Intelligent analysis and early warning method for medical imaging equipment log data

By extracting and analyzing medical imaging equipment log data and medical recording images, and combining the equipment model for quality scoring, the problem of not being able to automatically identify nonlinear abnormalities in medical imaging in the prior art is solved, and the equipment's self-test capability and early warning accuracy are improved.

CN118658602BActive Publication Date: 2024-11-29JICUISIHE TRUSTED DATA (NANJING) CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411151550.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-21
Publication Date
2024-11-29
Estimated Expiration
2044-08-21

AI Technical Summary

Technical Problem

Existing medical imaging equipment log data analysis and early warning technology is difficult to automatically identify nonlinear abnormalities in medical imaging, resulting in insufficient equipment self-test capability and early warning accuracy.

Method used

By extracting the operating status record data and medical record images in the log data of the medical imaging equipment, and combining the device model to assign samples and feature detection to the medical imaging quality indicators, finally the image is quality scored to generate early warning information.

Benefits of technology

It improves the self-test capability and analysis and early warning accuracy of medical imaging equipment, and can automatically identify and early warning of nonlinear abnormalities in medical imaging to ensure equipment stability and diagnostic accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118658602B_ABST
    Figure CN118658602B_ABST
Patent Text Reader

Abstract

The present invention discloses an intelligent analysis and early warning method for medical imaging equipment log data, which relates to the field of data processing technology. The method includes: extracting medical imaging equipment operation status record data; extracting medical recording images according to the medical imaging equipment log data; obtaining the first medical imaging quality index assignment result; performing feature detection on the medical imaging quality index to obtain the second medical imaging quality index assignment result; performing quality scoring on the medical recording images to obtain the quality score value; when the quality score value is less than or equal to the quality score threshold, generating the first medical imaging equipment early warning information. The method solves the technical problem that the existing medical imaging equipment log data analysis and early warning is limited to monitoring the linear anomalies of circuits and mechanical components, which leads to the inability to automatically identify the nonlinear anomalies of medical images, and achieves the technical effect of improving the self-checking ability of medical imaging equipment and the accuracy of analysis and early warning.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field related to data processing, and specifically to an intelligent analysis and early warning method for log data of medical imaging equipment. Background Art

[0002] In the medical field, medical imaging equipment is an important tool for diagnosis and treatment. Its stability and accuracy are directly related to the health and safety of patients. The traditional medical imaging equipment log self-check method has significant limitations. It mainly focuses on linear identifiable abnormal data such as circuit failure and mechanical failure. Although it can guarantee the basic operation of the equipment to a certain extent, its shortcomings are becoming more and more prominent in the face of increasingly complex medical image processing needs. Traditional self-check mechanisms often rely on preset thresholds and rules to monitor hardware-level faults of the equipment, such as circuit short circuits and wear of mechanical parts. These usually have clear physical manifestations and are easy to identify through simple sensors and algorithms. However, in the field of medical image processing, nonlinear abnormalities such as image clarity, contrast, and artifacts are difficult to effectively monitor through traditional self-check methods. For example, blurry or distorted images may lead to the omission or misdiagnosis of lesions, which in turn affects the formulation of treatment plans.

[0003] Therefore, the current medical imaging equipment log data analysis and early warning related technologies are limited to monitoring linear anomalies of circuits and mechanical components, which leads to the technical problem of being unable to automatically identify nonlinear anomalies in medical images. Summary of the invention

[0004] This application provides an intelligent analysis and early warning method for medical imaging equipment log data, and adopts technical means such as image extraction, sample assignment, feature detection, and quality scoring to solve the technical problem that the existing medical imaging equipment log data analysis and early warning are limited to monitoring linear anomalies of circuits and mechanical components, which leads to the inability to automatically identify nonlinear anomalies in medical images, thereby achieving the technical effect of improving the self-inspection capability of medical imaging equipment and the accuracy of analysis and early warning.

[0005] The present application provides a method for intelligent analysis and early warning of medical imaging device log data, the method comprising: extracting medical imaging device operation status record data according to the medical imaging device log data; extracting medical recording images according to the medical imaging device log data, wherein the medical recording images and the medical imaging device operation status record data have time domain consistency; performing sample assignment of medical image quality indicators according to the medical imaging device operation status record data and in combination with the medical imaging device model to obtain a first medical image quality indicator assignment result; performing feature detection on the medical image quality indicator based on the medical recording images to obtain a second medical image quality indicator assignment result; performing quality scoring on the medical recording images according to the first medical image quality indicator assignment result and the second medical image quality indicator assignment result to obtain a quality score value; and generating first medical imaging device early warning information when the quality score value is less than or equal to a quality score threshold.

