TWI753613B - Medical equipment verification method and verification system realizing the verification method - Google Patents

Abstract

The invention discloses a medical equipment verification system, which is characterized in that: a controller configures the operation of a simulator, and the controller sends test cases to the simulator; the simulator converts the received test cases into corresponding physiological analog signals, and sends the physiological analog signals to an equipment under test; the equipment under test executes test work sequentially according to the received physiological analog signals; a video and audio capture device is used to capture the test process and test results of the equipment under test, and transmit the captured video and audio to the controller; the controller compares the capture content according to the specifications of the test cases, to determine whether the equipment under test meets the requirements. The medical equipment verification system of the present invention has the following advantages: the controller is used to analyze and judge the test results, reduce labor costs, and the results are consistent and repeatable; the video and audio capture device is used to capture the verification process and verification results of the equipment under test, which solves the problem that the equipment can only be verified when the communication protocol of the instrument is known.

Description

Detection method for medical equipment and detection system for realizing the detection method

The invention belongs to the field of instrument detection, and particularly relates to a detection method for medical equipment and a detection system for realizing the detection method.

With the development of science and technology, various medical equipments play an extremely important role in the diagnosis, treatment and scientific research of medical institutions. Medical equipment is an important means and basis for hospitals to carry out diagnosis, testing, surgery, treatment and other links. Therefore, the quality of medical equipment directly affects the reliability and effectiveness of disease diagnosis and treatment, as well as the life and health of patients. . For medical equipment, various countries have formulated different standards and regulations. Medical equipment manufacturers need to conduct strict inspections before medical equipment leaves the factory to meet these standards and regulations; at the same time, hospitals also need to use these instruments before using them. Acceptance and regular inspection of medical equipment to ensure that these medical equipment can work normally.

At present, there are mainly the following problems in the field of medical device testing: (1) There are many standards and regulations on medical devices, and it takes a long time to research regulations and develop test tools. Taking ECG equipment as an example, relevant industry standards include YY1079: ECG Monitor, YY1139: ECG Diagnostic Equipment, YY0782: Recording and Analysis Single- and Multi-Channel ECG Machine, YY0885: Holter Monitoring System Safety and Basic Performance ; National metrological verification procedures include JJG760: ECG monitor verification procedures, JJG1042: Dynamic (movable) ECG machine verification procedures, JJG1163: Multi-parameter monitor verification procedures, etc. The above standards and regulations need to be met at the same time before the medical equipment leaves the factory and in the process of hospital acceptance and periodic verification. Therefore, it takes a long time to study these standards and regulations, and it also takes a long time to develop corresponding test tools for these standards and regulations. time. (2) The current medical equipment detection relies on manual detection. First, in the manual detection process, due to factors such as cumbersome operation, heavy workload, and long detection time, detection errors caused by personnel fatigue are prone to occur; The level of personnel and the judgment standards are not consistent, and the test results rely on manual judgment. Even the same test personnel are prone to the problem that the results are not repeatable. (3) At present, there are automatic detection equipment for medical equipment on the market, but because of the great differences in the interfaces and protocols of medical equipment of different manufacturers, these automatic detection equipment can only detect the products of the manufacturers they cover, not Has broad versatility.

In view of the above problems of medical equipment, there is a great need for a medical equipment detection method and a detection system for realizing the detection method, which can effectively realize automatic detection, meet the requirements of standards and regulations, and can adapt to the equipment of various manufacturers.

In order to solve the above existing technical problems, the present invention provides a medical equipment detection system, which is characterized by comprising a controller, a simulator, a device to be tested, and an audio and video capture device, wherein the controller configures the operation of the simulator, The controller sends a test case to the simulator; the simulator converts the received test case into a corresponding physiological simulation signal, and sends the physiological simulation signal to the device under test; the under test The device performs the test work in sequence according to the received physiological analog signal; the video and audio capture device is used for photographing the test process and test result of the device under test, and transmits the photographic sound to the controller; the controller according to Test case specifications, compare the content of the shooting audio, and determine whether the device under test meets the requirements. The detection system provided in the present invention has the following advantages: (1) The controller is used to analyze and judge the test results, reducing labor costs, and the results no longer rely on manual interpretation, and the analysis results are consistent and repeatable (2) The detection process and detection results of the equipment to be tested are photographed by an audio and video capture device, which solves the problem that equipment detection can only be performed by knowing the communication protocol of the instrument at present. Therefore, the detection system in the present invention has better adaptability In addition, the use of audio and video capture devices for video and audio shooting is the same as the way medical staff observe medical equipment, which is closer to the actual use situation. Compared with the single way of using a communication protocol for numerical comparison, the detection method in the present invention The system can more comprehensively test the equipment to be tested.

