KR102518105B1 - The ai integration monitoring system and method for the mdical equipments - Google Patents

Abstract

The present invention is an information collection unit that collects status information provided by each medical device and daily inspection information provided by a manager terminal, and an information classification unit that classifies and stores the information collected by the information collection unit by medical device and information type. , a failure prediction learning unit that predicts whether or not a component for each medical device fails and whether or not there is a failure for the operating state using the status information of each medical device and learns the predicted result, and for each inspection item of the daily inspection information Stores the status diagnosis unit that determines whether or not there is an abnormality for each inspection item through the result value, the information classified by the information classification unit, the information output from the failure prediction learning unit, and the information output from the status diagnosis unit A device/part information storage unit that stores the warranty time or the number of times of warranty use for each part of each medical device and stores the reference state for each type of state information, and the diagnosis result of the condition diagnosis unit and the failure It relates to an AI integrated monitoring system and method for medical equipment including a notification unit for generating failure guide information by receiving a prediction result of a prediction learning unit and providing the generated failure guide information.

Description

AI integrated monitoring system and method for medical equipment {THE AI INTEGRATION MONITORING SYSTEM AND METHOD FOR THE MDICAL EQUIPMENTS}

The present invention relates to a technology for monitoring the operation and condition of medical equipment, and in particular, to an AI integrated monitoring system and method for predicting failure and diagnosing the condition of medical equipment based on artificial intelligence (AI). will be.

Due to the advancement of the industry and the development of information and communication technology, the medical field is also attempting to combine with ICT (Information & Communication Technology) technology, and accordingly, medical devices in hospitals are being digitalized and networked.

Based on this, a large amount of patient medical information is collected and systematically managed based on ICT, or the patient's condition is monitored remotely in real time by digitizing medical equipment and tools. In addition, according to Korean Patent Publication No. 2019-0023023, a conventional monitoring technique for determining whether there is an abnormality in medical imaging devices and notifying when an abnormality is found is proposed.

However, technology for predicting failures of various medical devices in advance, diagnosing the condition, and providing information enabling correct handling has not been provided.

The problem to be solved by the present invention is to provide an AI integrated monitoring system and method for medical equipment that predicts failures of various medical equipment in advance, diagnoses the condition, and provides information enabling correct response.

In addition, the problem to be solved by the present invention is to provide an AI integrated monitoring system and method for medical equipment that diagnoses conditions and predicts failures based on AI to provide more accurate results.

In addition to the above tasks, embodiments according to the present invention may be used to achieve other tasks not specifically mentioned.

AI integrated monitoring system for medical equipment according to an embodiment of the present invention for solving the above problems is an information collection unit that collects status information provided by each medical device and daily inspection information provided by a manager terminal, and the information collection unit An information classification unit that classifies and stores the collected information by medical device and type of information. Using the status information of each medical device, it predicts whether or not there is a failure of the parts and operation status of each medical device, and predicts A failure prediction learning unit that learns the result, a condition diagnosis unit that identifies whether or not there is an abnormality for each inspection item through the result value of each inspection item of the daily inspection information, the information classified by the information classification unit, and the failure prediction A collection information storage unit that stores the information output from the learning unit and the information output from the condition diagnosis unit, stores the warranty time or number of times of warranty use for each part of each medical device, and stores the reference state for each type of status information and a device/part information storage unit, and a notification unit that generates failure guide information by receiving the diagnosis result of the state diagnosis unit and the prediction result of the failure prediction learning unit, and provides the generated failure guide information.

The failure prediction learning unit predicts a failure if the usage time of the part included in the status information exceeds the warranty period or is within the set margin period, and the usage count of the part included in the status information exceeds the set usage count or is within the set margin count. If it is, it is predicted as a failure, or a failure is predicted if the current state is different from the reference state by comparing the current state of the action item included in the state information with the reference state.

The notification unit analyzes the relationship between the items determined to be abnormal in the status diagnosis unit, or the correlation between the items determined to be abnormal in the status diagnosis unit and the failure prediction result received from the failure prediction learning unit. Analysis is performed to generate the failure guide information.

The AI integrated monitoring system for medical equipment according to an embodiment of the present invention ensures that only information stored in the collected information storage unit is stored within a certain period of time, and if there is information exceeding a certain period of time, the excess information is extracted to a cloud server. It may further include an information management unit to be stored in.

