Abnormity detection method and system for magnetic resonance imaging equipment

Abstract

The invention discloses a magnetic resonance imaging equipment anomaly detection method and system. The method comprises the following steps: acquiring sensor data in a period of time when the magnetic resonance equipment operates; preprocessing the sensor data; performing empirical mode decomposition on the sensor data to obtain a plurality of IMF components; performing Hilbert transform on each IMF component, and calculating an instantaneous amplitude and an instantaneous frequency of the IMF component; drawing a time-frequency map based on the instantaneous amplitude, the instantaneous frequency and the time information of the IMF component; constructing a data set based on the preprocessed time-frequency atlas, inputting the time-frequency atlas in the data set into a deep learning model to train the deep learning model, and obtaining a trained deep learning model; and acquiring to-be-detected sensor data, constructing a time-frequency map based on the to-be-detected sensor data, and inputting the time-frequency map into the trained deep learning model to obtain an equipment anomaly detection result. The method has the capability of identifying complex abnormal modes, and can improve the accuracy of abnormal detection of the magnetic resonance equipment.

Description

Technical Field

[0001] The present application relates to the technical field of medical device abnormality identification, and in particular to a magnetic resonance imaging device abnormality detection method, system, storage medium and electronic device. Background Art

[0002] Magnetic resonance imaging (MRI) is a highly sophisticated medical diagnostic device widely used in modern medicine. It is used to provide detailed imaging and diagnosis of the human body's internal structures. It uses a powerful magnetic field, radiofrequency pulses, and computer technology to cause hydrogen nuclei in human tissue to resonate and emit signals. These signals are then processed by a computer to form high-resolution images, helping doctors detect and diagnose various conditions, such as brain disorders, tumors, and cardiovascular diseases.

[0003] Magnetic resonance imaging (MRI) equipment relies on precise imaging to capture detailed information about the human body's interior. The accuracy of its diagnosis relies heavily on the equipment's high performance and stability. Failure or anomalies during operation can result in image degradation, artifacts, or even incorrect diagnostic information, impacting doctors' accurate assessments of conditions. Therefore, MRI anomaly detection must be able to promptly and accurately detect any system anomalies, avoiding diagnostic errors and maximizing patient accuracy.

[0004] Currently, anomaly detection in magnetic resonance imaging (MRI) equipment primarily relies on monitoring a single parameter and determining thresholds. For example, by monitoring changes in magnetic field strength, an alarm is triggered when the field exceeds a preset normal range, preventing magnetic field instability from affecting imaging quality. However, this anomaly detection method considers a relatively single factor, resulting in a high probability of false positives and missed negatives. Summary of the Invention

[0005] The embodiments of the present application provide a method, system, storage medium, and electronic device for detecting abnormalities in a magnetic resonance imaging device, which have the ability to recognize complex abnormal patterns and can improve the accuracy of abnormality detection in a magnetic resonance imaging device.

[0006] The present invention provides a method for detecting abnormalities in a magnetic resonance imaging device, including: Acquire sensor data during a period of time when the magnetic resonance device is in operation; wherein the sensor data includes a variety of sensor data collected by a variety of types of sensors; Preprocessing the sensor data; Performing empirical mode decomposition on the sensor data to obtain multiple IMF components; Performing a Hilbert transform on each of the IMF components, and calculating the instantaneous amplitude and instantaneous frequency of the IMF component; Drawing a time-frequency spectrum based on the instantaneous amplitude, the instantaneous frequency and time information of the IMF component, and preprocessing the time-frequency spectrum; A data set is constructed based on the preprocessed time-frequency spectrum, and the time-frequency spectrum in the data set is input into a deep learning model to train the deep learning model to obtain a trained deep learning model; The sensor data to be detected is obtained, a time-frequency spectrum is constructed based on the sensor data to be detected, and the time-frequency spectrum is input into a trained deep learning model to obtain a device anomaly detection result.

[0007] Furthermore, in the above-mentioned magnetic resonance imaging device abnormality detection method, the preprocessing of the sensor data includes: Identify missing values ​​and outliers in the sensor data, fill the missing values ​​using linear interpolation or spline interpolation, and correct or remove the outliers; Normalize the processed sensor data.