Preferably, the controller includes a detection item definition and selection module, a deep learning module, and a data storage and analysis module; the detection item definition and selection module is used to edit and store data compiled for different equipment and different regulations. Test case; the deep learning module is used for training image information and sound information to obtain an image deep learning model and a sound deep learning model; the data storage and analysis module is used for storing the deep learning model after training and test case specifications, and analyze the device's shooting audio data according to the deep learning model.

Preferably, the controller further includes a data management module for generating, analyzing and saving electronic reports. The data management module can automatically generate and save electronic reports, and can issue reports in real time, shortening the time from detection to issuing reports, greatly improving the efficiency of production and testing, and relevant supervisors can see the reports in time and deal with corresponding problems in a timely manner. processing to improve the operating efficiency of the equipment. In addition, the electronic report adopts a paperless form, which can selectively print the required report content, which reduces the use of paper and is more energy-saving and environmentally friendly.

Preferably, the test cases include single-item test cases for testing a certain item, continuous test cases for standards or verification procedures, continuous test cases for basic signal generation, and test cases customized according to test requirements.

Preferably, the standard or verification procedure includes at least one of the following standards or verification procedures: YY1079, YY1139, YY0782, YY0885, JJG760, JJG1042, JJG1163, JJG93, JJG1041, JJG1149, JJF1260, IEC60601-2-25, IEC60601- 2-27, IEC60601-2-47, IEC60601-2-4, JJF1260, IEC60601-2-21, YY1259, YY0585, IEC60601-2-24, IEC60601-2-12, JJF1234, IEC60601-2-13, YY1147, EC11, EC13, EC38, EC57. Built-in standards or testing procedures in continuous test cases reduce the time for companies or hospitals to study regulations, and companies or hospitals no longer need to develop corresponding testing tools for these standards and regulations, saving production and testing costs.

Preferably, the basic signal includes at least one of the following signals: sine wave, unipolar square wave, bipolar square wave, unipolar triangular wave, bipolar triangular wave, square pulse wave, triangular pulse wave.

Preferably, when the controller compares the content of the video recordings according to the test case specifications, and determines whether the device under test meets the requirements, the following steps are included: the controller uses a trained deep learning model to extract technical features from the content of the video recordings, Infer the type and location information of the image in the shooting sound, as well as the type and volume of the sound, and compare the extracted features with the test case specifications to determine whether the device under test meets the requirements.

Preferably, the simulator includes a heart and brain electrical simulation module, a blood pressure simulation module, a blood oxygen simulation module, an invasive blood pressure simulation module, a carbon dioxide simulation module, a carbon monoxide simulation module and a temperature simulation module.

Preferably, the medical equipment detection system further includes an elastic frame, the elastic frame is used for clamping the video and audio capture device and adjusting the position between the video and audio capture device and the device to be tested.

Preferably, the elastic frame includes a beam, a clamping structure for the device to be tested, a rotating arm, a clamping structure for an audio and video capture device, a rotating structure and a locking structure, the bottom surface of the beam is connected to the clamping structure for the equipment to be tested, and one end of the rotating arm is The rotating structure is connected to the upper surface of the beam, the other end of the rotating arm is connected to the clamping structure of the audio and video capture device, and the locking structure is arranged on the side of the beam, wherein the clamping structure of the device to be tested includes a first clamping arm, a second clamping An arm and two pulleys, the first clamping arm and the second clamping arm and the beam are connected in a sliding manner for adjustment according to different widths of the equipment to be tested.

The present invention also provides a medical device detection method for the above-mentioned medical device detection system, which is characterized by comprising the following steps: Step S1: the controller receives the detection parameters; Step S2: the controller sends a test case to the simulator; Step S3: simulates The device generates a physiological analog signal according to the test case, and sends it to the device under test; Step S4: the device under test performs the test according to the received physiological analog signal; The shooting audio is transmitted to the controller; step S6: the controller compares the shooting audio content according to the test case specification, and determines whether the device to be tested meets the requirements.

Preferably, the medical device detection method further includes: step S7: generating an electronic detection report according to the test result.

Preferably, after the execution of S1, the following steps are also performed: S11: the controller receives the detection parameters and obtains the detection items; S12: the controller determines whether there are test cases for the detection items stored, and when the controller determines that there are test cases for the detection items When there is no test case for the test item, perform the next step S2: the controller sends the test case to the simulator; when the controller determines that there is no test case for the test item, perform step S13: customize the test case according to the test item and save it to the controller.

Preferably, step S6 specifically includes the following steps: step S61: using the trained deep learning model to extract image features and sound features of the video captured by the video and audio capture device; step S62: predicting the image features according to the extracted image features Category and location information; Step S63: Predict the sound category and volume of the sound according to the extracted sound features; Step S64: Compare the test case specifications to determine whether the device under test meets the requirements.

