Domain driven medical image quality control architecture

Abstract

This invention discloses a domain-driven medical image quality control architecture, relating to the field of medical imaging technology. Key technical points include: extracting the quality control feature parameters of the medical image to be controlled; comparing the quality control feature parameters with standard quality control feature intervals to obtain target quality control features; performing attribute classification and correlation judgment processing on the target quality control features to generate a quality control feature set; obtaining the feature association parameters of the target quality control features to which the quality control feature set belongs, as well as the target quality control feature parameters; performing verification processing and analysis on the feature association parameters to obtain target quality control representation features; analyzing the target quality control feature parameters to obtain target quality control coefficients; obtaining quality control priority coefficients based on the target quality control coefficients; and outputting multi-dimensional medical image quality control processing instructions for the medical image to be controlled based on the quality control feature order and the target quality control representation features. The effect is to improve the quality stability of medical images.

Description

Technical Field

[0001] This invention relates to the field of medical imaging technology, and more specifically, to a domain-driven medical imaging quality control architecture. Background Technology

[0002] In modern medical diagnostic and treatment systems, medical imaging has become a core tool for disease screening, diagnosis, and efficacy evaluation. Its quality directly determines the accuracy of clinical decisions and the safety of patient treatment. With the widespread use of imaging equipment such as CT, MRI, and ultrasound, and the surge in examination volume, the importance of image quality control has become increasingly prominent. Traditional medical imaging quality control frameworks lack standardized procedures, and the quality judgment results of the same image may vary significantly, making it difficult to match the priority of key imaging features with clinical diagnostic needs. Summary of the Invention

[0003] In view of the shortcomings of existing technologies, the purpose of this invention is to provide a domain-driven medical image quality control architecture.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] Domain-driven medical imaging quality control architecture includes:

[0006] Extraction module: Extracts the quality control feature parameters of the medical images to be quality controlled;

[0007] The comparison and generation module compares the feature parameters to be controlled with the standard quality control feature range to obtain the target quality control features, and performs attribute classification and correlation judgment on the target quality control features to generate a quality control feature set;

[0008] Acquisition module: Acquires the feature association parameters of the target quality control feature to which the quality control feature set belongs, as well as the target quality control feature parameters;

[0009] Analysis module: Verifies and analyzes the feature correlation parameters to obtain the target quality control characterization features; analyzes the target quality control feature parameters to obtain the target quality control coefficient;

[0010] Processing module: Obtains quality control priority coefficient based on target quality control coefficient, and obtains quality control feature order of standard quality control features based on quality control priority coefficient;

[0011] Output module: Outputs multi-dimensional medical image quality control processing instructions for the medical images to be controlled based on the quality control characteristics, quality control order, and target quality control characterization characteristics.

[0012] Preferably, the target quality control feature is obtained by comparing the feature parameter to be controlled with the standard quality control feature range, specifically including the following steps:

[0013] Extract the parameter generation time series information of the corresponding examination process of the medical image to be quality controlled, and combine the parameter generation time series information to determine the original feature parameters;

[0014] Retrieve the standard quality control feature range adapted to the clinical scenario, compare the original feature parameters with the standard quality control feature range, filter out the original deviation feature parameters that deviate from the standard quality control feature range and the original qualified feature parameters that conform to the standard quality control feature range, and obtain the time-series node information of the original deviation feature parameters.

[0015] Based on the time-series node information of the original deviation characteristic parameters, the subordinate characteristic parameters of the corresponding time-series node information are located. The subordinate characteristic parameters are compared with the standard quality control characteristic range, and subordinate deviation characteristic parameters that deviate from the standard quality control characteristic range and subordinate qualified characteristic parameters that conform to the standard quality control characteristic range are selected.

[0016] The transmission relationship between the original deviation characteristic parameter and the subordinate deviation characteristic parameter is determined, and the transmission determination result is obtained by judging whether the subordinate deviation characteristic parameter transmits to trigger the original deviation characteristic parameter.

[0017] A temporary quality control feature set is constructed by integrating the original qualified feature parameters and the subordinate qualified feature parameters;

[0018] The target quality control features are obtained by regularizing the temporary quality control feature set based on the transmission judgment results.