[0008] Furthermore, in the above-mentioned magnetic resonance imaging device abnormality detection method, the step of performing empirical mode decomposition on the sensor data to obtain multiple IMF components includes: Determine all local maximum points and local minimum points of the sensor data within a time period; Using a cubic spline interpolation method to interpolate the local maximum point and the local minimum point respectively to obtain an upper envelope and a lower envelope, and calculating the mean of the upper envelope and the lower envelope; Obtaining a preliminary intrinsic mode function based on the normalized sensor data and the mean value; Determining whether the preliminary eigenmode function is an intrinsic mode function; Obtain the residual component and repeat the decomposition to obtain multiple IMF components.

[0009] Furthermore, the above-mentioned magnetic resonance imaging device abnormality detection method, wherein the performing of Hilbert transform on each of the IMF components and calculating the instantaneous amplitude and instantaneous frequency of the IMF components, comprises: Performing a Hilbert transform on the IMF component to obtain a Hilbert transform result; Obtain an analytical signal based on the Hilbert transform result and the IMF component; The modulus of the analytical signal is analyzed to obtain the instantaneous amplitude of the IMF component, the phase of the analytical signal is analyzed, and the instantaneous frequency of the IMF component is calculated based on the phase.

[0010] Furthermore, in the above-mentioned magnetic resonance imaging device abnormality detection method, the step of drawing a time-frequency spectrum based on the instantaneous amplitude, the instantaneous frequency, and the time information of the IMF component comprises: The time information of the IMF component is used as the abscissa of the time-frequency spectrum, the instantaneous frequency is used as the ordinate, and the magnitude of the instantaneous amplitude is used as the grayscale value in the time-frequency spectrum.

[0011] Furthermore, the above-mentioned magnetic resonance imaging device abnormality detection method, wherein the time-frequency spectrum is preprocessed, includes: Adjusting the size of the time-frequency spectrum; All time-frequency maps within a running time period are superimposed to form a multi-channel image of the time-frequency map.

[0012] Furthermore, in the above-mentioned magnetic resonance imaging device abnormality detection method, the deep learning model is an improved ResNet18, and the improved ResNet18 includes: An input layer, comprising an initial convolutional layer, wherein the number of input channels of the initial convolutional layer is the same as the number of channels of the time-frequency spectrum; 4 groups of residual blocks are used for feature extraction to obtain feature maps; The global average pooling layer and the binary classification layer are used to compress the resolution of the feature map and predict the classification.

[0013] The present application also provides a magnetic resonance imaging device abnormality detection system, comprising: A data acquisition module, configured to acquire sensor data over a period of time while the magnetic resonance device is in operation; wherein the sensor data includes a variety of sensor data acquired by various types of sensors; A data preprocessing module, used for preprocessing the sensor data; An empirical mode decomposition module, configured to perform empirical mode decomposition on the sensor data to obtain a plurality of IMF components; a Hilbert transform module, configured to perform a Hilbert transform on each of the IMF components and calculate the instantaneous amplitude and instantaneous frequency of the IMF component; a time-frequency spectrum generating module, configured to draw a time-frequency spectrum based on the instantaneous amplitude, the instantaneous frequency and the time information of the IMF component, and to pre-process the time-frequency spectrum; The training module is used to construct a data set based on the preprocessed time-frequency spectrum, input the time-frequency spectrum in the data set into the deep learning model to train the deep learning model, and obtain a trained deep learning model; The device anomaly detection module is used to obtain sensor data to be detected, construct a time-frequency spectrum based on the sensor data to be detected, and input the time-frequency spectrum into a trained deep learning model to obtain device anomaly detection results.

[0014] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a plurality of instructions, wherein the instructions are suitable for being loaded by a processor to execute any of the above-mentioned magnetic resonance imaging device abnormality detection methods.

[0015] An embodiment of the present application also provides an electronic device, including a processor and a memory, wherein the processor is electrically connected to the memory, the memory is used to store instructions and data, and the processor is used in the steps of any of the above-mentioned methods for detecting abnormalities in magnetic resonance imaging equipment.

[0016] The present application provides a method, system, storage medium, and electronic device for detecting abnormalities in magnetic resonance imaging devices. The present application has the following beneficial effects: (1) Generally, abnormality detection in magnetic resonance imaging devices focuses only on a single sensor, while the present invention comprehensively considers the data of multiple sensors of magnetic resonance imaging devices, can capture cross-dimensional correlated abnormalities, identify complex abnormal patterns, improve the accuracy of abnormality detection, and reduce the false alarm rate and missed alarm rate. (2) The sensor data of the magnetic resonance imaging device is represented by a time-frequency spectrum, and the fusion of multiple sensors adopts a multi-channel approach, which can then be processed using an image deep learning model without the need to train multiple models. (3) During the data collection process, an appropriate amount of redundant data will be collected at both ends to solve the endpoint effect problem in the empirical mode decomposition process. (4) Unlike the previous method of limiting the features of sensor data to traditional statistical values ​​such as maximum, minimum, and average values, the present invention adopts transfer learning based on the mature ResNet model, which has stronger feature extraction capabilities and lower training costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The following detailed description of the specific embodiments of the present application in conjunction with the accompanying drawings will make the technical solutions and other beneficial effects of the present application apparent.