The invention also provides an elastic frame, the elastic frame includes a beam, a clamping structure for the equipment to be tested, a rotating arm, a clamping structure for an audio and video capture device, a rotating structure and a locking structure, and the bottom surface of the beam is connected to the clamping structure for the equipment to be tested, One end of the rotating arm is connected to the upper surface of the beam via the rotating structure, and the other end of the rotating arm is connected to the clamping structure of the audio and video capture device, and the locking structure is arranged on the side of the beam, wherein the clamping structure of the device to be tested includes a first clamp An arm, a second clamping arm and two pulleys, the first clamping arm, the second clamping arm and the beam are connected in a sliding manner for adjustment according to different widths of the equipment to be tested.

Preferably, the elastic frame further includes a rotation positioning structure, which is arranged on the side of the beam and is used to fix the position between the rotating arm and the beam.

Preferably, several gears are provided on the rotating arm for adjusting the distance between the audio and video capture device and the device to be tested.

Preferably, the clipping structure of the audio and video capture device is provided with several gear positions for adjusting the height of the video and audio capture device relative to the device to be tested.

The elastic frame provided by the present invention has the following advantages: (1) the three directions of the elastic frame (ie, the X, Y and Z axis directions) can be adjusted, which can facilitate the user to find the best shooting angle and improve the quality of the shooting sound; In particular, the image quality is improved; in addition, by adjusting the position of the elastic frame, the standardization of the image can be assisted to avoid the problem that the size and position of the captured image and the training image are too different; (2) the elastic frame can After the storage is completed, the floor space can be effectively saved.

The use of the medical equipment detection method and medical equipment detection system in the present invention has the following advantages: (1) automatic testing is realized, no manual participation is required in the whole process, and a machine is used to collect and judge the results, which can not only improve the detection efficiency, but also improve the test results. It has better consistency and repeatability; (2) The test process and test results are photographed by the audio and video capture device, and the test can be carried out without knowing the communication protocol of each instrument, which broadens the application scenarios of this method and has the advantages of Better applicability; in addition, the use of audio and video capture devices for audio and video shooting is the same as the way medical staff observe medical equipment, which is closer to the actual use situation. Compared with the single way of using communication protocols for numerical comparison, The detection system in the present invention can more comprehensively detect the equipment to be tested; (3) there are continuous test cases for a certain set of standards or verification procedures, which can reduce the time for researching regulations and verification procedures, and does not need to focus on regulations and verification procedures. Procedures for re-development of detection tools, saving time and economic costs.

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments.

FIG. 1 is a schematic structural diagram of a medical device detection system of the present invention. As shown in FIG. 1 , the medical equipment detection system provided in the present invention includes a controller 1 , a simulator 2 , a device to be tested 3 and a video and audio capture device 4 . The controller 1 is respectively connected with the simulator 2 and the audio and video capture device 4 , the simulator 2 is connected with the device under test 3 , and the video and audio capture device 4 is installed near the device under test 3 , and is used for photographing the device under test 3 . Test process and test results. The controller 1 is used to configure the operation of the simulator 2. The controller 1 sends a test case to the simulator 2, and the simulator 2 converts the received test case into a corresponding physiological simulation signal, and sends the physiological simulation signal to the test case. The device 3 and the device to be tested 3 sequentially perform corresponding test work according to the received physiological analog signal. During the process of executing the physiological simulation signal by the device under test 3, the execution result will be displayed and there will be some voice prompts during the execution process. The audio and video capture device 4 is installed near the device under test 3, and is used to photograph the test process and test results. , and transmit these shooting audios to the controller 1. According to the test case specification, the controller 1 compares the content of the video recording to determine whether the device under test 3 meets the requirements.

Specifically, the controller 1 can run the test case, send the content of the test case to the simulator 2, receive and analyze the test results from the audio and video capture device 4, and finally determine whether the device under test 3 is qualified, and generate a test report . When analyzing the test results, the controller can use the trained deep learning model to extract the technical features of the audio and video captured by the audio and video capture device, and infer the type and location information of the image in the captured video, as well as the type and volume of the sound. , and compare the extracted features with the test case specifications to determine whether the device under test meets the requirements. Among them, the test case specifications include waveform characteristics, numbers, text, icons, sound types, volume levels, etc. The controller 1 and the audio and video capture device 4 can be directly connected electrically or via a network, as long as the controller 1 can read the video recording in the video capture device 4 in time. The controller 1 may be a PC, a host or an intelligent terminal for industrial control, or may be other devices with data storage and analysis functions.