[0006] In a possible implementation, based on the medical imaging device operating status record data and in combination with the medical imaging device model, sample assignment of the medical imaging quality index is performed to obtain a first medical imaging quality index assignment result, and the following processing is performed: with the medical imaging device operating status record data of the medical imaging device model as a restriction, qualified sample feature detection record data is collected online, wherein the qualified sample feature detection record data includes medical image quality index record values; the medical image quality index record values ​​are forward processed to obtain medical image quality index characteristic values; the medical image quality index characteristic values ​​are traversed to perform centralized data extraction to obtain centralized medical image quality index characteristic values; the centralized medical image quality index characteristic values ​​are traversed to perform minimum value sorting to obtain the first medical image quality index assignment result.

[0007] In a possible implementation, the qualified sample feature detection record data includes medical image quality index record values, and the following processing is performed: the medical image quality index includes at least image contrast index, image brightness index, spatial resolution index, temporal resolution index, and image clarity index; the medical imaging equipment operating status index includes at least X-ray tube voltage, X-ray tube current, exposure time, detector temperature, equipment operating temperature, magnetic field strength, and radio frequency pulse.

[0008] In a possible implementation, the medical record image is scored for quality according to the first medical image quality index assignment result and the second medical image quality index assignment result to obtain a quality score value, and the following processing is also performed: according to the first medical image quality index assignment result, the first assignment result of the jth index is extracted, and according to the second medical image quality index assignment result, the second assignment result of the jth index is extracted; twice the product of the first assignment result of the jth index and the second assignment result of the jth index is calculated, and a preset constant is added to set it as a first evaluation factor; the square sum of the first assignment result of the jth index and the second assignment result of the jth index is calculated, and the preset constant is added to set it as a second evaluation factor; the ratio of the first evaluation factor to the second evaluation factor is calculated, and the ratio is set as the jth indicator quality coefficient, and added to the indicator quality coefficient set; the medical image quality index is weighted to generate a medical image quality index weighted result; the indicator quality coefficient set is weighted averaged according to the medical image quality index weighted result to generate the quality score value.

[0009] In a possible implementation, the medical image quality indicator is weighted to generate a medical image quality indicator weighted result, and the following processing is performed: the medical imaging device model and the medical image quality indicator are sent in parallel to M evaluation parties to obtain M groups of medical image quality indicator importance scoring results, wherein the M evaluation parties are informationally isolated from each other; the medical image quality indicators are traversed to group and reconstruct the M groups of medical image quality indicator importance scoring results to obtain a first indicator importance scoring set up to an Nth indicator importance scoring set; the medical image quality indicators are weighted according to the first indicator importance scoring set up to the Nth indicator importance scoring set to generate a medical image quality indicator weighted result.

[0010] In a possible implementation, the medical image quality indicators are weighted according to the first indicator importance score set up to the Nth indicator importance score set to generate a medical image quality indicator weighted result, and the following processing is performed: the first indicator importance score set is added up to the Nth indicator importance score set to obtain the indicator importance score sum; the first indicator importance score sum of the first indicator importance score set is compared with the indicator importance score sum, and the result is set as the first indicator weighted result; until the Nth indicator importance score sum of the Nth indicator importance score set is compared with the indicator importance score sum, and the result is set as the Nth indicator weighted result; the first indicator weighted results up to the Nth indicator weighted result are added to the medical image quality indicator weighted result.

[0011] In a possible implementation, feature detection is performed on the medical image quality index based on the medical record image, and before obtaining the second medical image quality index assignment result, the following processing is also performed: the medical image quality index is classified to obtain an image time zone quality index and an image instantaneous quality index, wherein the image time zone quality index represents a feature type that requires analysis of differences in multiple moments, and the image instantaneous quality index represents a feature type that can be extracted by analyzing instantaneous images; a time series model is configured according to the image time zone quality index to generate a first feature detection channel; a convolutional neural network is configured according to the image instantaneous quality index to generate a second feature detection channel; and the medical record image is processed according to the first feature detection channel and the second feature detection channel to obtain the second medical image quality index assignment result.

[0012] In a possible implementation, the medical imaging device log data intelligent analysis and early warning method further performs the following processing: when the quality score value is greater than a quality score threshold, calculating multiple quality score values ​​at multiple moments based on the medical imaging device log data; calculating the variance of the quality score values ​​at multiple moments to obtain a device score fluctuation parameter; when the device score fluctuation parameter is greater than or equal to the fluctuation parameter threshold, generating a second medical imaging device early warning information.

[0013] Through the intelligent analysis and early warning method of medical imaging equipment log data proposed in this application, it is intended to extract the medical imaging equipment operation status record data; extract the medical image record according to the medical imaging equipment log data; perform sample assignment of the medical image quality index in combination with the medical imaging equipment model to obtain the first medical image quality index assignment result; perform feature detection on the medical image quality index to obtain the second medical image quality index assignment result; perform quality scoring on the medical image record to obtain the quality score value; when the quality score value is less than or equal to the quality score threshold, generate the first medical imaging equipment early warning information. It solves the technical problem that the existing medical imaging equipment log data analysis and early warning is limited to monitoring the linear anomalies of circuits and mechanical components, which leads to the inability to automatically identify the nonlinear anomalies of medical images, and achieves the technical effect of improving the self-detection capability of medical imaging equipment and the accuracy of analysis and early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solution of the embodiment of the present disclosure, the accompanying drawings of the embodiment of the present disclosure will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0015] Figure 1 A schematic diagram of the process of the intelligent analysis and early warning method of medical imaging equipment log data provided in an embodiment of the present application.

