP2006202088A - Transmitter and information monitoring and reporting system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a new system for automatically identifying alarm signals and transmitting information. <P>SOLUTION: A reporting unit 20 is provided for fetching via a microphone 50 information signals including an alarm sound issued by a medical instrument 11 and environmental sounds issued from around the medical instrument 11 and for converting the signals into predetermined transmission signals for output. A monitoring center 40 is provided with an information management unit 41 that receives an alarm signal of the reporting unit 20 via a telephone network 30 and outputs information about the alarm sound of the medical instrument 11. The reporting unit 20 identifies the alarm sound of the medical instrument 11 from the information signals. The reporting unit 20 comprises a neural network that learns a single or a plurality of signal patterns for identifying the alarm sound of the medical instrument 11 in advance, so that the alarm sound is identified from the information signals according to the signal patterns. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

    The present invention relates to a transmitter and an information monitoring and reporting system, and particularly relates to determination of a notification signal transmitted from a transmitter.

    Conventionally, as disclosed in Patent Document 1, the information monitoring and reporting system includes a reporting device worn by a user, so that emergency information output from the reporting device is received at home or a police box telephone. There is what I did.

In addition, as disclosed in Patent Document 2, another information monitoring and reporting system includes a sensor that detects a user's health condition, and the detection data such as body temperature detected by the sensor is communicated to a monitoring device. There is something to send through.
JP 2003-218994 A JP 2003-198759 A

    However, in the conventional information monitoring and reporting system (Patent Document 1), since emergency information is input to the notification device with a button or the like, it is necessary for the user to input emergency information, which is troublesome. There was a problem.

    In another conventional information monitoring and reporting system (Patent Document 2), the sensor captures and outputs user information, so the sensor accurately detects information such as body temperature from the user. There was a need.

    On the other hand, current home medical devices include an oxygen concentrator. Some of these medical devices, for example, output an alarm sound when the amount of supplied oxygen decreases. When the patient heard this alarm sound, the patient reported the abnormality to the monitoring center by telephone or the like. Therefore, there is a problem that it takes much time and effort for the patient.

    Thus, it has been desired to develop a new system that automatically identifies the notification signal when an alarm signal such as an alarm sound is output and transmits information to a monitoring center or the like.

    The present invention has been made in view of such a point, and an object thereof is to provide a new system that automatically identifies a notification signal and transmits information.

    First, the first invention is directed to a transmitter that takes in an information signal including a notification signal transmitted from a transmitter and an environmental signal transmitted from the periphery of the transmitter, converts the information signal into a predetermined transmission signal, and outputs the signal. . And the identification means for identifying the alerting | reporting signal of a transmitter from the said information signal is provided.

    Further, according to a second invention, in the first invention, the identifying means learns in advance a single or a plurality of signal patterns for identifying the notification signal of the transmitter, and from the information signal based on the signal pattern The broadcast signal is configured to be identified.

    According to a third aspect, in the second aspect, the identification means learns, as a signal pattern, a single or a plurality of target patterns corresponding to the notification signal and a non-target pattern corresponding to the environmental signal. It consists of a network.

    In a fourth aspect based on the first aspect, the transmitter is provided in a medical device.

    Moreover, 5th invention is provided in the medical device in the said 1st invention so that the said transmitter may transmit an alarm sound as a notification signal. The transmitter body is configured separately from the medical device, and is configured to collect the alarm sound.

    In addition, a sixth invention is the above-mentioned fifth invention, wherein the transmitter main body is attached to or close to the medical device and includes a microphone for collecting the alarm sound.

    In a seventh aspect based on the sixth aspect, one or a plurality of microphones are provided in the transmitter main body or in proximity to the transmitter main body.

    In an eighth aspect based on the sixth aspect, a sound absorbing material is provided on the attachment surface of the transmitter main body with respect to the medical device so as to contact the medical device and surround the microphone.

    Further, a ninth invention includes sound collection means for collecting information signals including an alarm sound collected by the microphone and an environmental sound which is an environmental signal in the sixth invention. Further, the identification means is connected to the sound collection means and is configured to identify the alarm sound of the medical device from the information signal.