AI integrated monitoring method for medical equipment according to an embodiment of the present invention for solving the above problems is the step of collecting the status information provided by each medical device by the information collection unit and the daily inspection information provided by the manager terminal, the information collection unit Step of classifying and storing the collected information by medical device and type of information by the information classification unit, using the status information of each medical device classified and stored by the information classification unit, the failure prediction learning unit of parts for each medical device Predicting whether or not there is a failure and operating state and learning the predicted result, Using the result predicted and learned by the failure prediction learning unit, the status diagnosis unit determines the result value for each inspection item of the daily inspection information. determining whether or not there is an abnormality for each inspection item through a step, and generating failure guide information by a notification unit receiving the diagnosis result of the condition diagnosis unit and the prediction result of the failure prediction learning unit, and providing the generated failure guide information. includes

According to an embodiment of the present invention, it is possible to prevent failure and quality deterioration ultimately by predicting failure through condition diagnosis of medical devices owned by medical institutions, hospitals, and the like.

In addition to the above tasks, embodiments according to the present invention may be used to achieve other tasks not specifically mentioned.

1 is a conceptual diagram of an integrated AI monitoring technology for medical equipment according to an embodiment of the present invention.
2 is a block diagram of an integrated AI monitoring system for medical equipment according to an embodiment of the present invention.
3 is a diagram showing daily inspection information used in the AI integrated monitoring system for medical equipment according to an embodiment of the present invention.
4 is a diagram showing detailed daily inspection information used in the AI integrated monitoring system for medical equipment according to an embodiment of the present invention.
5 is a flowchart showing an AI integrated monitoring method for medical equipment according to a first embodiment of the present invention.
6 is a flow chart showing an integrated AI monitoring method for medical equipment according to a second embodiment of the present invention.

Hereinafter, with reference to the accompanying drawings, embodiments of the present invention will be described in detail so that those skilled in the art can easily carry out the present invention. The present invention may be embodied in many different forms and is not limited to the embodiments described herein. In order to clearly describe the present invention in the drawings, parts irrelevant to the description are omitted, and the same reference numerals are used for the same or similar components throughout the specification. In addition, in the case of widely known known technologies, detailed description thereof will be omitted.

In this specification, when a certain component is said to "include", it means that it may further include other components, not excluding other components unless otherwise stated.

Hereinafter, an AI integrated monitoring system and method for medical equipment according to an embodiment of the present invention will be described with reference to the accompanying drawings.

1 is a conceptual diagram of an integrated AI monitoring technology for medical equipment according to an embodiment of the present invention. Referring to FIG. 1, the AI integrated monitoring technology for medical equipment according to an embodiment of the present invention is an AI-based monitoring system 100 (hereinafter referred to as "monitoring system ( 100) is achieved by).

At this time, the medical device 10 is all types of medical imaging devices connected to the monitoring system 100 through a network, and X-ray tube related devices such as digital diagnostic X-ray imaging devices and mobile X-ray imaging devices, and CT , MRI, imaging devices that take pictures of the patient's body and organs, detector-related devices, console-related devices, table-related devices, and PACS (Picture Archiving Communication System)-related devices. Among the medical imaging devices, a raspberry pi is installed in a device that does not have a digital processing processor device to enable data transmission.

In addition, the status information about the medical device 10 collected by the monitoring system 100 is information transmitted from the medical device 10, and includes identification information of the medical device 10 and parts of the medical device 10. Usage information for the medical device 10, operation status of the medical device 10 (e.g., image capture speed, short distribution of the tube, horizontality of the instrument (table, etc.), number of shots, use information such as the number of times each part is used, This is information that can identify tube temperature, current location information (corresponding to mobile medical devices), quality of captured images, etc. Status information is provided to the monitoring system 100 as log information.

The daily checkup includes information sent from the medical device 10 itself (automatic information) and information input and operated by the manager and transmitted from the manager terminal (manual information). As shown in FIG. 3, among daily inspection information, automatic information includes items such as shielding-related overall inspection information and result values for the items. For such automatic information, each medical device 10 is equipped with a sensor for detecting corresponding information.

Manual information, as shown in FIG. 4, checks the main power, checks the normal operation of the timer, checks the normal operation of the function key, checks the noise and vibration status of the device, checks whether there is an abnormal overheating of the device, checks the door status, and inputs the RPM. Includes items such as check, time input check, and the result value for the item.