[0016] Figure 2 A schematic diagram of a process for obtaining a first medical image quality indicator assignment result in the medical imaging equipment log data intelligent analysis and early warning method provided in an embodiment of the present application. DETAILED DESCRIPTION

[0017] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0018] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0019] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0020] The present application embodiment provides a medical imaging device log data intelligent analysis and early warning method, such as Figure 1 As shown, the method includes:

[0021] Step S100, extracting the medical imaging device operation status record data according to the medical imaging device log data. Extracting the medical imaging device operation status record data according to the medical imaging device log data specifically refers to extracting relevant information reflecting the operation status of the device from the log file of the medical imaging device, which may include but is not limited to basic device information (such as device model, serial number, manufacturer, etc.), operation status parameters (various parameters generated by the device during operation, such as temperature, voltage, current, power, exposure time, imaging parameters, etc., reflecting the real-time operation status and performance of the device), fault records (fault information generated by the device during operation, including the time, type, cause, etc. of the fault), usage records (equipment usage, such as power-on time, power-off time, usage time, number of operations, etc., which helps to understand the frequency of use and load of the device), etc. The purpose of extracting these operation status record data is to comprehensively monitor and analyze the medical imaging device so as to timely discover potential problems and faults and ensure the normal operation of the device and the safety of the patient. In actual operation, it is necessary to use professional log analysis tools to parse and process the imaging device log files to extract useful equipment operation status information.

[0022] Step S200: extracting medical record images according to the medical imaging device log data, wherein the medical record images and the medical imaging device operation status record data have temporal consistency. Medical record images are extracted from the log data of medical imaging equipment, and it is ensured that these images have temporal consistency with the medical imaging equipment operation status record data, where medical record images refer to image data generated by medical imaging equipment for medical diagnosis, treatment or research purposes, such as X-rays, CT scans, MRI images, ultrasound images, etc. Temporal consistency refers to the consistency or synchronization of two or more data sets (here, medical record images and medical imaging equipment operation status record data) in time, that is, the generation time of the medical record images is required to coincide with the recording time of the medical imaging equipment operation status record data (correspond to each other). Specifically, both the medical record images and the medical imaging equipment operation status record data should contain clear timestamp information, and these timestamps should be able to correspond to each other, indicating that the two are generated in the same or similar time period. By matching the timestamps, it can be ensured that the medical record images can be accurately associated with the equipment operation status record data during analysis. When the medical imaging equipment generates the medical record images, it will also record the equipment operation status data. The medical record images and the medical imaging equipment operation status record data are completely synchronized in time, that is, they have temporal consistency.

[0023] Step S300, based on the medical imaging device operation status record data and in combination with the medical imaging device model, sample assignment of medical imaging quality indicators is performed to obtain a first medical imaging quality indicator assignment result. Based on the medical imaging device operation status record data, a preliminary quantitative evaluation of the medical imaging quality indicators is performed by comprehensively considering the impact of the device operation status and device model on the medical image quality. Specifically, the model information of the medical imaging device is identified and recorded. Different models of equipment may differ in imaging quality, technical parameters, etc., and then the basic performance and characteristics of the device model in terms of medical image quality are understood based on the device model. In combination with the device operation status record data and device model information, a preliminary quantitative evaluation of the medical imaging quality indicators (such as image resolution, time resolution, image clarity, image contrast, image brightness, noise level, artifact degree, etc.) is performed, and the device operation status parameters are converted into an evaluation result of the medical image quality to obtain the first medical image quality indicator assignment result, which is used to reflect the The preliminary evaluation results of medical image quality under the current equipment operating status and model, among which the medical image quality index is an important basis for evaluating the quality of medical images. When selecting quality indicators, it is necessary to determine them according to the specific type of medical images (such as X-rays, CT scans, MRI images, etc.) and clinical application requirements. The sample assignment process may include organizing and analyzing the acquired medical imaging equipment operating status record data, and extracting key parameters and indicators related to medical image quality; establishing a corresponding mathematical model based on the medical imaging equipment model and known quality indicator standards, which can comprehensively consider the relationship between the equipment operating status and quality indicators; inputting the organized data into the model for calculation and assignment; and finally outputting the first medical image quality indicator assignment result, which reflects the imaging quality of the medical imaging equipment under specific conditions.

[0024] In one possible implementation, Figure 2 As shown, step S300 further includes step S310, which collects qualified sample feature detection record data through networking, with the medical imaging device operation status record data of the medical imaging device model as a restriction, wherein the qualified sample feature detection record data includes medical image quality index record values. The medical imaging device model refers to the specific device model used for medical imaging examination, such as CT machine, MRI machine, etc. Based on the medical imaging device operation status record data of the same model of medical imaging device, the corresponding record data is obtained from multiple data sources (such as hospital information system, equipment management system, etc.) using network technology, that is, qualified sample feature detection record data, which refers to the various index values ​​of the medical record images that are evaluated to be of qualified quality in the feature detection process in the medical imaging examination, that is, the medical image quality index record value.