[0018] Figure 1 This is a flowchart of a method for detecting abnormalities in a magnetic resonance imaging device provided in an embodiment of the present application.

[0019] Figure 2 This is the normalized result of the magnetic field intensity sensor data provided in the embodiment of the present application.

[0020] Figure 3 This is the normalized result of the temperature sensor data provided in the embodiment of the present application.

[0021] Figure 4This is the normalized result of the RF power sensor data provided in the embodiment of this application.

[0022] Figure 5 This is the normalized result of the gradient magnetic field intensity sensor data provided in the embodiment of the present application.

[0023] Figure 6 This is the normalized result of the RF frequency sensor data provided in the embodiment of the present application.

[0024] Figure 7 This is the normalized result of the coil current sensor data provided in the embodiment of the present application.

[0025] Figure 8 The IMF is obtained by empirical mode decomposition of the RF power sensor data provided in the embodiment of the present application.

[0026] Figure 9 This is the time-frequency spectrum corresponding to the RF power sensor data provided in the embodiment of the present application.

[0027] Figure 10 This is the time-frequency diagram of the six sensors finally spliced ​​together in the embodiment of the present application.

[0028] Figure 11 This is a structural diagram of the magnetic resonance imaging equipment abnormality detection system provided in an embodiment of the present application.

[0029] Figure 12 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0031] Currently, anomaly detection in magnetic resonance imaging (MRI) equipment primarily relies on monitoring a single parameter and determining thresholds. For example, by monitoring changes in magnetic field strength, an alarm is triggered when the field exceeds a preset normal range, preventing magnetic field instability from affecting imaging quality. However, this anomaly detection method considers a relatively single factor, resulting in a high probability of false positives and missed negatives.

[0032] Therefore, when performing anomaly detection, considering multi-dimensional data cross-validation from multiple sensors can significantly reduce false alarm rates and improve detection accuracy. Furthermore, multi-sensor-based anomaly detection systems have the ability to identify complex anomaly patterns and capture cross-dimensional correlation anomalies between different parameters. Some complex equipment anomalies may manifest as subtle changes in multiple parameters. While observing each parameter individually may not be obvious, analyzing these parameters together can accurately identify potential failure modes. For example, simultaneously monitoring parameters such as magnetic field strength, RF power, and temperature may reveal specific failure modes, allowing for a more accurate assessment of the equipment's operating status.

[0033] To address the aforementioned issues, embodiments of the present application provide a method, system, storage medium, and electronic device for detecting anomalies in magnetic resonance imaging equipment. The embodiment of the present application provides a magnetic resonance imaging equipment anomaly detection system that can be integrated into an electronic device, such as a terminal or server. The terminal may include a tablet computer, laptop computer, personal computer (PC), microprocessor box, or other device.

[0034] See also Figure 1 , Figure 1 This is a flowchart of a method for detecting abnormalities in a magnetic resonance imaging device provided in an embodiment of the present application, which is applied to an electronic device. The method for detecting abnormalities in a magnetic resonance imaging device includes the following steps: S1, acquiring sensor data for a period of time while the magnetic resonance device is in operation; wherein the sensor data includes a variety of sensor data collected by various types of sensors; The data from the time period [0, t] during one MRI run is used as a sample. Raw sensor data from the temperature sensor, magnetic induction intensity sensor, and RF power sensor are collected during this time period. To facilitate subsequent cubic spline interpolation of endpoints, two additional data points are collected before and after the sample, resulting in sensor data at time [-2, t+2].

[0035] Sensor data should contain timestamp information, recording the time point corresponding to each sensor measurement value. The time interval of the data should be kept relatively uniform.

[0036] In one embodiment, the processed data may be stored in CSV format for subsequent analysis and processing.

[0037] S2, preprocessing of sensor data.

[0038] In one embodiment, step S1 includes: S21, identifying missing values ​​and outliers in the sensor data, filling missing values ​​with linear interpolation or spline interpolation, and correcting or removing outliers.