Upon receiving the status information and daily inspection information, the monitoring system 100 predicts failure using the status information and diagnoses the state of the medical device 10 using the status information and daily inspection information. In addition, the monitoring system 100 generates information on countermeasures according to the failure prediction result and the diagnosis result, and transmits the information to the corresponding medical device 10 or/and the manager.

On the other hand, the monitoring system 100 according to an embodiment of the present invention stores the daily state information and daily inspection information for each medical device 10 in its own storage, and the corresponding failure prediction result and diagnosis result for a certain period of time. and data (information) after a certain period of time is uploaded to the cloud server 200 and managed.

Hereinafter, an example of implementing the concept of the present invention described with reference to FIG. 1 will be described with reference to FIG. 2 .

2 is a block diagram of an integrated AI monitoring system for medical equipment according to an embodiment of the present invention. Referring to FIG. 2 , the monitoring system 100 according to an embodiment of the present invention includes an information collection unit 110, an information classification unit 120, a current location determination unit 130, a failure prediction learning unit 140, a state It includes a diagnosis unit 150, an information management unit 160, a collection information storage unit 170, a device/part information storage unit 180, and a notification unit 190.

The information collection unit 110 collects status information and daily inspection information transmitted from each of the medical devices 11 to 14 and daily detailed inspection information transmitted from the manager terminal 20 . The information classification unit 120 classifies the information collected by the information collection unit 110 by medical device and by type of information, and stores it in the collection information storage unit 170 .

The current location detection unit 130 determines current location information transmitted from a mobile medical device among medical devices. The current location determining unit 130 may be omitted in some cases.

The failure prediction learning unit 140 predicts the failure of each medical device using the state information of each medical device classified by the information collection unit 110 and learns the predicted result. Failure prediction learning unit 140, if the state information is the warranty time (usage time) of the part or the number of times of use for each part, the warranty time or warranty set in response to the unique identification information of the corresponding part stored in the device/part information storage unit 180 When the set warranty time is exceeded compared to the number of uses or close to the warranty period (ie, within the margin period), a failure prediction warning is issued through the notification unit 190 because failure is highly likely to occur.

In addition, the failure prediction learning unit 140 stores device/part information for a reference state suitable for each type of information if the state information is information indicating an operating state (temperature, level of equipment, number of shots, quality of a shot image, etc.) The unit 180 reads and compares, and when the current state is different from the reference state, a failure prediction warning is issued through the notification unit 190.

By this failure prediction learning unit 140, it is possible to predict the life of the tube, and it is possible to determine that the quality of the image has deteriorated through comparison between the quality of the current captured image and the image captured immediately before. Fault guide information can be provided based on the identified result by figuring out the reason through items that show state information different from the reference state (level of equipment, tube operation, detector operation, tube temperature, etc.).

For example, the failure prediction learning unit 140 knows through learning that the equipment used for shooting should be horizontal in a reference state, but if the equipment is not horizontal, the image quality deteriorates, and the tube is evenly shorted. This is the standard condition, and if it is not, it is known through learning that the video quality deteriorates, and the video shooting speed is the set speed, which is the reference condition. You can know through learning.

In addition, when the failure prediction learning unit 140 shoots how much, that is, when the number of shots reaches a certain number of times, the tube life ends, how much time it takes to recalibrate, and the current tube temperature is a few degrees, but it is stable It is possible to predict how much cooling time is required for this through learning.

The condition diagnosis unit 150 receives and analyzes the daily inspection information and daily detailed inspection information, identifies inspection items that are determined to be abnormal (or malfunctioning) among the daily inspection information and daily detailed inspection information, and diagnoses the state of each medical device. do.

The information management unit 160 manages the information stored in the collection information storage unit 170 in association with the cloud server 200 . That is, the information management unit 160 stores the information stored in the collection information storage unit 170 only within a certain period of time, and if there is information exceeding a certain period of time, the excess information is extracted and stored in the cloud server 200. let it

The collection information storage unit 170 combines the information classified by the information classification unit 120, the information output from the failure prediction learning unit 140, and the information output from the condition diagnosis unit 150 into the identification information of the corresponding medical device. matched and stored.

The device/parts information storage unit 180 stores the warranty time or the number of guaranteed uses for each part of each medical device, and stores a reference state for each type of state information.