[0025] The method further includes step S320, in which the recorded values ​​of the medical image quality index are positively processed to obtain characteristic values ​​of the medical image quality index. Since the recorded values ​​of the medical image quality index may contain positive indicators (such as image clarity, contrast, etc.) and negative indicators (such as noise level, artifact degree, etc.), it is necessary to uniformly convert these indicators into positive indicators. The positive processing refers to the operations such as negation and reciprocal transformation of the negative indicators, so that the larger the value, the better the quality. After the positive processing, the recorded values ​​of the medical image quality index are converted into characteristic values ​​of the medical image quality index, and these characteristic values ​​more intuitively reflect the quality status of the medical image.

[0026] The method further includes step S330, traversing the medical image quality index characteristic values ​​to perform centralized data extraction to obtain centralized medical image quality index characteristic values. Traversing all medical image quality index characteristic values, i.e. checking each characteristic value one by one, centralized data extraction refers to extracting representative or key data from all characteristic values, deleting discrete data, and the extracted key data can reflect the overall status of medical image quality. After centralized data extraction, a set of characteristic values ​​that can represent medical image quality is obtained, i.e. centralized medical image quality index characteristic values.

[0027] The method further includes step S340, traversing the concentrated medical image quality index characteristic values ​​to sort the minimum values ​​and obtain the first medical image quality index assignment result. By comparing the sizes of the concentrated medical image quality index characteristic values ​​one by one, the minimum value is found. Specifically, first, a variable (such as m) is set to store the currently found minimum value, and then, each element of the concentrated medical image quality index characteristic value is traversed. During the traversal process, the current element (i.e., the medical image quality index characteristic value currently under investigation) is compared with m. If the value of the current element is less than m, m is updated to the value of the current element. As the traversal proceeds, m will always remain the minimum value of the traversed elements. After all elements are traversed, what is stored in m is the minimum value of the concentrated medical image quality index characteristic values, i.e., the first medical image quality index assignment result. The medical imaging equipment operating status indicator includes at least X-ray tube voltage, X-ray tube current, exposure time, detector temperature, equipment operating temperature, magnetic field strength, and radio frequency pulse.

[0028] In a possible implementation, step S310 also includes that the medical image quality index includes at least an image contrast index, an image brightness index, a spatial resolution index, a temporal resolution index, and an image clarity index. Image contrast refers to the grayscale difference between different structures in the image, that is, the degree of difference between light and dark areas, reflecting the brightness contrast between different tissues or structures in the image. Appropriate contrast can highlight pathological changes in the image, allowing doctors to more clearly identify the diseased area and make an accurate diagnosis; image brightness refers to the overall brightness of the image, which determines the degree of visualization of details in the image. Images with high brightness can display more detailed information, which helps doctors find tiny lesions or structural abnormalities; spatial resolution refers to the size of the smallest object or detail that can be distinguished in the image, reflecting the clarity and detail expression of the image. In medical images, the higher the spatial resolution, The clearer the image, the more details the doctor can observe and the more accurate the diagnosis of the disease. Temporal resolution refers to the ability of an imaging system to repeatedly acquire image data in a short period of time, reflecting the imaging system's ability to capture dynamic changes. In the imaging of dynamic organs such as the heart and blood vessels, temporal resolution is particularly important. High temporal resolution can more accurately capture the movement of organs, reduce the interference of motion artifacts, and improve imaging quality. Image clarity refers to the clarity of the edges of objects in the image, reflecting the image's ability to express details. High-definition images can more accurately present the morphology and structural characteristics of objects, helping doctors to accurately locate and judge lesions.