[0019] Preferably, the target quality control features are subjected to attribute segmentation and correlation judgment processing to generate a quality control feature set, specifically including the following steps:

[0020] After classifying the target quality control features by attributes to obtain compliant target quality control features and abnormal target quality control features, a preliminary screening feature set is formed.

[0021] The correlation between the quality control features of compliant targets and the quality control features of abnormal targets in the initial screening feature set is determined, and the quality control features of compliant targets and the quality control features of abnormal targets that have mutual influence are marked as feature groups;

[0022] Based on the degree of influence of the target quality control features on medical image diagnosis, the feature group is divided into primary features and secondary features to generate a quality control feature set.

[0023] Preferably, the target quality control characterization features are obtained by verifying and analyzing the feature association parameters, specifically including the following steps:

[0024] The feature association parameters are validated to form a valid set of association parameters.

[0025] By exploring the mutual constraints among parameters in the effective set of related parameters, and clarifying the positive linkages and negative inhibitions between different parameters based on these mutual constraints, a parameter constraint relationship map is formed.

[0026] Based on the parameter constraint relationship map, locate the target-related parameters that have a dominant influence on the target quality control characteristics;

[0027] Tracing the cause location results corresponding to the target's associated parameters;

[0028] Based on the results of the cause localization and the parameter constraint relationship map, the influence of the target quality control characteristics is judged and the characteristic influence status is formed;

[0029] The characteristics affecting the system are integrated to form the target quality control characterization features.

[0030] Preferably, the target quality control coefficient is obtained by analyzing the target quality control characteristic parameters, specifically including the following steps:

[0031] Based on the dependency and constraint attributes of the target quality control feature parameters, the strongly correlated feature parameter group and the weakly correlated feature parameter group are distinguished.

[0032] The core influence parameters are obtained by processing the strongly correlated feature parameter group and the weakly correlated feature parameter group.

[0033] Based on the interaction law of strongly correlated feature parameter groups, the basic influence values ​​of the core influence parameters are superimposed and adjusted to obtain the adjusted influence values.

[0034] The target quality control characteristic parameters are quantified and integrated based on the adjusted influence values ​​to obtain the target quality control coefficient.

[0035] Preferably, the core influencing parameters are obtained by processing the strongly correlated feature parameter set and the weakly correlated feature parameter set, specifically including the following steps:

[0036] The core dominant parameters are obtained by judging the interaction law of parameters within a strongly correlated feature parameter group, the dominant direction of the quality control influence of the core dominant parameters is clarified, and the dominant parameter set of the core dominant parameters is extracted based on the dominant direction of the quality control influence.

[0037] Assess the basic influence of the dominant parameter set and weakly correlated feature parameters on the quality control results of medical images, and determine the core influencing parameters based on the basic influence.

[0038] Preferably, the quality control priority coefficient is obtained based on the target quality control coefficient, specifically including the following steps:

[0039] The actual deviation of the corresponding quality control feature is defined based on the target quality control coefficient, and the deviation definition result is obtained by clarifying the extent to which each target quality control coefficient deviates from the standard quality control feature range based on the actual deviation.

[0040] The assessment results of the transmission effect are obtained based on the deviation definition results.

[0041] The associated clinical diagnosis and treatment risk level is determined based on the results of the deviation transmission impact assessment.

[0042] The priority coefficient for quality control is determined based on the results of the associated clinical diagnosis and treatment risk levels.

[0043] Preferably, the quality control priority of the standard quality control features is obtained according to the quality control priority coefficient, which specifically includes the following steps:

[0044] Several hierarchical intervals are determined based on the numerical distribution of the quality control priority coefficient, and each hierarchical interval corresponds to a unique priority level;

[0045] The correlation results are obtained by determining the correlation degree of standard quality control features with priority levels.

[0046] The correlation between the standard quality control characteristics and the target quality control characteristics is confirmed based on the correlation results and the target quality control characterization characteristics.

[0047] The quality control features are sorted according to their priority levels based on their degree of correlation to form a quality control feature quality control order.