    The tenth invention includes a transmitter that takes in an information signal including a notification signal transmitted from a transmitter and an environmental signal transmitted from the periphery of the transmitter, converts the information signal into a predetermined transmission signal, and outputs the signal. . Furthermore, a receiver for receiving a transmission signal of the transmitter via a communication circuit and outputting information of a notification signal of the transmitter is provided. And either of the said transmitter and a receiver is provided with the identification means which identifies the alerting | reporting signal of a transmitter from an information signal.

    In an eleventh aspect based on the tenth aspect, the identifying means learns in advance a single or a plurality of signal patterns for identifying the notification signal of the transmitter, and from the information signal based on the signal pattern. The broadcast signal is configured to be identified.

    In a twelfth aspect based on the eleventh aspect, the identification means learns, as a signal pattern, a single or a plurality of target patterns corresponding to a notification signal and a non-target pattern corresponding to an environmental signal. It consists of a network.

    In a thirteenth aspect based on the tenth aspect, the transmitter is provided in a medical device.

    In a fourteenth aspect based on the tenth aspect, the transmitter is provided in a medical device so as to transmit an alarm sound as a notification signal. The transmitter is configured separately from the medical device and is configured to collect the alarm sound.

    In a fifteenth aspect based on the tenth aspect, the communication circuit is constituted by a telephone network, and the transmitter is connected to the telephone network. On the other hand, the receiver is connected to a telephone network and is constituted by an information management unit of a monitoring center that monitors medical devices.

-Action-
In the first invention and the tenth invention, first, the transmitter transmits various notification signals in accordance with the operation status. For example, in the fourth invention, the fifth invention, the thirteenth invention, or the fourteenth invention, a transmitter of a medical device such as an oxygen concentrator transmits an alarm sound as a notification signal. This alarm sound is received by the transmitter from the medical device. For example, in the sixth invention, the alarm sound propagates from the medical device to the mounting surface of the transmitter. In addition to the oxygen concentrator, the medical device includes a ventilator, an infusion pump, a dialysis machine, an electrocardiograph, and the like.

    Therefore, the transmitter receives the notification signal. For example, in the seventh aspect, the microphone of the transmitter collects the alarm sound. At this time, the microphone collects environmental sounds transmitted from around the medical device as external sounds. The alarm sound and environmental sound collected by the microphone are output as information signals and input to the identification means. That is, the notification signal is input to the identification unit.

    In the first and tenth inventions, for example, the identification means is provided in the transmitter and identifies a notification signal such as an alarm sound from the information signal. In that case, in the second and eleventh aspects, the identification means learns a single or a plurality of signal patterns for identifying a notification signal such as an alarm sound in advance, and from the information signal based on the signal pattern. Identify the broadcast signal. That is, in the third and twelfth inventions, the identification means learns, as a signal pattern, a single or a plurality of target patterns corresponding to the notification signal and a non-target pattern corresponding to the environmental signal, as a signal pattern. To do. The receiver receives the notification signal identified by the identification means. For example, in the fifteenth aspect, the notification signal is transmitted from the telephone network to the monitoring center and displayed on the information management unit. In this monitoring center, for example, a monitor manages medical devices and the like based on the display of the information management unit.

    As described above, according to the present invention, since the notification signal of the transmitter is identified from the information signal, the information of the transmitter can be accurately determined. As a result, measures corresponding to the state of the transmitter can be quickly and accurately performed.

    Further, if the transmitter is provided with identification means, the transmission amount can be minimized.

    According to the second and eleventh aspects of the present invention, the identification means learns a single or a plurality of signal patterns for identifying the notification signal. Since the environmental signal can be accurately identified, the determination accuracy can be further improved.

    According to the third and twelfth inventions, since the identification means is configured by a neural network, it can learn a single signal pattern or a plurality of signal patterns for identifying a notification signal with high accuracy. Can do. As a result, the transmitter information can be more accurately determined as the learning accuracy improves, and the transmitter can be dealt with more quickly and accurately.

In particular, when learning with a neural network, since a single or a plurality of target patterns and non-target patterns are learned, the notification signal can be identified with high accuracy, thereby further improving the reliability of the determination. be able to.