The simulator 2 is used to simulate the waveform and data of physiological signals such as ECG, respiration, blood pressure, pulse, body temperature and cardiac output. In this embodiment, the simulator 2 includes an electroencephalogram simulation module, a blood pressure simulation module, a blood oxygen simulation module, an invasive blood pressure simulation module (ie, an IBP simulation module), a carbon dioxide simulation module, and a carbon monoxide simulation module. Group and temperature simulation mods. Simulator 2 is not limited to the above-mentioned modules, and other modules that can simulate physiological signal waveforms and data can also be added. The different modules in Simulator 2 can work individually or together, depending on the actual test case requirements. Each module in the simulator 2 can be controlled by the controller 1 , and each module works according to the related instructions in the test case in the controller 1 . The controller 1 can control any module in the simulator 2, and can also enable several modules in the simulator 2 to work together through the test case requirements.

The device under test 3 is connected to the simulator 2 , and each input interface in the device under test 3 is respectively connected with the output interface corresponding to each module in the simulator 2 . For example, the ECG detection input interface in the device under test 3 is connected with the output interface of the cardiac and EEG simulation module in the simulator 2; the blood pressure detection input interface in the device under test 3 is connected with the blood pressure simulation module in the simulator 2. Output interface connection. According to the difference of the device under test 3, different connection interfaces are selected respectively. The device to be tested 3 can be different devices, such as multi-parameter monitors, static electrocardiographs, dynamic electrocardiographs, ventilators, infant incubators, infant incubators, cooling blankets, infusion pumps, anesthesia machines, electric shocks, automatic external Fibrillators, infusion pumps, etc., are selected to be connected to the module output interface in the simulator 2 according to the required detection. The detection system in this embodiment can simultaneously adapt to different detection equipment, and can effectively solve the problem that the hospital equipment is numerous and scattered, and the regular detection takes a long time.

The audio and video capture device 4 includes a photosensitive electronic component and a sound capture component. The photosensitive electronic components may include Charge Coupled Device (CCD, Charge Coupled Device) and/or Complementary Metal Oxide Semiconductor (CMOS, complementary metal oxide semiconductor) sensors. The photosensitive electronic assembly can use single color, color or multicolor. The sound acquisition component can be a set of microphones or a microphone array. During the instrument detection process, different images and sounds will be produced. Among them, the images include the detection results of the instrument, and these images include the waveform of the physiological signal, the numbers, characters, and icons of the detection results. Taking a multi-parameter monitor as an example, the waveforms displayed on the test result interface include ECG, respiration monitoring waveform, volume wave pulse rate and pulse intensity, and the displayed numbers and text include ECG frequency, blood pressure, blood oxygen, respiration Frequency, body temperature and other information. Sounds include alarm sounds, such as high or low heart rate, stop beating, poor connection of cables, low power and other equipment alarm sounds, as well as some non-alarm sounds, including human sounds, such as speaking , door opening and closing sound, footsteps, etc. The video and audio capture device 4 captures the test process and test results of the instrument under test 3 to facilitate subsequent analysis. Use the video and audio capture device 4 to shoot the test process and test results, so that the system has better adaptability, and the identification and judgment of the results will no longer rely on manual or known protocols for the equipment to be tested, especially for hospitals. In scenarios where there are many and scattered devices, it is no longer necessary to know the communication protocol of each test instrument to carry out the test; in addition, the use of audio and video capture devices for audio and video shooting is the same as the way medical staff observe medical equipment. , which is closer to the actual use situation. Compared with the single approach of using the communication protocol for numerical comparison, the detection system in the present invention can perform comprehensive detection on the device under test.

In order to facilitate the erection and adjustment of the video capture device, the present invention also provides an elastic frame 5 . The structure of the elastic frame 5 is shown in FIG. 2 , FIG. 2 a is a schematic diagram of an unfolded state of the elastic frame 5 , and FIG. 2 b is a schematic diagram of a storage state of the elastic frame 5 . The elastic frame includes a beam 51 , a device-under-test clamping structure 52 , a rotating arm 53 , an audio-video capture device clamping structure 54 , a rotating structure 55 , a locking structure 56 and a rotating positioning structure 57 , which are parallel to the long side of the beam 51 . The direction of the rotating arm 53 perpendicular to the beam 51 is called the Y axis, and the direction parallel to the axis of the rotating structure 55 is called the Z axis.