[0048] Preferably, the multi-dimensional medical image quality control processing instructions for the medical images to be controlled are output according to the quality control feature order and the target quality control characterization features, specifically including the following steps:

[0049] Based on the quality control characteristics and quality control order, the target quality control characterization characteristics are hierarchically sorted out, and the quality control characterization characteristics corresponding to each priority are selected.

[0050] Determine the type and impact of quality control anomalies based on the quality control characteristics and the quality control sequence to determine the processing priority of each anomaly type.

[0051] A multi-dimensional collaborative processing framework is obtained by processing the types, impact levels, and processing priorities of quality control anomalies.

[0052] Based on the multi-dimensional collaborative processing framework, the adjustment rules for processing content in different quality control scenarios are clarified, resulting in multi-dimensional medical image quality control processing instructions.

[0053] Preferably, a multi-dimensional collaborative processing framework is obtained by processing the quality control anomaly type, impact level, and processing priority, specifically including the following steps:

[0054] Generate basic quality control processing content at the corresponding level based on the type and degree of impact of quality control anomalies;

[0055] By combining the basic quality control processing content to determine the correlation between different processing priorities, a multi-dimensional collaborative processing framework is formed.

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] This invention achieves a deep integration of quality control work with clinical diagnosis and treatment scenarios. Prioritizing quality control features related to nodule boundary clarity and density measurement accuracy ensures that quality control results effectively support clinical diagnostic decisions, avoiding ineffective quality control divorced from reality and enhancing the clinical value of quality control work. A fully traceable, closed-loop quality control system is constructed. When image quality issues arise, medical staff can trace back to the original causes, such as equipment calibration cycles and patient positioning. This not only allows for rapid rectification of the current problem but also enables the optimization of quality control rules through debriefing, forming a continuous iterative quality control improvement mechanism. Multi-dimensional collaborative processing enhances quality control efficiency and resource utilization. Quality control priorities are divided according to clinical risk levels, allowing high-risk equipment failures and core parameter drift issues to be addressed first, preventing low-priority issues from consuming excessive resources. The multi-dimensional collaborative processing framework supports the parallel execution of some low-correlation quality control tasks, significantly improving the quality stability of medical images. Attached Figure Description

[0058] Figure 1 A schematic diagram of a domain-driven medical image quality control architecture is provided for embodiments of the present invention;

[0059] Figure 2 This diagram illustrates the steps involved in obtaining a quality control feature set within a domain-driven medical image quality control architecture, as provided in this embodiment of the invention. Detailed Implementation

[0060] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0061] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0062] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places throughout this specification does not necessarily refer to the same embodiment, nor is it a single embodiment or an embodiment selectively excluded from other embodiments.

[0063] Reference Figures 1-2 As shown.

[0064] The embodiments further illustrate the domain-driven medical image quality control architecture proposed in this invention.

[0065] Domain-driven medical imaging quality control architecture includes:

[0066] Extraction module: Extracts the quality control feature parameters of the medical images to be quality controlled;

[0067] The comparison and generation module compares the feature parameters to be controlled with the standard quality control feature range to obtain the target quality control features, and performs attribute classification and correlation judgment on the target quality control features to generate a quality control feature set;

[0068] Acquisition module: Acquires the feature association parameters of the target quality control feature to which the quality control feature set belongs, as well as the target quality control feature parameters;

[0069] Analysis module: Verifies and analyzes the feature correlation parameters to obtain the target quality control characterization features; analyzes the target quality control feature parameters to obtain the target quality control coefficient;

[0070] Processing module: Obtains quality control priority coefficient based on target quality control coefficient, and obtains quality control feature order of standard quality control features based on quality control priority coefficient;

[0071] Output module: Outputs multi-dimensional medical image quality control processing instructions for the medical images to be controlled based on the quality control characteristics, quality control order, and target quality control characterization characteristics.

[0072] The target quality control feature is obtained by comparing the feature parameter to be controlled with the standard quality control feature range. This includes the following steps:

[0073] Extract the parameter generation time series information of the corresponding examination process of the medical image to be quality controlled, and combine the parameter generation time series information to determine the original feature parameters;

[0074] Retrieve the standard quality control feature range adapted to the clinical scenario, compare the original feature parameters with the standard quality control feature range, filter out the original deviation feature parameters that deviate from the standard quality control feature range and the original qualified feature parameters that conform to the standard quality control feature range, and obtain the time-series node information of the original deviation feature parameters.