[0029] The operating status indicators of medical imaging equipment also include at least X-ray tube voltage, X-ray tube current, exposure time, detector temperature, equipment operating temperature, magnetic field strength, and radio frequency pulses. The operating status indicators of medical imaging equipment are an important basis for evaluating the performance and working status of the equipment, and are directly related to the quality of medical images and the accuracy of diagnosis. Specifically, the X-ray tube voltage refers to the peak value of the operating voltage loaded between the cathode and anode of the X-ray tube. It is one of the important parameters for generating X-rays, affecting the penetration ability and radiation intensity of X-rays. The selection of X-ray tube voltage directly affects the contrast and clarity of imaging; the X-ray tube current refers to the current passing through the anode of the X-ray tube, which determines the output dose rate of X-rays, that is, the amount of X-rays generated per unit time. The size of the X-ray tube current directly affects the speed and quality of imaging; the exposure time refers to the length of time that X-rays are irradiated onto the detector or film, and is one of the key parameters for controlling the radiation dose of X-rays. Longer exposure times can improve the contrast and clarity of imaging, but will also increase the radiation dose, while shorter exposure times are suitable for occasions that require rapid imaging, but may reduce image quality; the detector temperature It refers to the temperature state of the detector during operation. The detector is a key component in medical imaging equipment for receiving and converting X-ray signals. Excessive temperature may cause the performance of the detector to deteriorate, affecting the clarity and contrast of the image. The operating temperature of the equipment refers to the temperature state of the medical imaging equipment during normal operation. This temperature range needs to be maintained within a certain range to ensure the performance and stability of the equipment. The stability of the operating temperature of the equipment is crucial to ensuring the imaging quality and extending the life of the equipment. The magnetic field strength refers to the size of the main magnetic field B0 in the MRI (magnetic resonance imaging) equipment. It is one of the important parameters of MRI imaging and determines the signal-to-noise ratio and imaging quality of the image. The higher the magnetic field strength, the higher the signal-to-noise ratio of the image and the better the imaging quality. The radio frequency pulse is an electromagnetic wave pulse used to excite the atomic nucleus to generate magnetic resonance signals during MRI imaging. It is an important part of MRI imaging technology. The accuracy and stability of the radio frequency pulse directly affect the clarity and contrast of the MRI image.

[0030] Step S400, performing feature detection on the medical image quality index based on the medical record image to obtain a second medical image quality index assignment result. Performing feature detection on the medical image quality index based on the medical record image to obtain a second medical image quality index assignment result specifically refers to using image processing technology and medical image analysis algorithms to extract and analyze key features in the medical record image to evaluate its quality and give a corresponding quantitative score. Specifically, the medical record image is preprocessed, including denoising, contrast enhancement, brightness correction, etc., to improve image quality and facilitate subsequent feature extraction. Image processing technologies such as edge detection, texture analysis, shape recognition, etc. are used to extract key features related to medical image quality from the medical record image. Features may include image resolution, contrast, noise level, artifacts, etc. The degree of shadow, clarity of lesions, etc. are then quantitatively evaluated for the extracted key features, and the quality of the image is reflected by calculating the corresponding index values. For example, the image contrast ratio, signal-to-noise ratio, artifact area ratio and other indicators can be calculated to evaluate the clarity and noise level of the image, and then the extracted features are compared and analyzed with the preset quality standards or reference images to evaluate whether the quality of the medical record images meets the requirements; according to the medical image quality indicators and actual application needs, a corresponding scoring system is established, and the results of quantitative evaluation and comparative analysis are input into the scoring system for calculation and assignment, and finally the second medical image quality index assignment result is output to reflect the quality level of the medical record images.

[0031] In a possible implementation, step S400 further includes step S401, classifying the medical image quality index to obtain an image time zone quality index and an image instantaneous quality index, wherein the image time zone quality index represents a feature type that requires analysis of differences between multiple moments, and the image instantaneous quality index represents a feature type that can be extracted by analyzing instantaneous images. Medical image quality indicators are divided into two categories, image time zone quality indicators and image instantaneous quality indicators, the difference being that the feature types they represent are different, image time zone quality indicators refer to those feature types that require analysis of differences between multiple moments (or frames) to be accurately evaluated, such as motion artifacts, temporal resolution, etc., and image instantaneous quality indicators refer to those feature types that can be extracted by analyzing instantaneous (or single-frame) images, such as noise level, contrast, clarity, etc.

[0032] The method further includes step S402, configuring a time series model according to the image time zone quality index to generate a first feature detection channel. Use a time series model (such as an autoregressive integrated moving average model, a long short-term memory network) to capture and analyze the features of the image time zone quality index. Specifically, configure the parameters of the model according to the specific requirements of the image time zone quality index and the characteristics of the medical image data, including setting the input dimension of the model (such as time step, number of features, etc.), hidden layer structure (such as number of layers, number of neurons, etc.), activation function, optimization algorithm, loss function, etc., to generate a first feature detection channel, which is responsible for receiving the time series data in the medical image data, and extracting feature information related to the image time zone quality index through analysis and processing of the model.

[0033] The method further includes step S403, wherein a convolutional neural network is configured according to the instantaneous image quality index to generate a second feature detection channel. The instantaneous image quality index refers to the type of medical image features that can be extracted by analyzing instantaneous (or single-frame) images, and is usually related to local attributes of the image, such as noise level, contrast, clarity, texture features, etc. Specifically, a detection model is constructed based on a convolutional neural network model, and the input image is abstracted and feature extracted layer by layer through structures such as convolutional layers and pooling layers. According to the specific requirements of the instantaneous image quality index, a suitable convolution kernel size and number are selected, and the step size and padding method of the convolution operation are set to control the size of the feature map and avoid information loss. An activation function is added to increase the nonlinear ability of the network. The pooling layer is used to reduce the spatial dimension (height and width) of the feature map, reduce the number of parameters and the amount of calculation, and maintain the main information of the feature. After the convolutional neural network is configured, it is used as a feature detection channel to receive a single-frame image in the medical image data, and extract feature information related to the instantaneous image quality index through layer-by-layer processing of the network.

