TW202116374A - Pipeline abnormality risk prediction method, device and related system - Google Patents

Abstract

A pipeline abnormality risk prediction method, device and related system are applied to a patient using a pipeline, wherein the method comprises the following steps: detecting a disturbance state of the pipeline or a specific action of the patient; Calculating a pipeline abnormality risk index according to the disturbance state of the road or the specific action or sound of the user; and corresponding to the range of the plurality of numerical ranges according to the pipeline abnormality risk index, and corresponding to the plurality of contents Select one of the different warning signals.

Description

Pipeline abnormal risk prediction method, device and related system

This case is a pipeline abnormality risk prediction method, device and related system, especially a pipeline abnormality risk prediction method, device and related system that can be applied to patients.

With the advent of an aging society, the number of patients requiring intubation for treatment has rapidly increased. The intubation treatment is mostly to maintain basic life, so if the tube accidentally falls off, it will cause serious consequences. However, most intubation treatments will cause discomfort to the patient, so that the patient may deliberately or subconsciously remove the tube by himself, causing irreversible errors. In order to prevent this kind of situation from happening, the traditional method needs to use a large number of nursing staff to conduct house inspections to detect the occurrence of abnormal pipeline events and prevent them in time. But this will take up a lot of human resources.

In order to improve the above-mentioned shortcomings of the prior art, this case developed a pipeline abnormality risk prediction method, which is applied to a patient using a pipeline. The method includes the following steps: detecting a disturbance state of the pipeline or the disease A specific action of the patient; a pipeline abnormality risk index is calculated according to the disturbance state of the pipeline or the specific action or sound of the user; and a pipeline abnormality risk index falls within a plurality of numerical ranges according to the pipeline abnormality risk index A range, and correspondingly select one output from a plurality of warning signals with different contents.

According to the above conception, in the pipeline abnormal risk prediction method of this case, the disturbance state is any or all of the change in the shape of the pipeline and the change in the flow rate of the fluid in the pipeline.

According to the above conception, in the pipeline abnormal risk prediction method of this case, the specific action of the patient is that the patient moves his free limb towards the pipeline and the patient’s free limb has any abnormal motion trajectory. One or all.

According to the above conception, the pipeline abnormality risk prediction method of this case, wherein the pipeline abnormality risk index is calculated according to the disturbance state of the fluid pipeline and a patient background data

According to the above-mentioned concept, the pipeline abnormality risk prediction method of this case, wherein the pipeline abnormality risk index is calculated according to the specific action of the user and a patient's background data.

According to the above conception, in the pipeline abnormality risk prediction method described in this case, the pipeline abnormality risk index falls within one of a plurality of numerical ranges. The first range represents that the patient pulls the pipeline or the free limb shakes violently. A first warning signal, it is recommended that caregivers use heavy sedatives and restraint bands on the patient, where the abnormal risk index of the pipeline falls within one of a plurality of numerical ranges. The second range represents the obvious movement of the pipeline or the patient’s The free limb has a large movement, corresponding to a second warning signal. It is recommended that the caregiver use a light tranquilizer on the patient and consider the use of a restraint band. The abnormal risk index of the pipeline falls in the third of a plurality of numerical ranges. The range represents the slight movement of the pipeline or the abnormal trajectory of the patient's free limbs. Correspondingly, a third warning signal is issued. It is recommended that the caregiver pay more attention to the patient and ask about the patient's condition.

According to the above conception, in the pipeline abnormality risk prediction method of this case, the specific action of the user is to make a sound.

Another aspect of this case is a pipeline abnormality risk prediction device, which is applied to patients who use a pipeline. The device includes: a sensor arranged near the pipeline to detect an abnormality of the pipeline. A disturbance state or a specific action of the patient; and an arithmetic unit, which is signal-connected to the sensor, and calculates a pipeline abnormality risk index according to the disturbance state of the pipeline or the specific action of the patient, And according to the abnormal risk index of the pipeline falling within one of a plurality of numerical ranges, correspondingly, one of a plurality of warning signals with different contents is selected and output.

According to the above-mentioned conception, in the pipeline abnormal risk prediction device of this case, the sensor includes a motion sensor or an image capture and analysis device, and the detected disturbance state is the appearance change of the pipeline.

According to the above conception, in the pipeline abnormality risk prediction device of the present case, the sensor includes a pipeline flow detector, and the detected disturbance state is the change of the fluid flow rate in the pipeline.

According to the above conception, in the pipeline abnormality risk prediction device of this case, the specific action of the patient detected by the sensor is the patient's freedom to move his free limbs toward the pipeline and the patient The limbs have any or all of the abnormal movement tracks.