The bottom surface of the beam 51 is provided with a track parallel to the long side of the beam, preferably a sliding rail. The sliding rail on the bottom surface of the beam 51 is matched with the pulley in the clamping structure 52 of the equipment to be tested, so that the clamping structure 52 of the equipment to be tested can be relatively The parallel movement of the beam 51 along the X-axis direction; the device-under-test clamping structure 52 includes a first clamping arm 520, a second clamping arm 521 and two pulleys, and the first clamping arm 520 and the second clamping arm 521 pass through the pulley and the beam 51. A sliding connection is formed, and the two clamping arms can slide left and right relative to the beam 51. Through this sliding connection, the clamping width of the clamping structure 52 of the device to be tested can be adjusted. In actual use, it can be adjusted according to the size of the screen of the device to be tested. Adaptive adjustment. In addition, this sliding connection method can also realize the storage of the first clamping arm 520 and the second clamping arm 521, thereby reducing the floor space; the rotating arm 53 is connected to the top of the beam through the rotating structure 55, and the rotating arm 53 is close to the axis of the rotating structure 55. One end of the wire is provided with a structure that matches the rotating structure 55 , which is a waist-shaped hole of a certain length in this embodiment. This waist-shaped hole is matched with the rotating structure 55 to make the rotating arm 53 rotate around the axis of the rotating structure 55 , to realize the storage of the rotating arm 53 and effectively reduce the floor space. The other end of the rotating arm 53 is connected to the video capture device holding structure 54, and a plurality of gears are set at one end of the video capture device holding structure 54 for adjusting the position of the video capture device in the Y-axis direction with the instrument The distance between the screens can be adjusted according to different distances. In the invention, three gear positions are set, and the present invention does not limit the number of gear positions. The video capture device clamping structure 54 is in a vertical relationship with the rotating arm 53, and is connected to one end of the rotary arm 53, and its length can be adjusted in the direction parallel to the Z axis, that is, this structure can adjust the relative position of the video capture device 54. Adjustment of the upper and lower positions of the device to be tested 3; the rotating structure 55 is a mushroom-like pin in this embodiment, and the bottom surface of the mushroom head on the top thereof is smooth and contacts the upper surface of the rotating arm 53, so that the surface between the two surfaces is smooth. Having a larger contact area is beneficial to ensure a better effect of parallel rotation of the rotating arm 53 in the plane. In addition, the rotating structure 55 enables the entire elastic frame 5 to be unfolded and retracted; the locking structure 56 is provided on one side of the beam to fix the state of the elastic frame 5 when unfolded or retracted. In this embodiment, a bolt locking structure is adopted, and a movable handle is arranged on the top thereof, so as to make the operator more convenient and labor-saving when rotating the bolt locking frame; On one side, during the unfolding and rotating process of the rotating arm, one side surface of the rotating arm 53 will be in contact with one side surface of the rotating positioning structure 57 to realize the positioning of the rotating position of the rotating arm 53. At this time, it is explained that the rotating arm 53 is perpendicular to the beam 51 The position state of the cross beam 51 and the swivel arm 53 are orthogonal to each other. When the rotating arm 53 is retracted and rotated, the same side surface of the rotating arm 53 will contact the other side surface of the rotating positioning structure 57, so as to realize the storage state in which the rotating arm 53 is parallel to the beam.

The above elastic frame has the following functions: (1) The three directions of the elastic frame (ie X, Y and Z axis directions) can be adjusted, which can facilitate users to find the best shooting angle, improve the quality of shooting sound, especially improve the Image quality; in addition, by adjusting the position of the elastic frame, the standardization of the image can be assisted to avoid the problem that the size and position of the captured image and the training image are too different; (2) the elastic frame can be stored, After the storage is completed, the floor space can be effectively saved.

The video and audio capture device 4 transmits the captured video and audio to the controller 1, and the controller 1 compares the content of the captured video according to the test case specifications to determine whether the device under test 3 meets the requirements.

The detection system in this embodiment has the following advantages: (1) The controller is used to analyze and judge the test results, reducing labor costs, and the results no longer rely on manual interpretation, and the analysis results are consistent and repeatable (2) The detection process and detection results of the equipment to be tested are photographed by an audio and video capture device, which solves the problem that equipment detection can only be performed by knowing the communication protocol of the instrument at present. Therefore, the detection system in the present invention has better adaptability In addition, the use of audio and video capture devices for video and audio shooting is the same as the way medical staff observe medical equipment, which is closer to the actual use situation. Compared with the single way of using a communication protocol for numerical comparison, the detection method in the present invention The system can more comprehensively test the equipment to be tested.

FIG. 3 is a schematic structural diagram of the controller 1 in the present invention. The controller 1 includes a detection item definition and selection module, a deep learning module, a data storage and analysis module, and a data management module.