[0034] The method further includes step S404, wherein the medical record image is processed according to the first feature detection channel and the second feature detection channel to obtain the second medical image quality index assignment result. The medical record image is processed using the configured first feature detection channel and the second feature detection channel, that is, the medical record image is input into the two channels respectively, and the feature information of the image time zone quality index and the image instantaneous quality index are extracted respectively, and quantitative evaluation is performed on them. Finally, the output results of the two channels are combined to obtain the second medical image quality index assignment result.

[0035] Step S500, the medical record image is scored for quality according to the first medical image quality index assignment result and the second medical image quality index assignment result to obtain a quality score value. The first medical image quality index assignment result and the second medical image quality index assignment result are used as input data and integrated into a comprehensive evaluation framework, and then the first and second assignment results are assigned corresponding weights according to the actual application requirements and the importance of medical image quality evaluation, and then the first and second assignment results are weighted and summed according to the assigned weights to obtain a preliminary quality score value, which is used to reflect the quality level of the medical record image, so as to understand the quality status of the medical image and take corresponding measures to improve the image quality. At the same time, the quality score value can also serve as an important basis for performance evaluation, fault warning and quality control of medical imaging equipment.

[0036] In a possible implementation, step S500 further includes step S510, extracting the first assignment result of the jth indicator according to the first medical image quality indicator assignment result, and extracting the second assignment result of the jth indicator according to the second medical image quality indicator assignment result. From the first medical image quality indicator assignment result and the second medical image quality indicator assignment result, the assignment result corresponding to the jth indicator is extracted, and j refers to one of the medical image quality indicators. It also includes step S520, calculating twice the product of the first assignment result of the jth indicator and the second assignment result of the jth indicator, adding a preset constant, and setting it as a first evaluation factor. Calculate twice the product of the first assignment result of the jth indicator and the second assignment result, and then add a preset constant as the first evaluation factor, the preset constant may be a value used for adjustment or standardization to ensure the rationality and comparability of the evaluation factor. It also includes step S530, calculating the square sum of the first assignment result of the jth indicator and the second assignment result of the jth indicator, adding the preset constant, and setting it as a second evaluation factor. The square sum of the first assignment result and the second assignment result of the j-th index is calculated, and then the same preset constant is added as the second evaluation factor.

[0037] It also includes step S540, calculating the ratio of the first evaluation factor to the second evaluation factor, setting it as the jth indicator quality coefficient, and adding it to the indicator quality coefficient set. The ratio of the first evaluation factor to the second evaluation factor is used as the quality coefficient of the jth indicator, which reflects the relative size between the two evaluation factors, so as to evaluate the comprehensive performance of the jth indicator under the two sets of assignment results, and then adding it to the indicator quality coefficient set, which includes the quality coefficients of all evaluation indicators (not just the jth indicator). It also includes step S550, weighting the medical image quality indicator to generate a weighted result of the medical image quality indicator. According to the importance of different indicators and the degree of influence on the quality of medical images, a weight value is assigned to each indicator, and the weight value reflects the relative importance of each indicator in the overall evaluation. After obtaining the weight of each indicator, the weighted result of the medical image quality indicator is formed, that is, a set containing all indicators and their corresponding weights. It also includes step S560, performing weighted mean calculation on the indicator quality coefficient set according to the weighted result of the medical image quality indicator to generate the quality score value. The previously calculated indicator quality coefficient and the corresponding weight value are used to perform a weighted mean calculation to generate a score value that comprehensively reflects the quality of the medical image, that is, to generate a quality score value.

[0038] In a possible implementation, step S550 further includes step S551, sending the medical imaging device model and the medical imaging quality index to M evaluators in parallel, obtaining M groups of medical imaging quality index importance scoring results, wherein the M evaluators are informationally isolated from each other. The medical imaging device model and related medical imaging quality index information are sent in parallel to M evaluators, wherein the M evaluators may be experts, doctors or other personnel with relevant professional knowledge, and the M evaluators need to maintain information isolation, that is, they cannot communicate with each other or see the scoring results of other evaluators to ensure the independence and objectivity of the scoring, each evaluator independently scores the importance of each quality indicator based on the received medical imaging device model and quality index information, each evaluator generates a set of scoring results on the importance of medical imaging quality indicators, and finally obtains M groups of medical imaging quality index importance scoring results.

[0039] The method further includes step S552, traversing the medical image quality indicators, grouping and reconstructing the M groups of medical image quality indicator importance score results, and obtaining a first indicator importance score set until an Nth indicator importance score set. Traverse all medical image quality indicators, and group and reconstruct the M groups of score results. Specifically, the score results of all evaluation parties for the same quality indicator are collected together to form a separate score set, and N score sets are obtained, each set corresponding to an importance score of a quality indicator, wherein N is the total number of quality indicators, and the score sets are respectively named as the first indicator importance score set until the Nth indicator importance score set.