[0075] Based on the time-series node information of the original deviation characteristic parameters, the subordinate characteristic parameters of the corresponding time-series node information are located. The subordinate characteristic parameters are compared with the standard quality control characteristic range, and subordinate deviation characteristic parameters that deviate from the standard quality control characteristic range and subordinate qualified characteristic parameters that conform to the standard quality control characteristic range are selected.

[0076] The transmission relationship between the original deviation characteristic parameter and the subordinate deviation characteristic parameter is determined, and the transmission determination result is obtained by judging whether the subordinate deviation characteristic parameter transmits to trigger the original deviation characteristic parameter.

[0077] A temporary quality control feature set is constructed by integrating the original qualified feature parameters and the subordinate qualified feature parameters;

[0078] The target quality control features are obtained by regularizing the temporary quality control feature set based on the transmission judgment results.

[0079] In practical operation, all parameters of the corresponding examination process for the medical images to be quality controlled are extracted and parameter generation time-series information is generated. For example, in a CT chest scan scenario, the tube voltage, tube current, slice thickness, and scanning parameters of the reconstruction algorithm are recorded in chronological order from the patient's positioning, as well as the post-processing parameters of window width and window level in the subsequent image reconstruction process. Combining the parameter generation time-series information, the discrete parameters are transformed into raw feature parameters with time node markers. For example, the tube voltage value at 10 seconds and the slice thickness value at 25 seconds are explicitly marked as the raw feature parameters of the corresponding time-series nodes.

[0080] Retrieve standard quality control feature ranges suitable for the current clinical scenario. Taking a routine chest CT examination as an example, the standard range for tube voltage is set between 100kV and 120kV, and the standard range for slice thickness is set between 5mm and 10mm. Each original feature parameter is compared one by one with its corresponding standard range, filtering out original deviation feature parameters that deviate from the standard range and original qualified feature parameters that conform to the standard range. Simultaneously, the timing information corresponding to each original deviation feature parameter is recorded. For example, when the tube voltage is 90kV at the 10th second, this parameter is marked as an original deviation feature parameter, and its corresponding timing point is recorded as the 10th second.

[0081] Based on the time-series node information of the original deviation characteristic parameters, the subordinate characteristic parameters corresponding to the node are located. Taking the tube voltage deviation at the 10th second as an example, the subordinate characteristic parameters of tube current and tube temperature at the same time node are extracted. The subordinate characteristic parameters have a direct equipment linkage relationship with the tube voltage. The subordinate characteristic parameters are compared with the corresponding standard quality control characteristic range to filter out the subordinate deviation characteristic parameters and the subordinate qualified characteristic parameters. For example, if the tube current at the 10th second is 200mA, and the corresponding standard range is 150mA to 180mA, the tube current parameter is marked as a subordinate deviation characteristic parameter.

[0082] After screening the deviation parameters, the transmission relationship between the original deviation characteristic parameters and the subordinate deviation characteristic parameters is determined. Based on the physical principles of medical imaging equipment and the logical connection of the examination process, it is determined whether the subordinate deviation is the cause of the original deviation. If an abnormal increase in tube current is found to lead to excessive tube load, thereby triggering the equipment to automatically reduce the tube voltage to protect the hardware, then it is determined that the subordinate deviation characteristic parameter has caused the original deviation characteristic parameter. If there is no direct logical connection between the abnormal tube current and the tube voltage deviation, it is determined that there is no transmission relationship.

[0083] All original qualified feature parameters and subordinate qualified feature parameters are integrated to construct a temporary quality control feature set. If the slice thickness parameter at the 15th second of a chest CT scan is 8mm, which meets the standard range of 5mm to 10mm, this parameter is included in the temporary quality control feature set; if the window width parameter at the 20th second is 350HU, which meets the standard range of 300HU to 400HU, it is also included in the temporary quality control feature set.