Further, according to the fourth and thirteenth inventions, since the information on the medical device can be identified, the user's sense of security for the medical device can be greatly improved, and the reliability for the medical device can be improved. Can be improved.

    In particular, according to the fifth invention and the fourteenth invention, since the alarm sound of the medical device can be identified, the reliability of the medical device can be reliably improved.

    In addition, since the transmitter is configured separately from the medical device, the transmitter can be applied to existing medical devices in addition to various medical devices. As a result, the use range of various medical devices can be expanded.

    According to the sixth invention, the transmitter is directly attached to the medical device or provided close to the device, and the alarm sound is collected by the microphone, so that the state of the medical device can be grasped easily and reliably. be able to. From this, it is possible to respond quickly and to select and receive a signal in an optimal state.

    According to the seventh invention, since the transmitter is provided with one or a plurality of microphones or provided close to each other, applicable medical devices can be expanded.

    According to the eighth invention, since the transmitter is provided with a sound absorbing material surrounding the microphone, the target alarm sound can be collected more accurately.

    According to the fifteenth aspect of the invention, the transmitter provides information to the monitoring center using a telephone network, so that the monitoring center can accurately grasp information on a large number of medical devices.

    Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

<Embodiment 1>
In the first embodiment, as shown in FIGS. 1 and 2, an information monitoring and reporting system 10 is applied to a medical device 11.

    The medical device 11 is composed of, for example, an oxygen concentrator. The medical device 11 constitutes, for example, a transmitter that transmits an alarm sound that is a vibration wave notification signal when the amount of supplied oxygen decreases.

    The information monitoring and reporting system 10 includes a reporting unit 20 that is a transmitter, a telephone network 30, and an information management unit 41 of a monitoring center 40 that is a receiver, and reports various types of information on the medical device 11 to the monitoring center 40. The system is configured.

    The notification unit 20 constitutes a transmitter that identifies the alarm sound of the medical device 11, converts information corresponding to the alarm sound into a predetermined transmission signal, and outputs the signal to the monitoring center 40. As shown in FIG. 3, the notification unit 20 includes three microphones 50, an alarm sound detection / identification unit 21, and a notification unit 22.

    The notification unit 20 is formed in a flat rectangular body as shown in FIG. FIG. 4 is a perspective view showing the lower surface of the reporting unit 20.

    The notification unit 20 is configured separately from the medical device 11 such as an oxygen concentrator. For example, the notification unit 20 is configured to be installed directly on the upper portion of the medical device 11. Three microphones 50 and a sound absorbing material 51 are provided on the lower surface of the reporting unit 20, that is, the attachment surface 2 a of the reporting unit 20 with respect to the medical device 11. The microphone 50 is provided at each of two corners of the center portion of the attachment surface 2a of the reporting unit 20 and the front side portions of the attachment surface 2a. And since the alarm sound which the said medical device 11 transmits propagates the top plate of the medical device 11, each said microphone 50 is comprised so that the alarm sound which propagates a top plate may be collected. The sound absorbing material 51 is provided over the entire periphery at the peripheral edge of the attachment surface 2 a of the reporting unit 20. The sound absorbing material 51 is for absorbing extraneous sound, and is provided so as to contact the upper surface of the medical device 11 and surround the microphone 50.

    The sensing identification unit 21 includes a sound collection board 60 and a DSP board 70 (Digital Signal Processor Board).

    The sound collecting board 60 includes sound collecting means for collecting information signals including alarm sounds of the medical device 11 collected by the microphone 50 and various environmental sounds that are environmental signals transmitted from around the medical device 11. It is composed. The sound collection board 60 includes a CPU 61 and two multiplexers 62 and 63 to which the microphone 50 is connected. Then, the output signal of the first multiplexer 62 is removed from the signal of a predetermined frequency by the high-pass filter 64, amplified by the amplifier 65, and then input to the DSP board 70 via the AD converter 66. The output of the second multiplexer 63 is amplified by the amplifier 67 and then input to the CPU 61.

    The CPU 61 is configured to output control signals to the two multiplexers 62 and 63 while outputting various display signals to the display board 23. The display board 23 is provided with a plurality of display lamps 24 made of LEDs.