[0039] Perform a preliminary check on the collected data to identify and address any outliers or missing values. Missing values ​​are filled using linear or spline interpolation. Clearly unreasonable outliers, such as those outside the measurement range, are corrected or removed based on the statistical characteristics of the data (e.g., mean).

[0040] S22, normalizing the processed sensor data.

[0041] The data of each sensor is normalized and scaled to the range of [0, 1] to eliminate the difference in the magnitude of the data and improve the accuracy and stability of subsequent analysis. The formula is:

[0042] in, It is i The sensors in t The original data at the moment, and 、 are the historical minimum and maximum values ​​of the i-th sensor data, is the normalized data.

[0043] Figure 2-Figure 7 They are respectively the normalized results of the magnetic field strength sensor data, the normalized results of the temperature sensor data, the normalized results of the RF power sensor data, the normalized results of the gradient magnetic field strength sensor data, the normalized results of the RF frequency sensor data, and the normalized results of the coil current sensor data provided in the embodiments of the present application.

[0044] S3, perform empirical mode decomposition on the sensor data to obtain multiple IMF components.

[0045] In one embodiment, step S3 includes: S31, determining all local maximum points and local minimum points of sensor data within a time period.

[0046] S32, using a cubic spline interpolation method to interpolate the local maximum point and the local minimum point respectively to obtain an upper envelope and a lower envelope, and calculating the mean of the upper envelope and the lower envelope.

[0047] The upper envelope is calculated by the following formula and lower envelope The mean :

[0048] S33, obtaining a preliminary intrinsic mode function based on the normalized sensor data and the mean.

[0049] Using normalized sensor data minus , and obtain the preliminary eigenmode function: .

[0050] S34, determining whether the preliminary eigenmode function is an intrinsic mode function.

[0051] Determine whether it is an intrinsic mode function (IMF) and check whether the two conditions of IMF are met: (1) Throughout the entire time history, the number of extreme points and the number of zero crossing points must be equal or differ by no more than one; (2) At any time, the average value of the upper envelope formed by the local maximum points and the average value of the lower envelope formed by the local minimum points is zero.

[0052] if If the above conditions are not met, Consider it as a new signal and repeat steps S31 to S33 until a signal that meets the conditions is obtained. , which is recorded as the first IMF component .

[0053] S35, obtaining the residual component and repeatedly decomposing it to obtain multiple IMF components.

[0054] Calculate residual components .

[0055] Will As a new input signal, repeat the above steps to get the second IMF component , and so on, until the residual component It becomes a monotonic function and can no longer be decomposed into IMF components. The final sensor signal is It can be expressed as .

[0056] Figure 8 The IMF is obtained by empirical mode decomposition of the RF power sensor data provided in the embodiment of the present application.

[0057] S4, performing Hilbert transform on each IMF component and calculating the instantaneous amplitude and instantaneous frequency of the IMF component.

[0058] In one embodiment, step S4 includes: S31, performing Hilbert transform on the IMF component to obtain a Hilbert transform result.

[0059] For the IMF component , and perform the Hilbert transform:

[0060] in, is the Hilbert transform result.

[0061] S32, obtaining an analytical signal based on the Hilbert transform result and the IMF component.

[0062] It is expressed by the following formula:

[0063] in, To analyze the signal, Is an imaginary unit.

[0064] S33, analyzing the modulus of the analytical signal to obtain the instantaneous amplitude of the IMF component, analyzing the phase of the analytical signal, and calculating the instantaneous frequency of the IMF component based on the phase.

[0065] Specifically, parse the signal The modulus is the instantaneous amplitude

[0066] Analyzing the signal Phase

[0067] Instantaneous frequency .

[0068] S5, drawing a time-frequency spectrum based on the instantaneous amplitude, instantaneous frequency and time information of the IMF component, and preprocessing the time-frequency spectrum.

[0069] S51, using the time information of the IMF component as the horizontal coordinate of the time-frequency spectrum, the instantaneous frequency as the vertical coordinate, and the instantaneous amplitude as the grayscale value in the time-frequency spectrum.

[0070] Specifically, the instantaneous amplitude, instantaneous frequency and time information of all IMF components are combined to draw a time-frequency spectrum. The horizontal axis of the time-frequency spectrum is time. , the vertical axis is frequency , the grayscale in the spectrum represents the instantaneous amplitude , thereby intuitively showing the energy distribution of the signal at different times and frequencies. Figure 9 This is the time-frequency spectrum corresponding to the RF power sensor data provided in the embodiment of the present application.