[0084] Based on the conduction determination results, the temporary quality control feature set is regularized to generate target quality control features. If a conduction relationship is determined, redundant qualified parameters caused by the conduction relationship in the temporary quality control feature set are removed, and key deviation parameters of the conduction chain are added. For example, in the case where pipe current causes pipe voltage deviation, the pipe current deviation parameter is added to the quality control features, while some qualified parameters affected by conduction are removed, ultimately forming target quality control features that can reflect the quality risk conduction path. If no conduction relationship is determined, the temporary quality control feature set is directly retained as the target quality control features to ensure that it can fully represent the true state of image quality.

[0085] When quantifying the degree of deviation, the deviation rate of the original deviation characteristic parameter is calculated as follows: Original deviation rate = (Original characteristic parameter value - Standard interval median) / Standard interval width × 100%. For example, the standard range of tube voltage is 100kV to 120kV, the median is 110kV, and the width is 20kV. When the original characteristic parameter value is 90kV, substituting it into the formula, we get the original deviation rate = (90-110) / 20 × 100% = -100%. A negative value indicates that the parameter is lower than the lower limit of the standard interval.

[0086] Subordinate deviation rate = (subordinate characteristic parameter value - standard interval median) / standard interval width × 100%. For example, the standard interval for tube current is 150mA to 180mA, the median is 165mA, and the width is 30mA. When the subordinate characteristic parameter value is 200mA, substituting it into the formula, we get subordinate deviation rate = (200-165) / 30×100%≈116.67%. A positive value indicates that the parameter is higher than the upper limit of the standard interval.

[0087] The target quality control features are divided into attributes and their correlation is determined to generate a quality control feature set. This process includes the following steps:

[0088] After classifying the target quality control features by attributes to obtain compliant target quality control features and abnormal target quality control features, a preliminary screening feature set is formed.

[0089] The correlation between the quality control features of compliant targets and the quality control features of abnormal targets in the initial screening feature set is determined, and the quality control features of compliant targets and the quality control features of abnormal targets that have mutual influence are marked as feature groups;

[0090] Based on the degree of influence of the target quality control features on medical image diagnosis, the feature group is divided into primary features and secondary features to generate a quality control feature set.

[0091] First, the target quality control features are categorized into compliant and abnormal target quality control features, thus forming a preliminary screening feature set. Taking a chest CT scan as an example, if the slice thickness parameter is 8mm, falling within the standard range of 5mm to 10mm, this parameter is classified as a compliant target quality control feature; if the tube voltage parameter is 90kV, deviating from the standard range of 100kV to 120kV, it is classified as an abnormal target quality control feature. All features that have undergone attribute categorization are integrated to form a preliminary screening feature set containing both categories of features.

[0092] The correlation between compliant and abnormal target quality control features in the initial screening feature set is determined, and features with mutual influence are marked as feature groups. For example, in a chest CT scenario, a low tube voltage, an abnormal target quality control feature, directly affects the contrast of compliant target quality control images. This is because a low tube voltage leads to insufficient X-ray penetration, resulting in an abnormally high image contrast; the two have a clear causal relationship and are therefore marked as the same feature group. If the abnormal target quality control feature is patient position deviation, while the compliant target quality control feature is lesion boundary clarity, the position deviation causes the lesion to deviate from the preset scan center in the image, thus affecting the determination of boundary clarity; the two are also marked as one feature group.

[0093] After labeling the feature groups, each feature group is divided into primary and secondary features based on the degree of influence of the target quality control features on medical image diagnosis, ultimately generating a quality control feature set. In the feature group of tube voltage and image contrast, abnormal tube voltage is the direct root cause of image quality problems and has a more fundamental impact on diagnostic results, therefore it is classified as a primary feature; while changes in image contrast are a result of abnormal tube voltage and have a relatively indirect impact on diagnosis, therefore it is classified as a secondary feature. In the feature group of body position deviation and lesion boundary clarity, body position deviation is the core cause of subsequent image quality problems, and is therefore classified as a primary feature; changes in lesion boundary clarity are a derivative effect of body position deviation, and are therefore classified as secondary features.