The test item definition and selection module is used to edit and store test cases prepared for different equipment and different regulations. These test cases include single-item test cases for different equipment to test a certain item, continuous test cases for a set of standards or verification procedures, and test cases that can be customized according to test requirements. These test cases can be selected according to different test situations. For example, during the factory inspection of the equipment, it is necessary to test each index of the equipment to make it meet the requirements of the standard or verification regulations. At this time, continuous test cases can be selected. Execute the test items in the standard or verification procedures in sequence; when the equipment is tested and some indicators are abnormal and need to be re-tested, a single test case can be selected to test a certain test item; when the equipment is in production or debugging In the process, when you need to test some of the test items, you can customize these items to be tested, customize the test cases according to the test requirements, edit and store these test cases in the test item definition and selection module. Among them, continuous test cases include test cases written according to YY1079: ECG monitors, test cases written according to YY1139: ECG diagnostic equipment, YY0782: Test cases written for recording and analyzing single-channel and multi-channel ECG machines, according to YY0885 : Test cases written for the safety and basic performance of the Holter monitoring system; Test cases written according to JJG760: ECG monitor verification procedures; Test cases written according to JJG1042: Dynamic (mobile) ECG verification procedures; JJG1163: Multi-parameter monitoring The test cases written by the instrument verification regulations, etc., also include standards and verification regulations including: JJG93, JJG1041, JJG1149, JJF1260, IEC60601-2-25, IEC60601-2-27, IEC60601-2-47, IEC60601-2-4, JJF1260, IEC60601-2-21, YY1259, YY0585, IEC60601-2-24, IEC60601-2-12, JJF1234, IEC60601-2-13, YY1147, EC11, EC13, EC38, EC57, etc. The continuous test case also includes the continuous test case generated for the basic signal, wherein the basic signal includes sine wave, unipolar square wave, bipolar square wave, unipolar triangular wave, bipolar triangular wave, square wave pulse wave, triangular pulse wave Wait. The test item definition and selection module can save test cases of different equipment, and can customize test cases to adapt to different types of medical equipment and different testing needs, and improve production and testing efficiency.

The deep learning module is used to train image information and sound information, and obtain an image deep learning model and a sound deep learning model. As shown in Figure 4, the deep learning module is divided into an image training sub-module and a sound training sub-module. The image training sub-module uses the deep learning method to train to obtain a deep learning model according to the medical equipment test result images marked with image categories, wherein the image categories include physiological signal waveforms, numbers, text, icons, etc. The sound training sub-module uses the deep learning method to train to obtain a deep learning model according to the medical equipment test sounds marked with sound categories and volume levels. The sound categories include alarm sound categories and non-alarm sound categories. The sound of high or low heart rate, the sound of stopping the pulsation, the poor connection of the connecting line, the alarm sound of the low power supply and other equipment; the non-alarm sound includes the sound of talking, the sound of opening and closing doors, and the sound of footsteps. Deep learning methods include artificial neural network (ANN), taking convolutional neural network (CNN) as an example, convolutional neural network includes multiple hidden layers, and the parameters of each hidden layer can be preset to Obtain the framework of the deep learning model, and then complete the preset deep learning model. The parameters include the weight value, bias value, structure of each hidden layer, the number of output layer nodes, the size of the convolution kernel, and the number of convolution kernels. It is also possible to obtain the framework of a mature deep learning model by initializing a mature deep learning model, and use the framework of the mature deep learning model as the framework of the preset deep learning model. These mature deep learning models include GoogLeNet, VGG, MobileNet, Inception, ResNet, EfficientNet, EfficientDet, Yolo, etc. The present invention does not limit these deep learning models.

The data storage and analysis module is used to store the trained deep learning model and test case specifications, and analyze the test audio and video data of the device according to the deep learning model. The data storage and analysis module uses the deep learning model trained by the deep learning module to perform feature extraction on the audio and video captured by the audio and video capture device, predicts the image category and location information in the image according to the extracted features, and predicts The category and volume of the sound. Finally, according to the information predicted by the deep learning model and the test case specifications, it is judged whether the equipment under test meets the requirements.

The data management module is used to generate, analyze and save electronic reports. The data management module can automatically generate and save electronic reports, and can issue reports in real time, shortening the time from detection to issuing reports, greatly improving the efficiency of production and testing, and relevant supervisors can see the reports in time and deal with corresponding problems in a timely manner. processing to improve the operating efficiency of the equipment. In addition, the electronic report adopts a paperless form, which can selectively print the required report content, which reduces the use of paper and is more energy-saving and environmentally friendly. In addition, the data management module also has a report analysis function, which conducts data statistics for the content of the unqualified report, so as to facilitate the timely discovery of batch problems in production and rectification.

The controller with the above structure can improve the automation degree in the equipment testing process, improve the production and testing efficiency, avoid manual testing and judgment, and have better consistency and repeatability.