[0071] S52, adjusting the size of the time-frequency spectrum.

[0072] The time-frequency maps of all sensor data are adjusted to obtain the same resolution, such as 112×224, to adapt to the model input.

[0073] S53: superimpose all the time-frequency spectra within a running time period to form a multi-channel image of the time-frequency spectra.

[0074] Using Python's OpenCV, the grayscale image list is converted into a NumPy array, and the time-frequency maps of all sensors are superimposed in a fixed order to obtain a multi-channel image of the sensor data during the runtime period. Figure 10 This is the time-frequency diagram of the six sensors finally spliced ​​together in the embodiment of the present application.

[0075] S6, constructing a data set based on the preprocessed time-frequency spectrum, inputting the time-frequency spectrum in the data set into the deep learning model to train the deep learning model, and obtaining a trained deep learning model.

[0076] In one embodiment, the deep learning model is a modified ResNet18, which includes an input layer, including one initial convolutional layer whose input channels are the same as the number of channels in the time-frequency spectrum; four groups of residual blocks for feature extraction to generate feature maps; a global average pooling layer and a binary classification layer for performing resolution compression and predicting classification on the feature maps.

[0077] The improved ResNet18 uses transfer learning to ensure strong image feature extraction capabilities while reducing training costs. Compared with the traditional ResNet18, the improved ResNet18 has the following structural differences: (1) Modify the input layer. Replace the first convolution layer of ResNet and change the number of input channels from 3 to N (N is the number of channels after stacking) to adapt to the multi-channel image input of the magnetic resonance imaging device sensor. The weights of the new convolution layer are randomly initialized, while the pre-trained weights of other layers are retained.

[0078] (2) Modify the output layer. Replace the fully connected layer with a binary classification output.

[0079] The following is the training process of the deep learning model: (1) Dataset division: Divide the training set, validation set, and test set into proportions (e.g., 70%-15%-15%) to ensure that each data set has a close ratio of positive and negative samples.

[0080] (2) Cross entropy loss is used as the loss function, and the original scores (logits) output by the model are converted into probability distribution through softmax.

[0081] (3) The optimizer uses SGD, the momentum coefficient is set to 0.9, and the learning rate is set to 0.001.

[0082] (4) Layer-wise learning rate: The pre-training layer uses a lower learning rate (such as 0.00001), and the new layer uses a higher learning rate (such as 0.0001).

[0083] (5) Learning rate scheduling: use cosine annealing or reduce it as needed.

[0084] (6) Weight decay: L2 regularization (e.g., 0.0001).

[0085] (7) Dropout: Add Dropout (e.g., probability 0.5) before the fully connected layer.

[0086] (8) Batch training: Use a reasonable batch size (such as 32) and monitor GPU memory usage.

[0087] (8) Early stopping method: Stop training when the validation set loss does not decrease for multiple consecutive epochs.

[0088] (10) Training indicators: Loss, Accuracy, F1 Score.

[0089] (11) Visualization: Use TensorBoard to record the training curve.

[0090] S7, obtain the sensor data to be detected, construct a time-frequency spectrum based on the sensor data to be detected, input the time-frequency spectrum into the trained deep learning model, and obtain the device anomaly detection result.

[0091] The trained deep learning model is deployed on the server to perform real-time detection of sensor data from the magnetic resonance imaging equipment. The deep learning model classifies the data as abnormal or not, and decides whether to issue an alarm based on the classification results.

[0092] Based on the method described in the above embodiment, this embodiment will be further described from the perspective of a magnetic resonance imaging device anomaly detection system. The magnetic resonance imaging device anomaly detection system can be implemented as an independent entity or integrated into an electronic device. The electronic device can be a terminal, a server, or other device. The terminal can include a tablet computer, a laptop computer, a personal computer (PC), a micro processing box, or other devices.