[0094] Feature influence weight = feature deviation rate × clinical diagnostic dependence coefficient, where the feature deviation rate is the original deviation rate or subordinate deviation rate, and the clinical diagnostic dependence coefficient is the importance coefficient of the feature in image diagnosis, with a value range of 0 to 1. For example, the clinical diagnostic dependence coefficient of tube voltage is 0.85, the clinical diagnostic dependence coefficient of image contrast is 0.75, the clinical diagnostic dependence coefficient of body position deviation is 0.90, and the clinical diagnostic dependence coefficient of lesion boundary clarity is 0.80.

[0095] Taking tube voltage anomaly as an example, if its original deviation rate is -100% and the clinical diagnostic dependence coefficient is 0.85, substituting it into the formula yields a feature influence weight of -100% × 0.85 = -0.85. Meanwhile, the compliance feature deviation rate for image contrast is 0%, and the clinical diagnostic dependence coefficient is 0.75, resulting in a feature influence weight of 0% × 0.75 = 0. By comparing the absolute values ​​of the feature influence weights, it is clear that tube voltage anomaly has a higher impact, thus classifying it as the primary feature, while image contrast is classified as a secondary feature.

[0096] Taking body position deviation as an example, if its original deviation rate is 60% and its clinical diagnostic dependence coefficient is 0.90, the feature influence weight is 60% × 0.90 = 0.54; the compliant feature deviation rate of lesion boundary clarity is 0%, the clinical diagnostic dependence coefficient is 0.80, and the feature influence weight is 0% × 0.80 = 0. Based on the weight, body position deviation is classified as the primary feature, and lesion boundary clarity is classified as the secondary feature.

[0097] The target quality control characterization features are obtained by verifying and analyzing the feature association parameters, specifically including the following steps:

[0098] The feature association parameters are validated to form a valid set of association parameters.

[0099] By exploring the mutual constraints among parameters in the effective set of related parameters, and clarifying the positive linkages and negative inhibitions between different parameters based on these mutual constraints, a parameter constraint relationship map is formed.

[0100] Based on the parameter constraint relationship map, locate the target-related parameters that have a dominant influence on the target quality control characteristics;

[0101] Tracing the cause location results corresponding to the target's associated parameters;

[0102] Based on the results of the cause localization and the parameter constraint relationship map, the influence of the target quality control characteristics is judged and the characteristic influence status is formed;

[0103] The characteristics affecting the system are integrated to form the target quality control characterization features.

[0104] First, the feature-related parameters are validated to form a valid set of related parameters. In a chest CT scan scenario, feature-related parameters typically include multiple external parameters such as equipment calibration cycle, room temperature and humidity, patient body mass index, and scanning bed position accuracy. The reasonableness of each parameter is verified. For example, if the equipment calibration cycle is recorded as 180 days, exceeding the standard 90-day cycle, the parameter is marked as invalid and discarded; if the room temperature is recorded as 28 degrees Celsius, within the reasonable range of 18 to 30 degrees Celsius, it is retained. All parameters that meet the reasonableness requirements after validation are integrated into a valid set of related parameters.

[0105] By exploring the interrelationships among parameters in the effective set of correlated parameters, and clarifying the positive linkages and negative inhibitions between different parameters, a parameter constraint relationship map can be formed. For example, in a chest CT scenario, increased humidity in the machine room will negatively inhibit the signal acquisition sensitivity of the detector, showing a negative inhibition relationship; while increased equipment tube voltage will positively enhance X-ray penetration, thus positively linking to the enhancement of image contrast, showing a positive linkage relationship.

[0106] Based on parameter constraint relationship mapping, target-related parameters that have a dominant influence on target quality control features were identified. Taking abnormal tube voltage in chest CT as an example, the parameter constraint relationship mapping revealed that expired equipment calibration cycles led to decreased tube voltage output accuracy, and excessively high room temperatures accelerated tube aging, thus affecting voltage stability. Among these factors, the impact of expired equipment calibration cycles was greater, and therefore it was identified as a target-related parameter. In the target quality control feature of patient positional deviation, the parameter constraint relationship mapping revealed that insufficient scanning bed position accuracy was the core related parameter causing positional deviation, and thus it was identified as a target-related parameter.