The notification unit 190 receives the diagnosis result of the condition diagnosis unit 150 and the prediction result of the failure prediction learning unit 140, and receives prescription information, that is, failure guide information (eg, corresponding part replacement notification, abnormal parts/condition notification, parts prediction life notification, etc.) is generated and the generated failure guide information is transmitted to the corresponding medical device or/and the manager terminal 20 to inform it.

On the other hand, the notification unit 190 analyzes the relationship between the items determined to be abnormal in the status diagnosis unit 150, such as correlation, or the items determined to be abnormal in the status diagnosis unit 150 and the failure prediction learning unit ( 140), failure guide information, which is a diagnosis (check) result, may be generated by analyzing the relationship between failure prediction results received.

Hereinafter, an integrated AI monitoring method for medical equipment according to an embodiment of the present invention will be described with reference to FIGS. 5 and 6 .

5 is a flowchart showing an AI integrated monitoring method for medical equipment according to a first embodiment of the present invention. Referring to FIG. 5 , the monitoring system 100 collects state information from each of the medical devices 11 to 14 (S501), and classifies and stores the collected state information by device and type (S502).

The monitoring system 100 starts predicting a failure using state information for each classified device and for each type (S503).

The failure prediction operation includes failure prediction using the warranty period (warranty time) for each part, failure prediction according to the number of times each part is used, and failure (abnormality) discovery or prediction through comparison between the previous state and the reference state.

Describing the failure prediction process using the warranty period (guarantee time) for each part, the monitoring system 100 checks the warranty period of each part based on the state information of the part through pre-stored information (S504), and The period of use of each part included in the information is compared with the warranty period of each pre-stored part, and it is determined whether there are parts whose period of use exceeds the period of warranty or is close to the period of guarantee (ie, within the set margin period) (S505).

Here, the monitoring system 100 may obtain life information of each part through the determination process.

On the other hand, the monitoring system 100 identifies the part if the usage time exceeds the warranty period or there is a part within the margin period (S506). Then, the monitoring system 100 repeats the steps S505 and S506 to be performed for all parts having status information (S507).

The monitoring system 100 generates a failure prediction result, that is, failure guide information for the parts identified through steps S505 to S507, and informs the corresponding medical device or/and manager terminal 20 (S508).

Next, the failure prediction process according to the number of times of use of each part is described. The monitoring system 100 identifies the number of times of use of each part through status information including information on the number of times of use of each part (S510), and pre-stored It is compared with the guaranteed use count of each part, and it is determined whether there is a part whose use count exceeds the guaranteed use count or is close to the guaranteed use count (ie, within the set margin count) (S511). Here, the monitoring system 100 may obtain life information of each part through the determination process.

The monitoring system 100 identifies the part if the number of uses exceeds the number of guaranteed uses or if there is a part within the number of margins (S506). Then, the monitoring system 100 repeats the steps S511 and S506 to be performed for all parts having status information on the number of uses (S507).

The monitoring system 100 generates failure guide information, which is a failure prediction result, for the parts identified through steps S511, S505, and S507, and informs the corresponding medical device or/and manager terminal 20 (S508).

Finally, describing the process of discovering or predicting a failure (abnormality) through comparison between a previous state and a reference state, the monitoring system 100 uses state information related to operation items of a medical device rather than information related to parts among state information. It is read from the collection information storage unit 170, and the current state of each item is compared with the reference state (S512), and it is determined whether there is a malfunction or abnormal item (S513).

When the monitoring system 100 determines that there is a failure (or anomaly) in the process S513, it identifies an operation item with a failure (or anomaly) (S515). A failure prediction result corresponding to the identified operation item is generated (S516) and provided to the corresponding medical device or/and manager terminal 20 (S517).

For example, if it is determined that the image capturing speed is slower than the reference state in the item for the image capturing speed, the failure is determined, the monitoring system 100 generates a failure prediction result of "the chain or the motor is abnormal and needs to be inspected". As another example, when the tube temperature is higher than the reference temperature in the tube temperature item, the monitoring system 100 generates a failure prediction result of "1 minute cooling time required".

Meanwhile, the monitoring system 100 repeats steps S512 and S513 to be performed for status information of all items.

The monitoring system 100 learns the failure prediction result related to the life of the part provided through the process S508 and the failure prediction result for each item provided through the process S517 to quickly and accurately predict the failure of the state information that is input thereafter. (S509).