According to the above conception, in the pipeline abnormality risk prediction device of this case, the calculation unit calculates the pipeline abnormality risk index according to the disturbance state of the fluid pipeline and a patient background data

According to the above conception, in the pipeline abnormality risk prediction device of this case, the calculation unit calculates the pipeline abnormality risk index according to the specific action of the user and a patient background data.

According to the above-mentioned conception, in the pipeline abnormality risk prediction device of this case, the pipeline abnormality risk index falls within one of a plurality of numerical ranges. The first range represents that the patient pulls the pipeline or the free limb shakes violently, and the computing unit Corresponding to a first warning signal, it is recommended that caregivers use heavy sedatives and restraint bands on the patient. The abnormal risk index of the pipeline falls within one of a plurality of numerical ranges. The second range represents the obvious movement of the pipeline or When the patient’s free limbs move in a large scale, the computing unit correspondingly sends out a second warning signal. It is recommended that the caregiver use mild tranquilizers on the patient and consider the use of restraint bands. The abnormal risk index of the pipeline falls on a plurality of values. One of the ranges and the third range represent slight movement of the pipeline or abnormal trajectory of the patient's free limbs. The computing unit correspondingly sends out a third warning signal. It is recommended that the caregiver strengthen observation of the patient and inquire about the patient's condition.

According to the above-mentioned conception, in the pipeline abnormality risk prediction device of the present case, the specific action of the user is to emit a sound, and the sensor is used to detect the sound to form the intensity and frequency spectrum data representing the sound.

Another aspect of this case is a pipeline abnormality risk prediction system, which is applied to a plurality of patients using multiple beds. The system includes: a plurality of sensors, respectively set corresponding to the plurality of patients Near the plurality of pipelines to detect the disturbance state of the plurality of pipelines or the specific actions of the plurality of patients; and an arithmetic unit, connected to the plurality of sensors via a network, according to the plurality of The disturbance state of the pipeline or the specific actions of the plurality of patients respectively calculate the corresponding plurality of pipeline abnormality risk indexes, and any one of the plurality of pipeline abnormality risk indexes falls within a plurality of numerical ranges One of the ranges, and correspondingly select one output from a plurality of warning signals with different contents. In the pipeline abnormality risk prediction method, the specific action of the patient is that the patient moves his free limb toward any or all of the abnormal motion trajectories of the pipeline and the patient's free limb.

Please refer to Figure 1A and Figure 1B, which are functional block diagrams of a pipeline abnormal event prediction device developed in this case, which is applied to patients who use a pipeline 10 (not shown). The pipeline can It is an endotracheal tube, an intravenous injection tube, a hemodialysis tube, etc. The pipeline abnormal event prediction device 1 mainly includes a sensor 11 and an arithmetic unit 12. The sensor 11 of the embodiment shown in FIG. 1A is set in the tube The vicinity of the road 10 is used to detect a disturbance state of the pipeline 10 or a specific action of the patient. As for the calculation unit, the signal is connected to the sensor 11, and a pipeline abnormality risk index is calculated according to the disturbance state of the pipeline 10 or the specific action of the patient, and the pipeline abnormality risk index falls on a complex number according to the pipeline abnormality risk index. One of a range of numerical values, and correspondingly select one of a plurality of different warning signals to output. The disturbance state detected by the sensor is the change in the appearance of the pipeline and the change in the flow rate of the fluid in the pipeline. Any or all of them.

The aforementioned sensor 11 can be one or any combination of a motion sensor 110, an image capture and analysis device 111, a pipeline flow detector 112, and a sound sensor 113. The sensor 11 in FIG. 1A can be a combination of a motion sensor 110 and a pipeline flow detector 112. It requires a pipeline 10 closer to the patient. In order to be concise and easy to read, the following embodiments take the motion sensor 110 as an example for description.

The motion sensor 110 can be attached to the pipeline 10 and close to the patient, such as a gravitational acceleration sensor (G-sensor). It can be used to sense the disturbance state of the pipeline 10 caused by the patient's movement (for example, the rapid displacement or rotation of the pipeline caused by the unfree hand shaking of the fixed pipeline) corresponding to the generation of a first sensing signal. The calculation unit 12 can calculate a single data or multiple data representing the abnormal risk (for example, falling off) of the pipeline according to the disturbance state of the fluid pipeline represented by the first sensing signal and a patient background data. Data array formed by data. The patient’s background information can be gender, age, number of extubation events in the patient’s medical history, hospitalization days, pain index and agitation index corresponding to the patient’s symptoms, etc. Therefore, the patient’s background information is useful for representative management. A single data or multiple data with abnormal road risk (such as falling off) will constitute a certain degree of impact.