[0040] The method further includes step S553, weighting the medical image quality indicators according to the first indicator importance score set to the Nth indicator importance score set, and generating a medical image quality indicator weighting result. After obtaining the N quality indicators, which reflect the importance of the indicator in the overall evaluation, the medical image quality indicators are weighted according to these scoring results. Specifically, the weight of each indicator is calculated and weighted, and finally a medical image quality indicator weighting result is obtained, which includes all quality indicators and their corresponding weights.

[0041] In a possible implementation, step S553 further includes step S5531, adding the first indicator importance score set until the Nth indicator importance score set to obtain the sum of indicator importance scores. All score values ​​from the first indicator importance score set to the Nth indicator importance score set are summed up respectively to obtain the total importance score of each indicator and the total importance score of the N indicators. It also includes step S5332, comparing the sum of the first indicator importance scores of the first indicator importance score set with the sum of the indicator importance scores, and setting it as the first indicator weighting result. The first indicator weighting result = the total importance score of the first indicator / the total indicator importance score, which represents the importance of the indicator among all indicators. It also includes step S5333, until the sum of the Nth indicator importance score of the Nth indicator importance score set is compared with the sum of the indicator importance scores, and setting it as the Nth indicator weighting result. Calculate the ratio of the total importance score of each indicator to the total importance score of the indicator, and obtain N indicator weighting results. The weighting result of each indicator is a value between 0 and 1, and the sum of the weighting results of all indicators should be equal to 1 or very close to 1. It also includes step S5334, adding the first indicator weighting result to the Nth indicator weighting result to the medical image quality indicator weighting result. The calculated weighting result of each indicator (from the first indicator weighting result to the Nth indicator weighting result) is added to the medical image quality indicator weighting result to determine the contribution of different quality indicators in the overall evaluation.

[0042] Step S600, when the quality score value is less than or equal to the quality score threshold, generate the first medical imaging device warning information. In the process of medical imaging quality assessment, if the quality score value of the medical recording image obtained by the comprehensive assessment method fails to reach the preset quality score threshold, the warning mechanism will be automatically triggered, and the warning information about the medical imaging device will be generated and issued, so as to timely remind relevant personnel to pay attention to the possible problems of the medical imaging device, so as to take necessary measures to intervene and repair, and ensure that the medical imaging device can continuously and stably provide high-quality medical imaging services. Specifically, the quality score threshold is set according to the actual situation and the assessment requirements, and can reflect the minimum quality level that the medical imaging device should achieve under normal operating conditions. When the calculated quality score value is less than or equal to the preset quality score threshold, the system will automatically trigger the warning mechanism, which may include but is not limited to sending emails, SMS notifications, generating eye-catching warning information, etc., to ensure that relevant personnel can receive the warning information in time. After receiving the warning information, the relevant personnel immediately inspect and evaluate the medical imaging device to determine the specific reasons for the decrease in the quality score value. According to the assessment results, measures are taken to repair the equipment failure, optimize the equipment performance, etc., to improve the quality level of medical imaging.

[0043] In a possible implementation, step S600 further includes step S610, when the quality score value is greater than the quality score threshold, multiple quality score values ​​at each moment are calculated according to the medical imaging device log data. When the quality score value is greater than the quality score threshold, it indicates that the quality of these images has reached a higher standard, and the medical imaging device log data associated with the medical image is accessed and read, and the medical imaging device log data is further analyzed, and image data at multiple representative moments (or batches) are selected from a large number of images collected by the device according to the timestamp or other identification information in the log data, and images at multiple moments (such as different time periods, different exposure settings, different patients, etc.) are selected for quality scoring, and corresponding quality score values ​​are calculated to reflect the specific quality characteristics of the images at that moment (or batch).

[0044] It also includes step S620, calculating the variance values ​​of the quality score values ​​at the multiple moments to obtain the equipment score fluctuation parameter. Then, the variance values ​​of the quality score values ​​at the multiple moments are calculated to obtain the equipment score fluctuation parameter, which reflects the stability of the image quality produced by the equipment at different moments or under different conditions. If the variance value is large, it means that the equipment score fluctuates greatly and there may be potential problems. It also includes step S630, when the equipment score fluctuation parameter is greater than or equal to the fluctuation parameter threshold, a second medical imaging equipment warning message is generated. When the equipment score fluctuation parameter is greater than or equal to the fluctuation parameter threshold, it means that there may be problems with the stability of the equipment, and the second medical imaging equipment warning message is immediately generated to remind relevant personnel to pay attention to the stability of the equipment and possible problems, so as to take timely measures for maintenance and repair, wherein the fluctuation parameter threshold is used to determine whether the equipment score fluctuation exceeds an acceptable range.