    Further, a connection terminal 25 for connecting an external microphone is connected to the two multiplexers 62 and 63. That is, when the reporting unit 20 is used for another device, an external microphone is connected to the connection terminal 25, and the reporting unit 20 is configured to be able to collect the alarm sound of the other device.

    The reporting unit 20 is configured such that the power supply 12 is connected to the output power supply unit 26 and supplied with power.

    The DSP board 70 includes an identification unit 71 constituted by a DSP. The identification unit 71 is connected to a FIFO memory 72, a memory 73 composed of a static RAM, and an interface 74 composed of RS232C. The identification unit 71 is configured to input an information signal collected by the microphone 50 from the AD converter 66 via the FIFO memory 72.

    The identification unit 71 constitutes identification means for identifying an alarm sound from an information signal. That is, since the microphone 50 collects the alarm sound and picks up various environmental sounds transmitted from the surroundings of the notification unit 20, the identification unit 71 identifies only the necessary alarm sound. Yes. Although not shown, the identification unit 71 includes a band pass filter (BPF) and the like, and applies fast Fourier transform (FFT) to the decomposition into frequency components to identify an alarm sound transmitted from the medical device 11 in advance. Is configured to learn a single or a plurality of signal patterns and identify an alarm sound from an information signal based on the signal patterns. In particular, the identification unit 71 includes a neural network that learns, as a signal pattern, a single or a plurality of target patterns corresponding to an alarm sound and a non-target pattern corresponding to an environmental sound.

    Specifically, the schematic configuration of the identification unit 71 will be described with reference to FIG.

    The alarm sound of the medical device 11 collected by the microphone 50 and various environmental sounds are input to the fast Fourier transform unit (7a) via the AD converter 66 and normalized. The output signal of the fast Fourier transform unit (7a) is subjected to feature extraction in the slab value creation unit (7b) and input to the neural network unit (7c).

    The neural network unit (7c) is composed of a multilayer perceptron, and is composed of an input layer (7d), one or more intermediate layers (7e) and an output layer (7f), and the identification operation is performed from the input layer (7d). Processing is performed only in the direction of the output layer (7f). The identification unit 71 is configured to output, for example, an alarm signal of three types of alarm sounds from the information signal, and an exclusion signal that does not correspond to any alarm sound. Examples of the alarm signal include a signal indicating disconnection, a signal indicating a decrease in oxygen concentration, and a signal indicating cannula breakage.

    The reason for learning the non-target pattern in addition to the target pattern is as follows. When only the target pattern is used as learning data, it may be identified as any of the target patterns with respect to the non-target pattern. Therefore, an output cell corresponding to the non-target pattern is newly added as a cell for exclusion, and the non-target pattern is added as learning data. As a result, it is possible to reduce the frequency of occurrence of a phenomenon that an output cell corresponding to a target pattern is activated when a non-target pattern is input (a phenomenon that is determined as a target pattern).

    The role of the non-target pattern cell is to suppress the activity of the output cell corresponding to the target pattern. Therefore, instead of setting the output cells according to the possible non-target pattern variations, A group of patterns such as a “trash box” in which non-target patterns are collected is used as a learning pattern.

    The alarm signal of the alarm sound identified by the identification unit 71 is input to the CPU 61 from the interface 74 of the DSP board 70 via the interface 68 including the RS232C of the sound collection board 60. The alarm signal is transmitted from the CPU 61 to the telephone network 30 via the notification unit 22 which is a modem.

    The telephone network 30 constitutes a communication circuit that receives an alarm signal transmitted from the reporting unit 20 via the telephone 31 and transmits the alarm signal to the monitoring center 40.

    On the other hand, the monitoring center 40 includes an information management unit 41 and constitutes a receiver that receives an alarm signal (transmission signal) transmitted from the telephone network 30. The information management unit 41 is configured to receive an alarm signal from the telephone network 30 via the modem 42.

    In the monitoring center 40, the administrator manages the medical device 11 with the information management unit 41, and takes measures corresponding to the alarm signal.

-Driving action-
Next, the notification monitoring operation of the information monitoring notification system 10 will be described.