In order to more accurately sense the patient’s specific movements, the motion sensor 110 can be a gravity acceleration sensor (G-sensor) attached to the pipeline 10, which can also include Other components work together, such as a multi-axis sensor 1100 produced by a combination of an angular velocity sensor (i.e. gyroscope) and a magnetic induction sensor (i.e. magnetometer), such as a 3-axis sensor, a 6-axis sensor, 9-axis sensor, etc. The functional block diagram inside the motion sensor shown in Figure 2 shows that the motion sensor 110 may additionally include an optical sensor 1101 that can detect the intensity of infrared (IR). The magnetic sensor 1102 that changes the magnetic flux caused by the proximity (wearing a magnet on the movable hand), or the capacitive sensor 1103 that can detect changes in capacitance, of course, can also be other devices that can detect changes in electric field. Type sensor. In this way, it can be used to determine whether the patient's other movable hand (a free hand without a tube) is close to the tube 10, and then sense that the patient's specific action affects the magnetic flux, capacitance, or The electric field is changed, and sufficient information is further provided for the calculation unit 12 to analyze, and then to perform an early warning judgment on the abnormal risk (such as falling off) of the pipeline caused by the manual operation.

As for another embodiment of the sensor 11, the image capture and analysis device 111 can be used. The captured images can also be used to detect specific actions of the patient, for example, the patient moves his free limbs toward the pipeline. If there is any or all of the abnormal movement trajectories with the free limbs of the patient, it can be judged that the patient is struggling or has a tendency to stretch out the tube. In addition, tags using indoor positioning technology can also be attached to the hands. In this way, by detecting the movement of the tag, it is also possible to determine whether the patient's free limbs have abnormal motion trajectories. The motion trajectory image of the aforementioned free limb can be compared with a normal image by the computing unit 12 to obtain one or more image difference values representing the degree of difference between the two images. The above-mentioned one or more image difference values can be part or all of the above-mentioned pipeline abnormality risk index.

As for the fluid flow rate change in the pipeline, it can also be regarded as a part of the disturbance state of the pipeline 10, and the fluid flow rate is used to be detected by the pipeline flow detector 112 to detect the cause The improper bending of the pipeline reduces the flow rate, which leads to a tendency to increase the risk of abnormal pipelines, which in turn allows the calculation unit 12 to perform analysis and judgment. In addition, because the patient may make some specific sounds in uncomfortable conditions, such as sighing, groaning or wailing, the sound intensity and frequency spectrum data sensed by the sound sensing device 113 can be used by the computing unit 12 Carry out analysis and judgment, and then make necessary adjustments to the pipeline abnormal risk index.

After one or a group of pipeline abnormality risk indexes are determined according to the above-mentioned various methods, the pipeline abnormality risk can be judged based on the value. Next, as shown in FIG. 3, as shown in the schematic flow chart of the pipeline abnormal event prediction method, the numerical range of the above-mentioned pipeline abnormal risk index can be defined as multiple numerical ranges. Therefore, step 30 is used to make a judgment based on the value of the abnormal risk index of the pipeline (for example, 0-100). When the value range of the index falls within the first range (for example, 70-100) among a plurality of numerical ranges, It means that the patient pulls the pipeline or the free limb shakes violently, and then proceeds to step 31, so that the arithmetic unit 12 correspondingly sends a first warning signal, and the caregiver is advised to use a heavy sedative and a restraint band on the patient. When the value range of the index falls within the second range (for example, 50-70) of the plurality of value ranges, it means that the pipeline moves significantly or the patient's free limb has a large movement, and then enters step 32 to make the The arithmetic unit 12 correspondingly sends out a second warning signal, and it is recommended that the caregiver use a light tranquilizer for the patient and consider using a restraint band. When the numerical range of the index falls within one of the plural numerical ranges, the third range (for example, 10-50) represents slight movement of the pipeline or abnormal trajectory of the patient’s free limb, and then proceeds to step 33, the calculation unit 12 Correspondingly, a third warning signal is issued, and it is recommended that the caregiver strengthen observation of the patient and ask about the patient's condition. Of course, if the value range of the index is less than 10, no notification will be made and the data will be continuously monitored and the amount of data collected will be reduced, thereby achieving the benefits of not disturbing patients and saving energy.