[0093] See also Figure 11 , Figure 11 The present invention specifically describes a magnetic resonance imaging device anomaly detection system provided by an embodiment of the present invention, which is applied to an electronic device. The magnetic resonance imaging device anomaly detection system may include: A data acquisition module, configured to acquire sensor data over a period of time while the magnetic resonance device is in operation; wherein the sensor data includes a variety of sensor data acquired by various types of sensors; A data preprocessing module, used for preprocessing the sensor data; An empirical mode decomposition module, configured to perform empirical mode decomposition on the sensor data to obtain a plurality of IMF components; a Hilbert transform module, configured to perform a Hilbert transform on each of the IMF components and calculate the instantaneous amplitude and instantaneous frequency of the IMF component; a time-frequency spectrum generating module, configured to draw a time-frequency spectrum based on the instantaneous amplitude, the instantaneous frequency and the time information of the IMF component, and to pre-process the time-frequency spectrum; The training module is used to construct a data set based on the preprocessed time-frequency spectrum, input the time-frequency spectrum in the data set into the deep learning model to train the deep learning model, and obtain a trained deep learning model; The device anomaly detection module is used to obtain sensor data to be detected, construct a time-frequency spectrum based on the sensor data to be detected, and input the time-frequency spectrum into a trained deep learning model to obtain device anomaly detection results.

[0094] During specific implementation, the above modules and / or units can be implemented as independent entities, or can be arbitrarily combined to be implemented as the same or several entities. The specific implementation of the above modules and / or units can refer to the previous method embodiments. The specific beneficial effects that can be achieved can also be found in the beneficial effects in the previous method embodiments, which will not be repeated here.

[0095] In addition, an embodiment of the present application further provides an electronic device, which may be a computer, tablet computer, or other device. This electronic device can implement the steps of any embodiment of the magnetic resonance imaging device anomaly detection method provided in the embodiments of the present application, and thus can achieve the beneficial effects achieved by any magnetic resonance imaging device anomaly detection method provided in the embodiments of the present application. For details, please refer to the previous embodiments and will not be repeated here.

[0096] Figure 12 The following is a block diagram of the structure of an electronic device provided in an embodiment of the present invention, which can be used to implement the magnetic resonance imaging device anomaly detection method provided in the above embodiments. The electronic device 500 can be a terminal, server, or other device. The terminal can include a tablet computer, laptop computer, personal computer (PC), microprocessor, or other device.

[0097] RF circuit 510 is used to receive and transmit electromagnetic waves, converting them into electrical signals, thereby enabling communication with a communications network or other devices. RF circuit 510 may include various existing circuit components for performing these functions, such as an antenna, a radio frequency transceiver, a digital signal processor, an encryption / decryption chip, a subscriber identity module (SIM) card, memory, and the like. RF circuit 510 can communicate with various networks, such as the Internet, an intranet, or a wireless network, or with other devices via a wireless network. These wireless networks may include cellular telephone networks, wireless local area networks, or metropolitan area networks. The wireless networks may utilize various communication standards, protocols, and technologies, including but not limited to Global System for Mobile Communication (GSM), Enhanced Data GSM Environment (EDGE), Wideband Code Division Multiple Access (WCDMA), Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Wireless Fidelity (Wi-Fi) (such as Institute of Electrical and Electronics Engineers standards IEEE 802.11a, IEEE 802.11b, IEEE802.11g, and / or IEEE802.11n), Voice over Internet Protocol (VoIP), Worldwide Interoperability for Microwave Access (Wi-Max), other protocols for email, instant messaging, and short messaging, and any other suitable communication protocols, including those currently undeveloped.

[0098] The memory 520 can be used to store software programs and modules, such as the corresponding program instructions / modules in the above-mentioned embodiments. The processor 580 executes various functional applications and data processing by running the software programs and modules stored in the memory 520, that is, realizing functions such as taking pictures with the front camera, processing the captured images, and switching the display color of the displayed content on the display screen. The memory 520 may include a high-speed random access memory and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 520 may further include a memory remotely located relative to the processor 580, and these remote memories may be connected to the electronic device 500 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0099] The input unit 530 may be used to receive input digital or character information, and generate a keyboard and a mouse related to user settings and function control. The display unit 540 can be used to display information input by the user or information provided to the user, as well as various graphical user interfaces. These graphical user interfaces can be composed of graphics, text, icons, videos, or any combination thereof. The display unit 540 may include a display panel 541. Optionally, the display panel 541 can be configured in the form of an LCD (Liquid Crystal Display), an OLED (Organic Light-Emitting Diode), or the like.

[0100] Audio circuit 560, speaker 561, and microphone 562 provide an audio interface between the user and electronic device 500. Audio circuit 560 converts received audio data into electrical signals and transmits them to speaker 561, which then converts them into sound signals for output. Microphone 562, on the other hand, converts collected sound signals into electrical signals, which are then received by audio circuit 560 and converted into audio data. The audio data is then processed by output processor 580 and transmitted via RF circuit 510 to, for example, another terminal. Alternatively, the audio data may be output to memory 520 for further processing. Audio circuit 560 may also include an earphone jack to allow communication between external headphones and electronic device 500.