    First, the medical device 11 transmits various alarm sounds according to the operating state. For example, when the amount of oxygen supplied to the oxygen concentrator decreases, the medical device 11 that is the oxygen concentrator transmits an alarm sound. This alarm sound propagates through the top plate of the medical device 11 to the mounting surface 2a, which is the lower surface of the notification unit 20.

    Therefore, the microphone 50 of the reporting unit 20 collects the alarm sound. At that time, the microphone 50 also collects environmental sounds transmitted from around the medical device 11 as external sounds. The alarm sound and environmental sound collected by the microphone 50 are output as information signals and input to the two multiplexers 62 and 63. The information signal is input from the second multiplexer 63 to the CPU 61 via the amplifier 67, while the information signal is amplified by the amplifier 65 after the first multiplexer 62 removes a signal having a predetermined frequency by the high pass filter 64. After that, the signal is input to the DSP board 70 via the AD converter 66.

    The information signal is input from the AD converter 66 of the sound collection board 60 to the identification unit 71 via the FIFO memory 72 of the DSP board 70.

    In the identification unit 71, only the alarm sound is identified among the information signals, and the alarm signal of the identified alarm sound is input to the CPU 61 via the interface 74 of the DSP board 70 and the interface 68 of the sound collection board 60. . The CPU 61 turns on the display lamp 24 based on the alarm signal, and the alarm signal is transmitted to the telephone network 30 via the modem and the telephone 31.

    Thereafter, the alarm signal is transmitted from the telephone network 30 to the monitoring center 40, and the alarm signal information is displayed on the information management unit 41. In the monitoring center 40, the monitoring person manages the medical device 11 based on the information of the alarm signal displayed on the information management unit 41, and performs predetermined measures such as replacement of the oxygen concentrator.

    Therefore, the reporting operation in the reporting unit 20 will be described based on the flowchart of FIG.

    In step ST1, when an alarm sound and an environmental sound information signal are input to the microphone 50, the process proceeds to step ST2, and the identification unit 71 determines the alarm sound from the information signal. Specifically, the identification unit 71 outputs any one of alarm 1 output, alarm 2 output, alarm 3 output, and exclusion output. In response to the output of the identification unit 71, the following operation is repeated.

    Next, the process proceeds to step ST3, where it is determined whether or not the output of the identification unit 71 is any one of an alarm 1 output, an alarm 2 output, and an alarm 3 output. That is, it is determined whether or not there are three types of alarm sounds transmitted by the medical device 11.

    If the determination in step ST3 is an alarm sound, the process proceeds to step ST4, and the number of times the alarm sound is output is counted. Then, it moves to step ST5 and returns to step ST3 until the count of step ST4 becomes five times. When the count reaches 5, the process moves from step ST5 to step ST6, and it is determined whether or not it is within 30 seconds. If 30 seconds have elapsed, the process moves to step ST7, the count is cleared, and the process returns to step ST3. That is, the identification unit 71 determines whether to output an alarm signal five times within 30 seconds.

    When the identification unit 71 outputs an alarm signal five times within 30 seconds, the process proceeds from step ST6 to step ST8, and transmits to the monitoring center 40 and is displayed on the display board 23. Specifically, the detection LED that is the display lamp 24 is turned on, and the notification LED that is the display lamp 24 is blinked. Thereafter, the process proceeds to step ST9, where it is determined whether or not the connection to the monitoring center 40 is completed. When the connection is completed, the process proceeds to step ST10. Then, the notification unit 20 transmits an alarm signal, clears the alarm counter, turns off the detection LED that is the display lamp 24, turns off the notification LED that is the display lamp 24, and turns off the communication that is the display lamp 24. The abnormal LED is turned off, the process returns to step ST3, and the above operation is repeated.

    In step ST10, if the connection with the monitoring center 40 is not completed, retry is performed until the number of retries reaches 30. That is, the process proceeds from step ST10 to step ST11, and it is determined whether or not the number of retries has reached 30. The process proceeds to step ST12 until the number of retries reaches 30, and the number of retries is counted and the process returns to step ST8. When the number of retries reaches 30 or more, the process proceeds from step ST11 to step ST13, the communication abnormality LED as the display lamp 24 is turned on, the process returns to step ST3, and the above operation is repeated.