In order to clearly distinguish the classification levels, it can be seen that the above methods all need to set one or more threshold values for judgment, so as to determine which level of warning signal is to be issued. For this reason, this case can also collect a large amount of information to optimize the training mode of threshold value. Among them, the training mode established by artificial intelligence related technology is better. It depends on the collected image data of various patients, especially disease. The image data of the patient before manual extubation. Then feed a large amount of collected image data and the subsequent results of whether extubation occurs for machine learning, so as to establish a high-risk action mode for extubation, so that the threshold can be re-adjusted based on the results The distribution pattern of the data, and then reaches the threshold value optimization and can enter the state of accurate judgment without misjudgment. Of course, in addition to the image data, the sound data, magnetic flux, capacitance, or electric field and other disturbance-related data captured by the various sensors can also be used for machine learning. The high-risk action mode of extubation can further optimize the judgment and increase the probability of no misjudgment.

Furthermore, in order to complete the processing quickly, the system can also place the computing unit 12 in the cloud, as shown in FIG. 4, and improve the computing capability by the computing unit 4 composed of multiple servers 40. In addition, the sensors 41 scattered on multiple hospital beds 48 corresponding to different patients can also be received through the network 49 at the same time, so as to monitor the status of multiple patients at the same time. In this way, when the condition of a certain hospital bed satisfies the above-mentioned conditions for issuing a warning signal, a corresponding warning signal containing bed information can be sent to a portable device 42 and then displayed by the portable device 42 Ability to inform the medical staff of the abnormal risk of the tube falling off of the patient in that bed. Moreover, multiple portable devices can receive the warning signal to achieve the purpose of information sharing and common care.

In summary, this system will construct a smart predictive device for clinical tubing shedding, allowing nursing staff to immediately discover the risk of extubation in elderly patients, which will benefit the clinical practice of reducing tubing shedding, as well as clinical practicability. Cost-effectiveness to reduce medical disputes. This case will collect physical measurements in various pipelines (such as endotracheal tube, intravenous injection tube, hemodialysis pipeline, etc.) and physical quantities in the pipeline, and use patient's different background data (including age, pain, etc.) Index, hospitalization days, etc.), more accurate dynamic feedback to determine the status of different types of patients, so as to collect and analyze Digital Biomarker (Digital Biomarker), to achieve real-time warning. It is also possible to use machine learning in artificial intelligence to train the system in this case to achieve the purpose of optimizing the accuracy of judgment.

Although the present invention has been disclosed in the preferred embodiment as above, it is not intended to limit the present invention. Anyone familiar with the art can make some changes and modifications without departing from the spirit and scope of the present invention. Therefore, the present invention The scope of protection shall be subject to those defined by the appended claims.

10: Pipeline 113: Sound sensing device 1100: Multi-axis sensor 1101: optical sensor 1102: Magnetic Sensor 1103: Capacitive Sensor 40: server 4: Operation unit 49: Network 48: hospital bed 41: Sensor 42: portable device

1A and 1B are functional block diagrams of various embodiments of the pipeline abnormal event prediction device developed in this case. Figure 2 is a schematic diagram of the functional block inside the motion sensor of this case. Figure 3 is a schematic diagram of the pipeline abnormal event prediction method in this case. Figure 4 is a functional block diagram of the pipeline abnormal event prediction system in this case.

10: Pipeline

1: Pipeline abnormal event prediction device

11: Sensor

12: arithmetic unit

110: Motion sensor

111: Image capture and analysis device

112: Pipeline flow detector

113: Sound sensing device

Claims (16)