Hereinafter, a diagnosis process for daily inspection items and daily detailed inspection items will be described with reference to FIG. 6 . 6 is a flow chart showing an integrated AI monitoring method for medical equipment according to a second embodiment of the present invention.

The monitoring system 100 collects daily inspection information and daily detailed inspection information from each of the medical devices 11 to 14 and the manager terminal 20 (S601), and determines whether there is an abnormal inspection item (S602). An abnormal inspection item is identified (S603).

The monitoring system 100 repeats steps S602 and S603 until all items are performed.

The monitoring system 100 correlates items with abnormalities. And a correlation between the items with abnormality and the failure prediction result identified by the failure prediction learning unit 140 is identified (S605), and failure guide information, which is a diagnosis (check) result, is generated to correspond to the medical device or/and manager terminal (20). ) is provided.

Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements made by those skilled in the art in the field to which the present invention belongs are also the rights of the present invention. belong to the range

100: monitoring system 110: information collection unit
120: information classification unit 130: current location determination unit
140: failure prediction learning unit 150: status diagnosis unit
160: information management unit 170: collection information storage unit
180: device/part information storage unit 190: notification unit

Claims (8)

An information collection unit that collects status information provided by each medical device and daily inspection information provided by the manager terminal;
An information classification unit that classifies and stores the information collected by the information collection unit by medical device and by type of information;
A failure prediction learning unit that predicts whether or not a component for each medical device fails and whether or not there is a failure for an operating state using the state information of each medical device, and learns the predicted result;
A condition diagnosis unit that determines whether or not there is an abnormality for each inspection item through the result value of each inspection item of the daily inspection information;
A collection information storage unit for storing the information classified by the information classification unit, the information output from the failure prediction learning unit, and the information output from the status diagnosis unit;
A device/parts information storage unit that stores the warranty time or the number of times of warranty use for each part of each medical device and stores the standard state for each type of state information; and
And a notification unit for generating failure guide information by receiving the diagnosis result of the state diagnosis unit and the prediction result of the failure prediction learning unit, and providing the generated failure guide information.
AI integrated monitoring system for medical equipment. According to claim 1,
The failure prediction learning unit predicts a failure if the usage time of the part included in the status information exceeds the warranty period or is within the set margin period, and the usage count of the part included in the status information exceeds the set usage count or is within the set margin count. To predict if this is a failure,
AI integrated monitoring system for medical equipment. According to claim 2,
The failure prediction learning unit compares the current state of the action item included in the state information with a reference state and predicts a failure when the current state is different from the reference state.
AI integrated monitoring system for medical equipment According to claim 1,
The notification unit analyzes the correlation between the items determined to be abnormal in the status diagnosis unit, or analyzes the correlation between the items determined to be abnormal in the status diagnosis unit and the failure prediction result received from the failure prediction learning unit. generating failure guide information;
AI integrated monitoring system for medical equipment. According to claim 1,
Further comprising:
AI integrated monitoring system for medical equipment. Collecting status information provided by each medical device and daily inspection information provided by the manager terminal by the information collection unit;
Classifying and storing the information collected by the information collection unit by the information classification unit by medical device and by type of information;
Using the state information of each medical device classified and stored by the information classification unit, a failure prediction learning unit predicts whether a component for each medical device fails and whether a failure occurs in an operating state, and learns the predicted result;
Using the results predicted and learned by the failure prediction learning unit, the state diagnosis unit determines whether or not there is an abnormality for each inspection item through the result value for each inspection item of the daily inspection information, and
A notification unit receiving a diagnosis result of the status diagnosis unit and a prediction result of the failure prediction learning unit, generating failure guide information, and providing the generated failure guide information.
AI integrated monitoring method for medical equipment. According to claim 6,
In the step of predicting failure and learning the predicted result, if the usage time of the parts included in the status information exceeds the warranty period or is within the set margin period, it is predicted to be faulty, and the number of uses of the parts included in the status information is set. If the number of uses is exceeded or within the set margin, it is predicted to be a failure.
AI integrated monitoring method for medical equipment. According to claim 7,
The step of predicting whether or not the failure is and learning the predicted result is to compare the current state of the action item included in the state information with the reference state to predict failure when the current state is different from the reference state,
AI integrated monitoring method for medical equipment.

Images