    If step ST3 determines the exclusion output of the identification unit 71, since it is an environmental sound such as a foreign sound, the process proceeds to step ST14, the number of rejections is counted, then the process proceeds to step ST15, and the count value has reached 10. Determine whether or not. Until this count value becomes 10, it moves to step ST16 and determines whether it is within 1 minute or not. That is, when the exclusion output is output 10 times within 1 minute, the process proceeds from step ST16 to step ST17, and the noise detection LED as the display lamp 24 is lit for 15 seconds. Thereafter, the process proceeds to step ST18, the exclusion counter is cleared, the process returns to step ST3, and the above-described operation is repeated. If the exclusion output is not output 10 times within 1 minute, the determination in step ST16 is NO, the process proceeds to step ST18, and the above operation is repeated.

    Next, the alarm identification operation by the neural network of the identification unit 71 will be described.

    As shown in FIG. 7, in step ST <b> 21, rise detection of sound data such as an alarm sound collected by the microphone 50 is performed on the signal input to the identification unit 71. When this rising edge is detected, the process proceeds to step ST22 and data is extracted. That is, the fast Fourier transform unit (7a) extracts data from the information signal using a window function having a sufficient width for performing fast Fourier transform (FFT). Then, it moves to step ST23 and performs a fast Fourier transform on the information signal (extracted data) to convert it into a power spectrum.

    Subsequently, the process proceeds to step ST24, and feature extraction is performed from the power spectrum in the slab value creation unit (7b). At this time, paying attention to the fundamental frequency of the alarm sound and its constant multiple component, the maximum frequency search is performed with a certain width from the search start point of the basic frequency of the alarm sound given as known information, thereby obtaining the fundamental frequency and its constant multiple. The power spectrum value is identified and used as a part of the feature amount. Further, paying attention to the frequency component not including the alarm sound, it is set as a part of the feature amount.

    Thereafter, the process proceeds to step ST25, where it is determined whether or not the feature amount is sufficient, data is sought until it is sufficient, the process returns to step ST22, and the above operation is repeated.

    On the other hand, if the feature amount is sufficient in step ST25, the determination is YES, the process proceeds to step ST27, the feature quantity is normalized, then the process proceeds to step ST28, and an evaluation by a neural network is performed. In other words, the alarm sound is identified only when the maximum firing neuron fires sufficiently and the difference between the firing rate and the second largest neuron firing rate is sufficient. Thereafter, the process proceeds to step ST210, and when the evaluation by the neural network identifies the alarm sound, the alarm identification operation is terminated, and the above-described output of step ST2 in FIG. 6 is performed. If the evaluation by the neural network does not identify the alarm sound, the process returns to step ST21 and the above operation is repeated.

    Next, the rising detection operation in step ST21 in FIG. 7 will be described.

    As shown in FIG. 8, in step ST <b> 31, data such as an alarm sound collected by the microphone 50 is extracted from the signal input to the identification unit 71. At that time, a rectangular window having a sufficient width to be used for the rising determination process is applied.

    Subsequently, the process proceeds to step ST32, where a band pass filter (BPF) is applied to the extracted data to remove background noise. Then, it moves to step ST33 and performs a rising process. That is, the extracted data is divided into two buffers so that there is no data duplication and omission. Then, the ratio of the data exceeding the rising detection sound pressure threshold is obtained from the sound pressure data in the two buffers. When the ratio in the first buffer is equal to or lower than the rising detection ratio and the ratio in the second buffer is equal to or higher than the rising detection ratio, it is determined as rising and the head position of the second buffer is returned as the rising position.

    Thus, when the rising edge is detected, the determination in step ST34 is YES, the rising edge detection operation is terminated, and the process proceeds from step ST21 to step ST22 in FIG. On the other hand, when the above step ST34 does not detect a rising edge, the process proceeds to step ST35, the data is sought, the process returns to step ST31, and the above-described operation is repeated.

-Effect of Embodiment 1-
As described above, according to the first embodiment, since the alarm sound of the medical device 11 is identified from the information signal, the information of the medical device 11 can be accurately determined. As a result, measures corresponding to the state of the medical device 11 can be performed quickly and accurately.