A pipeline abnormality risk prediction method, applied to a patient using a pipeline, the method includes the following steps: Detect a disturbance state of the pipeline or a specific action of the patient; Calculate a pipeline abnormality risk index based on the disturbance state of the pipeline or the specific action or sound of the user; and According to the abnormal risk index of the pipeline falling within one of a plurality of numerical ranges, correspondingly, one of a plurality of warning signals with different contents is selected and output. The pipeline abnormality risk prediction method described in item 1 of the scope of patent application, wherein the disturbance state is any or all of the appearance change of the pipeline and the change of the fluid flow rate in the pipeline. The pipeline abnormality risk prediction method described in item 1 of the scope of the patent application, wherein the specific action of the patient is that the patient moves his free limb toward the pipeline and the patient’s free limb has an abnormal movement trajectory Any or all of them. According to the pipeline abnormality risk prediction method described in item 1 of the scope of patent application, the pipeline abnormality risk index is calculated according to the disturbance state of the fluid pipeline and a patient's background data For example, the pipeline abnormality risk prediction method described in item 1 of the scope of patent application, wherein the pipeline abnormality risk index is calculated according to the specific action of the user and a patient's background data. For example, the pipeline abnormality risk prediction method described in item 1 of the scope of the patent application, wherein the pipeline abnormality risk index falls within one of a plurality of numerical ranges. The first range represents that the patient pulls the pipeline or the free limb shakes violently, which corresponds to Corresponding to a first warning signal, it is recommended that caregivers use heavy sedatives and restraint bands on the patient. The abnormal risk index of the pipeline falls within one of a plurality of numerical ranges. The second range represents obvious movement or illness of the pipeline. The patient's free limb has a large movement, corresponding to a second warning signal. It is recommended that the caregiver use a light tranquilizer on the patient and consider using a restraint band. The abnormal risk index of the pipeline falls within one of a plurality of numerical ranges. The third range represents slight movement of the pipeline or abnormal trajectory of the patient's free limbs. Correspondingly, a third warning signal is issued. It is recommended that the caregiver strengthen observation of the patient and ask about the patient's condition. In the pipeline abnormality risk prediction method described in item 1 of the scope of patent application, the specific action of the user is to make a sound. A pipeline abnormality risk prediction device, which is applied to patients who use a pipeline. The device includes: A sensor arranged near the pipeline to detect a disturbance state of the pipeline or a specific action of the patient; and An arithmetic unit, signal connected to the sensor, calculates a pipeline abnormality risk index according to the disturbance state of the pipeline or the specific action of the patient, and falls into a plurality of pipeline abnormality risk indexes according to the pipeline abnormality risk index One of the numerical ranges, and correspondingly select one of a plurality of warning signals with different contents to output. The pipeline abnormality risk prediction device described in item 8 of the scope of patent application, wherein the sensor includes a motion sensor or an image capture and analysis device, and the detected disturbance state is the appearance change of the pipeline . According to the pipeline abnormality risk prediction device described in item 8 of the scope of patent application, the sensor includes a pipeline flow detector, and the detected disturbance state is the change of the fluid flow rate in the pipeline. The pipeline abnormality risk prediction device described in item 8 of the scope of patent application, wherein the specific action of the patient detected by the sensor is that the patient moves his free limb toward the pipeline and the patient Of the free limbs have any or all of the abnormal motion trajectories. The pipeline abnormality risk prediction device according to item 8 of the scope of patent application, wherein the calculation unit calculates the pipeline abnormality risk index according to the disturbance state of the fluid pipeline and a patient background data According to the pipeline abnormality risk prediction device described in item 8 of the scope of patent application, the calculation unit calculates the pipeline abnormality risk index according to the specific action of the user and a patient's background data. For example, the pipeline abnormality risk prediction device described in item 8 of the scope of patent application, wherein the pipeline abnormality risk index falls within one of a plurality of numerical ranges. The first range represents that the patient pulls the pipeline or violently shakes the free limb. The arithmetic unit correspondingly sends out a first warning signal. It is recommended that the caregiver use severe tranquilizers and restraint bands on the patient. The abnormal risk index of the pipeline falls within one of the multiple numerical ranges. The second range represents the obvious pipeline. If the patient moves or the patient’s free limb has a large movement, the arithmetic unit correspondingly sends out a second warning signal. It is recommended that the caregiver use a light tranquilizer on the patient and consider using a restraint band. The abnormal risk index of the pipeline falls in the plural One of the numerical ranges, and the third range represents slight movement of the pipeline or abnormal trajectory of the patient’s free limbs. The computing unit correspondingly sends out a third warning signal. It is recommended that the caregiver pay more attention to the patient and ask the patient. situation. For example, in the pipeline abnormality risk prediction device described in item 8 of the patent application, the specific action of the user is to emit a sound, and the sensor is used to detect the sound to form the intensity and spectrum data representing the sound. A pipeline abnormality risk prediction system, applied to multiple patients using multiple beds, the system includes: A plurality of sensors are respectively arranged near the plurality of pipelines corresponding to the plurality of patients, and are used to detect the disturbance state of the plurality of pipelines or the specific actions of the plurality of patients; and An arithmetic unit, connected to the plurality of sensors via a network, respectively calculates corresponding plurality of pipeline abnormal risk indexes according to the disturbance state of the plurality of pipelines or the specific actions of the plurality of patients, and According to the plurality of pipeline abnormal risk indexes, any one of the indexes falls within one of the plurality of numerical ranges, and correspondingly, one of the plurality of warning signals with different contents is selected and output.

Images