The present invention also provides a method for detecting medical equipment, and FIG. 5 is a flowchart of the method for detecting medical equipment in the present invention. The medical equipment detection method in FIG. 5 includes the following steps: Step S1: the controller receives detection parameters, wherein the detection parameters include equipment model, equipment category, equipment detection history, detection purpose, detection items, ambient temperature and humidity parameters, and detection personnel information Wait. Step S2: The controller sends a test case to the simulator, wherein the test case includes a single test case for testing a certain item, a continuous test case for a certain set of standards or verification procedures, and a test case customized according to test requirements. Step S3: The simulator generates a physiological simulation signal according to the test case, and sends it to the device to be tested. Step S4: The device to be tested performs a test according to the received physiological analog signal. Step S5: The video and audio capture device captures the test process and test result of the device under test, and transmits the captured video and audio to the controller. Step S6: The controller compares the content of the video recordings according to the specifications of the test case, and determines whether the device to be tested complies with the regulations. The use of the medical equipment detection method in Figure 5 has the following advantages: (1) This method realizes automatic testing, and does not require manual participation in the whole process. Using machines to collect and judge results can not only improve the detection efficiency, but also have better test results. (2) In this method, the video and audio capture device is used to shoot the test process and test results, and the test can be carried out without knowing the communication protocol of each instrument, which broadens the application scenarios of this method. It has better applicability; in addition, the use of audio and video capture devices for audio and video shooting is the same as the way medical staff observe medical equipment, which is closer to the actual use situation, compared to the single way of using communication protocols for numerical comparison. , the detection system in the present invention can more comprehensively detect the equipment to be tested; (3) the method includes continuous test cases for a certain set of standards or verification procedures, which can reduce the time for researching regulations and verification procedures, and more Targeted instrument testing.

In order to improve the automation degree of the method, the method may further include step S7: generating an electronic test report according to the test result. In this step, the electronic inspection report can be generated immediately after the inspection of the device to be tested is completed, which shortens the time from inspection to issuing the report, greatly improves the efficiency of production and inspection, and the relevant supervisors can see the report in time and address the corresponding problems. Timely processing to improve the operating efficiency of the equipment; in addition, the electronic inspection report is a paperless office, which is more energy-saving and environmentally friendly.

Please refer to FIG. 6 , which is a flowchart of selecting a test case according to an embodiment of the present application. The method for selecting a test case includes: S11: The controller receives the detection parameter and acquires the detection item. S12: The controller judges whether there is a test case for the test item stored. When the controller determines that there is a test case for the inspection item, the next step S2 is performed: the controller sends a test case to the simulator; when the controller determines that there is no test case for the inspection item, step S13 is performed: customize the test according to the inspection item case, and save it to the controller. In the step of selecting test cases, you can customize the test cases through the test items in the test parameters, which can facilitate customers to customize test cases for different products and different processes according to their needs, adapt to different products and different test needs, and can meet different medical equipment. Product demand, improve the efficiency of production and testing.

Please refer to FIG. 7 , which is a flowchart of judging whether the device under test meets the requirements according to an embodiment of the present application. The judging process includes the following steps: Step S61 : Using the trained deep learning model, extract the image features and sound features of the video sound captured by the video and audio capture device. Step S62: Predict the category and location information of the image according to the extracted image features. Step S63: Predict the sound category and volume of the sound according to the extracted sound features. Step S64: Compare the test case specifications to determine whether the device to be tested meets the requirements, wherein the test case specifications include physiological signal waveform characteristics, numbers, characters, icons, sound types and sound levels. The judgment and recognition of images and sounds through the deep learning model can effectively reduce the workload of the inspectors, and can realize automatic machine judgment with better consistency and repeatability.

The above description is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited to this. The equivalent replacement or change of the inventive concept thereof shall be included within the protection scope of the present invention.

1: Controller 2: Simulator 3: Equipment to be tested 4: Video capture device 5: Elastic frame 51: Beam 52: Clamping structure of the device to be tested 520:First clamp arm 521: Second clamp arm 53: Swivel Arm 54: Video and audio capture device clamping structure 55: Rotary structure 56: Locking structure 57: Rotary positioning structure

FIG. 1 is a schematic structural diagram of a medical device detection system of the present invention. FIG. 2a is a schematic diagram of an unfolded state of the elastic frame; FIG. 2b is a schematic diagram of a storage state of the elastic frame. FIG. 3 is a schematic structural diagram of the controller in the present invention. FIG. 4 is a schematic structural diagram of a deep learning module in the present invention. FIG. 5 is a flow chart of the medical device detection method in the present invention. FIG. 6 is a flow chart of selecting a test case in the present invention. FIG. 7 is a flow chart of judging whether the device under test conforms to the regulations in the present invention.

1: Controller

2: Simulator

3: Equipment to be tested

4: Video capture device

Claims (14)