    Further, since the identification unit 71 learns a single signal pattern or a plurality of signal patterns for identifying the alarm sound, it can accurately identify the alarm sound of the medical device 11 and the surrounding environmental sound. Therefore, the determination accuracy can be further improved.

    In addition, since the identification unit 71 is configured by a neural network, it can accurately learn a single or a plurality of signal patterns for identifying an alarm sound. As a result, the information of the medical device 11 can be more accurately determined as the learning accuracy is improved, and the medical device 11 can be dealt with more quickly and accurately.

    In particular, when learning with a neural network, since a single or a plurality of target patterns and non-target patterns are learned, the notification signal can be identified with high accuracy, thereby further improving the reliability of the determination. be able to.

    Moreover, since the information of the medical device 11 can be identified, the user's sense of security with respect to the medical device 11 can be greatly improved, and the reliability with respect to the medical device 11 can be improved.

    In particular, since the alarm sound of the medical device 11 at home can be identified, the reliability of the medical device 11 can be reliably improved.

    Further, since the reporting unit 20 is configured separately from the home medical device 11, the reporting unit 20 can be applied to existing medical devices 11 in addition to various medical devices 11. As a result, the use range of various medical devices 11 can be expanded.

    Further, since the reporting unit 20 is directly attached to the medical device 11 and collects the alarm sound by the microphone 50, the situation of the medical device 11 can be grasped easily and surely, so that a prompt response can be made. In addition, it is possible to select and receive an optimal signal.

    Moreover, since the reporting unit 20 is provided with a plurality of microphones 50, the applicable medical devices 11 can be expanded.

    Further, since the reporting unit 20 is provided with the sound absorbing material 51 surrounding the microphone 50, the target alarm sound can be collected more accurately.

    Moreover, since the identification unit 71 is provided in the reporting unit 20, the transmission amount can be minimized.

    In addition, since the reporting unit 20 provides information to the monitoring center 40 using the telephone network 30, the monitoring center 40 can accurately grasp information on a large number of medical devices 11.

<Embodiment 2 of the invention>
Next, Embodiment 2 of the present invention will be described in detail based on FIG.

    In the present embodiment, the notification signal of the device is identified to various types in the hospital. A transmission unit 80 as a communication circuit is connected to the notification unit 20, and the transmission unit 80 is connected to the nurse call parent device 81. It is connected.

    That is, the notification unit 20 is installed in the vicinity of a measuring instrument, for example, detects an alarm sound of the measuring instrument, identifies the alarm sound, and transmits an alarm signal to the transmission unit 80 via the signal line 8a. . On the other hand, the nurse call parent device 81 is installed in a nurse center to constitute a receiver. When the nurse call master device 81 receives an alarm signal from the notification unit 20 from the transmission unit 80, it displays a call sound or the like. In other respects, the configuration, operation, and effects are the same as those of the first embodiment.

    A nurse call terminal 82 is connected to the transmission unit 80. The notification unit 20 is configured to be applicable even when the alarm signal is an optical signal, that is, has a detection function for detecting the display of the alarm lamp transmitted from the measuring instrument. At that time, the reporting unit 20 transmits an alarm signal of an optical signal to the transmission unit 80 via the signal line 8b. Further, the notification unit 20 is configured to be applicable even when the alarm signal is an electrical signal, that is, has a reception function of receiving an alarm signal of an electrical signal transmitted from the measuring instrument. At that time, the notification unit 20 transmits an alarm signal of an electrical signal to the transmission unit 80 via the signal line 8c.

<Other embodiments>
The present invention may have the following configurations in addition to the first and second embodiments.

    The identification means that is the identification unit 71 may be provided in the receiver that is the information management unit 41 or the nurse call parent device 81 in addition to the transmission means that is the notification unit 20.

    The transmitter is not limited to the medical device 11 and the medical-related device, and may be any device that outputs a notification signal.

    Moreover, although the said microphone 50 was provided in the notification unit 20, you may make it provide close to the main body of the notification unit 20. FIG.

    Further, the notification signal may be an optical signal emitted from a notification lamp or the like in addition to a sound signal such as an alarm sound.