[0101] Electronic device 500, through a transmission module 570 (e.g., a Wi-Fi module), can help users receive requests, send information, and so on, providing users with wireless broadband Internet access. Although the figure shows transmission module 570, it is understood that it is not a required component of electronic device 500 and can be omitted as needed without changing the essence of the invention.

[0102] Processor 580 is the control center of electronic device 500. It connects all components of the phone using various interfaces and circuits. By running or executing software programs and / or modules stored in memory 520 and accessing data stored in memory 520, it executes various functions of electronic device 500 and processes data, thereby providing overall monitoring of the electronic device. Optionally, processor 580 may include one or more processing cores. In some embodiments, processor 580 may integrate an application processor and a modem processor. The application processor primarily handles the operating system, user interface, and application programs, while the modem processor primarily handles wireless communications. It is understood that the modem processor may not be integrated into processor 580.

[0103] Electronic device 500 also includes a power supply 590 (e.g., a battery) for powering various components. In some embodiments, the power supply can be logically connected to processor 580 via a power management system, thereby enabling the power management system to manage charging, discharging, and power consumption. Power supply 590 can also include any components, such as one or more DC or AC power supplies, a recharging system, a power failure detection circuit, a power converter or inverter, and a power status indicator.

[0104] Although not shown, the electronic device 500 also includes a camera (such as a front camera and a rear camera), a Bluetooth module, etc., which will not be described in detail here. Specifically, in this embodiment, the display unit of the electronic device is a touch screen display, and the mobile terminal also includes a memory and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by one or more processors. The one or more programs include instructions for performing the following operations: Acquire sensor data during a period of time when the magnetic resonance device is in operation; wherein the sensor data includes a variety of sensor data collected by a variety of types of sensors; Preprocessing the sensor data; Performing empirical mode decomposition on the sensor data to obtain multiple IMF components; Performing a Hilbert transform on each of the IMF components, and calculating the instantaneous amplitude and instantaneous frequency of the IMF component; Drawing a time-frequency spectrum based on the instantaneous amplitude, the instantaneous frequency and time information of the IMF component, and preprocessing the time-frequency spectrum; A data set is constructed based on the preprocessed time-frequency spectrum, and the time-frequency spectrum in the data set is input into a deep learning model to train the deep learning model to obtain a trained deep learning model; The sensor data to be detected is obtained, a time-frequency spectrum is constructed based on the sensor data to be detected, and the time-frequency spectrum is input into a trained deep learning model to obtain a device anomaly detection result.

[0105] In specific implementation, the above modules can be implemented as independent entities, or can be arbitrarily combined and implemented as the same or several entities. The specific implementation of the above modules can be found in the previous method embodiments and will not be repeated here.

[0106] Those skilled in the art will appreciate that all or part of the steps in the various methods of the above embodiments can be accomplished through instructions, or by controlling related hardware through instructions. The instructions can be stored in a computer-readable storage medium and loaded and executed by a processor. To this end, an embodiment of the present invention provides a storage medium storing a plurality of instructions that can be loaded by a processor to execute the steps of any of the embodiments of the magnetic resonance imaging device abnormality detection method provided in the embodiments of the present invention.

[0107] The computer-readable storage medium may include a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0108] Since the instructions stored in the storage medium can execute the steps of any embodiment of the magnetic resonance imaging device abnormality detection method provided in the embodiments of the present invention, the beneficial effects that can be achieved by any magnetic resonance imaging device abnormality detection method provided in the embodiments of the present invention can be achieved. Please refer to the previous embodiments for details and will not be repeated here.

[0109] The above is a detailed introduction to a magnetic resonance imaging device abnormality detection method, system, storage medium and electronic device provided in the embodiments of the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core ideas of the present application. At the same time, for those skilled in the art, based on the ideas of the present application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present application.

Claims

1. A method for detecting abnormality in a magnetic resonance imaging device, characterized in that: The method comprises: Acquire sensor data during a period of time when the magnetic resonance device is in operation; wherein the sensor data includes a variety of sensor data collected by a variety of types of sensors; Preprocessing the sensor data; Performing empirical mode decomposition on the sensor data to obtain multiple IMF components; Performing a Hilbert transform on each of the IMF components, and calculating the instantaneous amplitude and instantaneous frequency of the IMF component; Drawing a time-frequency spectrum based on the instantaneous amplitude, the instantaneous frequency and time information of the IMF component, and preprocessing the time-frequency spectrum; A data set is constructed based on the preprocessed time-frequency spectrum, and the time-frequency spectrum in the data set is input into a deep learning model to train the deep learning model to obtain a trained deep learning model; The sensor data to be detected is obtained, a time-frequency spectrum is constructed based on the sensor data to be detected, and the time-frequency spectrum is input into a trained deep learning model to obtain a device anomaly detection result.