A medical equipment detection system, characterized in that it includes a controller, a simulator, a device to be tested, and an audio and video capture device, wherein the controller configures the operation of the simulator, and the controller sends a message to the simulator. A test case; the simulator converts the received test case into a corresponding physiological simulation signal, and sends the physiological simulation signal to the device to be tested; the device to be tested executes sequentially according to the received physiological simulation signal test work and generate sound prompts during the execution process; the video and audio capture device is used to photograph the test process and test results of the equipment under test, and transmit the photographed images and sound prompts to the controller; the controller According to the specifications of the test case, compare the content of the shooting audio to determine whether the device under test meets the requirements. The medical device testing system according to the first item of the scope of application, wherein the controller includes a testing item definition and selection module, a deep learning module, and a data storage and analysis module; the testing item definition and The selection module is used to edit and store test cases prepared for different devices and different regulations; the deep learning module is used to train image information and sound information, and obtain an image deep learning model and a sound deep learning model; The data storage and analysis module described above is used to store the trained deep learning model and test case specifications, and to analyze the device's shooting audio data according to the deep learning model. The medical device testing system according to item 2 of the scope of the application, wherein the controller further includes a data management module, and the data management module is used for generating, analyzing and saving electronic reports. The medical device testing system according to item 1 of the scope of application, characterized in that, the test cases include a single test case for testing a certain item, a continuous test case for standards or verification procedures, and a continuous test for basic signal generation. Cases, as well as custom test cases based on test requirements. The medical equipment testing system according to item 4 of the scope of application, characterized in that the standard or verification procedure includes at least one of the following standards or verification procedures: YY1079, YY1139, YY0782, YY0885, JJG760, JJG1042, JJG1163, JJG93, JJG1041, JJG1149, JJF1260, IEC60601-2-25, IEC60601-2-27, IEC60601-2-47, IEC60601-2-4, JJF1260, IEC60601-2-21, YY1259, YY0585, IEC60601-2-24, IEC60601-2-12, JJF1234, IEC60601-2-13, YY1147, EC11, EC13, EC38, EC57. The medical device detection system according to item 4 of the claimed scope, characterized in that the basic signal includes at least one of the following signals: sine wave, unipolar square wave, bipolar square wave, unipolar triangular wave, bipolar Triangle wave, square wave pulse wave, triangle pulse wave. The medical equipment testing system as described in item 1 of the scope of the patent application is characterized in that, when the controller compares the content of the video recordings according to the test case specifications, and determines whether the equipment to be tested meets the requirements, the following steps are included: The trained deep learning model extracts the technical features of the content of the shooting sound, infers the category and location information of the image in the shooting sound, as well as the sound category and sound level, and compares the extracted features with the test case specifications to determine the test to be tested. Whether the equipment complies with regulations. The medical equipment detection system according to the first item of the patent application scope, characterized in that, the simulator includes a heart and brain electrical simulation module, a blood pressure simulation module, a blood oxygen simulation module, an invasive blood pressure simulation module, and a carbon dioxide simulation module. Simulation module, carbon monoxide simulation module and temperature simulation module. The medical equipment detection system according to the first item of the patent application, characterized in that, the medical equipment detection system further comprises an elastic frame, and the elastic frame is used for holding the video and audio capture device, and adjusts the video and video capture device to match the elastic frame. The position between the devices under test. The medical equipment detection system as described in item 9 of the patent application scope, wherein the elastic frame comprises a beam, a clamping structure for the equipment to be tested, a rotating arm, a clamping structure for an audio and video capture device, a rotating structure and a locking structure , the bottom surface of the beam is connected to the clamping structure of the equipment to be tested, one end of the rotating arm is connected to the upper surface of the beam through the rotating structure, the other end of the rotating arm is connected to the clamping structure of the audio and video capture device, and the locking structure is arranged on the side of the beam. The clamping structure of the device to be tested includes a first clamping arm, a second clamping arm and two pulleys, and the first clamping arm and the second clamping arm are connected with the beam in a sliding manner, and are used to Adjust for different widths of the measuring device. A medical device detection method for the medical device detection system described in the first item of the scope of the application, characterized in that it includes the following steps: Step S1: the controller receives the detection parameters; Step S2: the controller sends the test to the simulator Case; Step S3: the simulator generates a physiological analog signal according to the test case, and sends it to the device under test; Step S4: the device under test performs the test according to the received physiological analog signal and generates a sound prompt during the execution process; Step S5: video and audio capture The device captures the test process and test results of the device to be tested, and transmits the captured image and audio prompts to the controller; Step S6: the controller compares the content of the video and audio to determine whether the device to be tested meets the requirements according to the specifications of the test case. The medical device testing method described in item 11 of the scope of the application, further comprising: step S7: generating an electronic testing report according to the testing result. The medical equipment detection method described in item 11 of the scope of the application is characterized in that, after the execution of S1, the following steps are also performed: S11: the controller receives the detection parameters, and acquires the detection items; S12: the controller determines whether there are stored For the test cases in the test items, when the controller determines that there are test cases for the test items, the next step S2 is performed: the controller sends the test cases to the simulator; when the controller determines that there are no test cases for the test items, the steps are executed S13: Customize the test case according to the test item and save it to the controller. The method for detecting medical equipment as described in item 11 of the scope of the application, wherein step S6 specifically includes the following steps: Step S61 : using the trained deep learning model to extract image features and sounds of the audio captured by the audio and video capture device Step S62: Predict the category and location information of the image according to the extracted image features; Step S63: Predict the sound category and volume of the sound according to the extracted sound features; Step S64: Compare the test case specifications, determine the Check whether the equipment complies with the regulations.

Images