    As described above, the present invention is useful for an information monitoring notification system that identifies a notification signal transmitted by a medical device or the like.

It is a schematic block diagram which shows the information monitoring notification system of Embodiment 1. It is a block diagram which shows the information monitoring notification system of Embodiment 1. It is a circuit block diagram which shows the notification unit of Embodiment 1. It is the perspective view which looked at the reporting unit of Embodiment 1 from the lower surface. 2 is a configuration diagram illustrating a schematic configuration of an identification unit according to Embodiment 1. FIG. It is a flowchart which shows the notification operation of Embodiment 1. FIG. 6 is a flowchart illustrating an alarm identification operation according to the first embodiment. FIG. 5 is a flowchart illustrating a rising edge detection operation according to the first embodiment. It is a schematic block diagram which shows the information monitoring notification system of Embodiment 2.

Explanation of symbols

10 Information monitoring and reporting system 11 Medical device 20 Reporting unit (transmitter)
2a Mounting surface 30 Telephone network (communication circuit)
40 Monitoring Center 41 Information Management Unit (Receiver)
50 Microphone 51 Sound absorbing material 60 Sound collecting board 70 DSP board 71 Identification part (identification means)
80 Transmission unit 81 Nurse call base unit

Claims (15)

A transmitter that captures an information signal including a notification signal transmitted from a transmitter and an environmental signal transmitted from around the transmitter, converts the information signal into a predetermined transmission signal, and outputs the signal.
A transmitter comprising identification means for identifying a notification signal of a transmitter from the information signal. In claim 1,
The identification means is configured to learn a single or a plurality of signal patterns for identifying a notification signal of a transmitter in advance and identify the notification signal from an information signal based on the signal pattern. Transmitter. In claim 2,
The transmitter comprises a neural network that learns, as a signal pattern, a single or a plurality of target patterns corresponding to a notification signal and a non-target pattern corresponding to an environmental signal. In claim 1,
The transmitter is provided in a medical device. In claim 1,
The transmitter is provided in the medical device to transmit an alarm sound as a notification signal,
The transmitter main body is configured separately from the medical device, and is configured to collect the alarm sound. In claim 5,
The transmitter body is attached to a medical device or provided close to the transmitter body and includes a microphone for collecting an alarm sound. In claim 6,
A transmitter characterized in that one or a plurality of microphones are provided in the transmitter body or close to the transmitter body. In claim 6,
A transmitter characterized in that a sound absorbing material is provided on an attachment surface of the transmitter body with respect to the medical device so as to be in contact with the medical device and to surround the microphone. In claim 6,
A sound collecting means for collecting information signals including an alarm sound collected by the microphone and an environmental signal that is an environmental signal;
The transmitter is characterized in that the identification means is connected to the sound collection means and is configured to identify an alarm sound of a medical device from an information signal. A transmitter that captures an information signal including an informing signal transmitted from the transmitter and an environmental signal transmitted from around the transmitter, converts the information signal into a predetermined transmission signal, and
A receiver for receiving a transmission signal of the transmitter via a communication circuit, and outputting information of a notification signal of the transmitter;
Either of the transmitter and the receiver is provided with identification means for identifying the notification signal of the transmitter from the information signal. In claim 10,
The identification means is configured to learn a single or a plurality of signal patterns for identifying a notification signal of a transmitter in advance and identify the notification signal from an information signal based on the signal pattern. Information monitoring and reporting system. In claim 11,
The identification means is composed of a neural network that learns, as a signal pattern, a single or a plurality of target patterns corresponding to a notification signal and a non-target pattern corresponding to an environmental signal. . In claim 10,
An information monitoring and reporting system, wherein the transmitter is provided in a medical device. In claim 10,
The transmitter is provided in the medical device to transmit an alarm sound as a notification signal,
An information monitoring and reporting system, wherein the transmitter is configured separately from a medical device and is configured to collect the alarm sound. In claim 10,
The communication circuit is composed of a telephone network,
While the transmitter is connected to the telephone network,
An information monitoring and reporting system, wherein the receiver is connected to a telephone network and is constituted by an information management unit of a monitoring center that monitors medical devices.

Images