2. The method for detecting abnormality of a magnetic resonance imaging device according to claim 1, wherein: The preprocessing of the sensor data includes: Identify missing values ​​and outliers in the sensor data, fill the missing values ​​using linear interpolation or spline interpolation, and correct or remove the outliers; Normalize the processed sensor data.

3. The method for detecting abnormality of a magnetic resonance imaging device according to claim 1, wherein: The sensor data is subjected to empirical mode decomposition to obtain multiple IMF components, including: Determine all local maximum points and local minimum points of the sensor data within a time period; Using a cubic spline interpolation method to interpolate the local maximum point and the local minimum point respectively to obtain an upper envelope and a lower envelope, and calculating the mean of the upper envelope and the lower envelope; Obtaining a preliminary intrinsic mode function based on the normalized sensor data and the mean value; Determining whether the preliminary eigenmode function is an intrinsic mode function; Obtain the residual component and repeat the decomposition to obtain multiple IMF components.

4. The method for detecting abnormality in a magnetic resonance imaging device according to claim 3, wherein: The step of performing Hilbert transform on each of the IMF components and calculating the instantaneous amplitude and instantaneous frequency of the IMF components comprises: Performing a Hilbert transform on the IMF component to obtain a Hilbert transform result; Obtain an analytical signal based on the Hilbert transform result and the IMF component; The modulus of the analytical signal is analyzed to obtain the instantaneous amplitude of the IMF component, the phase of the analytical signal is analyzed, and the instantaneous frequency of the IMF component is calculated based on the phase.

5. The method for detecting abnormality in a magnetic resonance imaging device according to claim 1, wherein: The drawing of a time-frequency spectrum based on the instantaneous amplitude, the instantaneous frequency, and the time information of the IMF component includes: The time information of the IMF component is used as the abscissa of the time-frequency spectrum, the instantaneous frequency is used as the ordinate, and the magnitude of the instantaneous amplitude is used as the grayscale value in the time-frequency spectrum.

6. The method for detecting abnormality in a magnetic resonance imaging device according to claim 5, wherein: The time-frequency spectrum is preprocessed, including: Adjusting the size of the time-frequency spectrum; All time-frequency maps within a running time period are superimposed to form a multi-channel image of the time-frequency map.

7. The method for detecting abnormality in a magnetic resonance imaging device according to claim 1, wherein: The deep learning model is an improved ResNet18, which includes: An input layer, comprising an initial convolutional layer, wherein the number of input channels of the initial convolutional layer is the same as the number of channels of the time-frequency spectrum; 4 groups of residual blocks are used for feature extraction to obtain feature maps; The global average pooling layer and the binary classification layer are used to compress the resolution of the feature map and predict the classification.

8. A magnetic resonance imaging equipment abnormality detection system, characterized in that: include: A data acquisition module is used to obtain sensor data over a period of time while the magnetic resonance device is running; The sensor data includes a variety of sensor data collected by various types of sensors; A data preprocessing module, used for preprocessing the sensor data; An empirical mode decomposition module, configured to perform empirical mode decomposition on the sensor data to obtain a plurality of IMF components; a Hilbert transform module, configured to perform a Hilbert transform on each of the IMF components and calculate the instantaneous amplitude and instantaneous frequency of the IMF component; a time-frequency spectrum generating module, configured to draw a time-frequency spectrum based on the instantaneous amplitude, the instantaneous frequency and the time information of the IMF component, and to pre-process the time-frequency spectrum; The training module is used to construct a data set based on the preprocessed time-frequency spectrum, input the time-frequency spectrum in the data set into the deep learning model to train the deep learning model, and obtain a trained deep learning model; The device anomaly detection module is used to obtain sensor data to be detected, construct a time-frequency spectrum based on the sensor data to be detected, and input the time-frequency spectrum into a trained deep learning model to obtain device anomaly detection results.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the magnetic resonance imaging device abnormality detection method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the processor is electrically connected to the memory, the memory is used to store instructions and data, and the processor is used to execute the steps of the magnetic resonance imaging device abnormality detection method according to any one of claims 1 to 7.

Images