[0045] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. An intelligent analysis and early warning method for medical imaging equipment log data, characterized in that: include: Extracting the operation status record data of the medical imaging equipment according to the log data of the medical imaging equipment; Extracting medical record images according to the medical imaging device log data, wherein the medical record images have temporal consistency with the medical imaging device operation status record data; According to the medical imaging device operation status record data, combined with the medical imaging device model, the medical imaging quality index is sampled and assigned to obtain a first medical imaging quality index assignment result; Performing feature detection on the medical image quality index based on the medical record image to obtain a second medical image quality index assignment result, wherein performing feature detection on the medical image quality index based on the medical record image to obtain a second medical image quality index assignment result specifically comprises using image processing technology and medical image analysis algorithms to extract and analyze key features in the medical record image to evaluate its quality and give a corresponding quantitative score; Performing a quality score on the medical record image according to the first medical image quality index assignment result and the second medical image quality index assignment result to obtain a quality score value; When the quality score value is less than or equal to the quality score threshold, generating first medical imaging equipment warning information; According to the medical imaging device operation status record data, in combination with the medical imaging device model, the medical imaging quality index is sampled and assigned to obtain a first medical imaging quality index assignment result, including: Restricted by the medical imaging device operation status record data of the medical imaging device model, the qualified sample feature detection record data is collected online, wherein the qualified sample feature detection record data includes the medical imaging quality index record value; Performing positive processing on the recorded value of the medical image quality index to obtain a characteristic value of the medical image quality index; Traversing the medical image quality index characteristic values ​​to perform centralized data extraction, and obtaining centralized medical image quality index characteristic values; Traversing the concentrated medical image quality index characteristic values ​​to perform minimum value sorting to obtain the first medical image quality index assignment result; The method comprises: The medical image quality index at least includes an image contrast index, an image brightness index, a spatial resolution index, a temporal resolution index, and an image clarity index; The operating status indicators of medical imaging equipment include at least X-ray tube voltage, X-ray tube current, exposure time, detector temperature, equipment operating temperature, magnetic field strength, and radio frequency pulses; Performing a quality score on the medical record image according to the first medical image quality index assignment result and the second medical image quality index assignment result to obtain a quality score value, including: Extracting a first assignment result of the jth indicator according to the first medical image quality indicator assignment result, and extracting a second assignment result of the jth indicator according to the second medical image quality indicator assignment result; Calculate twice the product of the first assignment result of the j-th index and the second assignment result of the j-th index, add a preset constant, and set the result as the first evaluation factor; Calculate the square sum of the first value assignment result of the j-th index and the second value assignment result of the j-th index, add the preset constant, and set it as the second evaluation factor; Calculate the ratio of the first evaluation factor to the second evaluation factor, set it as the jth indicator quality coefficient, and add it to the indicator quality coefficient set; Weighting the medical image quality index to generate a medical image quality index weighting result; Performing weighted mean calculation on the indicator quality coefficient set according to the weighted result of the medical image quality indicator to generate the quality score value; The medical image quality index is weighted to generate a medical image quality index weighting result, including: Sending the medical imaging device model and the medical imaging quality index to M evaluation parties in parallel to obtain M groups of medical imaging quality index importance score results, wherein the M evaluation parties are informationally isolated from each other; Traversing the medical image quality indicators, grouping and reconstructing the M groups of medical image quality indicator importance score results, and obtaining a first indicator importance score set until an Nth indicator importance score set; weighting the medical image quality indicators according to the first indicator importance score set to the Nth indicator importance score set to generate a medical image quality indicator weighting result; The medical image quality indicators are weighted according to the first indicator importance score set to the Nth indicator importance score set to generate a medical image quality indicator weighting result, including: Adding the first indicator importance score set until the Nth indicator importance score set to obtain a sum of indicator importance scores; Compare the sum of the first indicator importance scores of the first indicator importance score set with the sum of the indicator importance scores, and set it as the first indicator weighting result; Until the sum of the Nth indicator importance score of the Nth indicator importance score set is compared with the sum of the indicator importance scores, and the result is set as the Nth indicator weighting result; Add the first indicator weighting result to the Nth indicator weighting result into the medical image quality indicator weighting result; Performing feature detection on the medical image quality index based on the medical record image to obtain a second medical image quality index assignment result, the method comprising: Classifying the medical image quality index to obtain an image time zone quality index and an image instantaneous quality index, wherein the image time zone quality index represents a feature type that requires analysis of differences in multiple moments, and the image instantaneous quality index represents a feature type that can be extracted by analyzing instantaneous images; Performing time series model configuration according to the image time zone quality index to generate a first feature detection channel; Performing convolutional neural network configuration according to the instantaneous image quality index to generate a second feature detection channel; Processing the medical record image according to the first feature detection channel and the second feature detection channel to obtain a second medical image quality index assignment result; The method further comprises: When the quality score value is greater than a quality score threshold, calculating quality score values ​​at multiple moments according to the medical imaging device log data; Calculate the variance of the quality score values ​​at the multiple moments to obtain a device score fluctuation parameter; When the device score fluctuation parameter is greater than or equal to the fluctuation parameter threshold, second medical imaging device warning information is generated.

Citation Information

Patent Citations

  • Medical imaging systems and methods with auto-correction of image quality-based on the log analysis of medical devices

    CN113747838A

  • Medical image data quality monitoring management method and system

    CN115393353A