[0107] The causes of the target-related parameters were traced and located. For target-related parameters whose equipment calibration cycle had expired, the equipment maintenance log was traced, and it was found that the calibration was not performed on time due to the equipment maintenance personnel's shift scheduling delay. For target-related parameters with insufficient scanning bed position accuracy, the equipment operation data was traced, and it was found that the position deviation exceeded the allowable range due to the wear of the scanning bed guide rail.

[0108] By combining the causal localization results with the parameter constraint relationship map, the impact of the target quality control characteristics is judged, thus forming the characteristic impact status. Taking the abnormal tube voltage caused by the equipment calibration cycle exceeding the deadline as an example, the low tube voltage leads to insufficient X-ray penetration, resulting in abnormally high image contrast, decreased boundary recognition of small lung nodules, and increased risk of missed diagnosis; taking the positional deviation caused by the wear of the scanning bed guide rail as an example, the positional deviation causes the lesion to deviate from the scanning center, and artifacts appear at the edge of the lesion during image reconstruction, thus affecting the accurate measurement of the lesion size.

[0109] All influencing factors are integrated to form the target quality control characterization features. These characterization features not only include deviation information of the target quality control features, but also cover the causes of deviations, the constraints between parameters, and their specific impact on image diagnosis. For example, the target quality control characterization features for abnormal tube voltage fully present the causes of expired equipment calibration cycles, the positive correlation between tube voltage and image contrast, and the impact on lung nodule identification; the target quality control characterization features for postural deviation fully present the causes of scan bed guide rail wear, the inverse suppression relationship between scan bed position and lesion artifacts, and the impact on lesion measurement accuracy.

[0110] The parameter constraint coefficient = deviation rate of the affected parameter × correlation sensitivity coefficient, where the deviation rate of the affected parameter is the deviation rate of the target correlated parameter, and the correlation sensitivity coefficient is the sensitivity of the affected parameter to the target correlated parameter, with a value ranging from 0 to 1. For example, in the constraint relationship between room humidity and detector sensitivity, if the room humidity deviation rate is 30% and the correlation sensitivity coefficient is 0.7, then the parameter constraint coefficient is 30% × 0.7 = 0.21. This value reflects the reverse suppression strength of room humidity on detector sensitivity. In the constraint relationship between tube voltage and image contrast, if the tube voltage deviation rate is -100% and the correlation sensitivity coefficient is 0.6, then the parameter constraint coefficient is -100% × 0.6 = -0.6. This value reflects the positive linkage strength of tube voltage on image contrast.

[0111] The target quality control coefficient is obtained by analyzing the target quality control characteristic parameters, specifically including the following steps:

[0112] Based on the dependency and constraint attributes of the target quality control feature parameters, the strongly correlated feature parameter group and the weakly correlated feature parameter group are distinguished.

[0113] The core influencing parameters are obtained by processing the strongly correlated feature parameter set and the weakly correlated feature parameter set. The specific steps include:

[0114] The core dominant parameters are obtained by judging the interaction law of parameters within a strongly correlated feature parameter group, the dominant direction of the quality control influence of the core dominant parameters is clarified, and the dominant parameter set of the core dominant parameters is extracted based on the dominant direction of the quality control influence.

[0115] Evaluate the basic influence of the dominant parameter set and weakly correlated feature parameters on the quality control results of medical images, and determine the core influencing parameters based on the basic influence.

[0116] Based on the interaction law of strongly correlated feature parameter groups, the basic influence values ​​of the core influence parameters are superimposed and adjusted to obtain the adjusted influence values.

[0117] The target quality control characteristic parameters are quantified and integrated based on the adjusted influence values ​​to obtain the target quality control coefficient.

[0118] Based on the dependency and constraint attributes of the target quality control characteristic parameters, they are divided into strongly correlated characteristic parameter groups and weakly correlated characteristic parameter groups. In the context of chest CT examination, the strongly correlated characteristic parameter group includes tube voltage, tube current, and image contrast. For example, changes in tube voltage directly affect the output of tube current, thereby affecting image contrast. The weakly correlated characteristic parameter group includes room temperature and humidity, and patient respiratory rate. For example, slight fluctuations in room humidity usually do not directly cause abnormalities in tube voltage.

[0119] The strongly correlated and weakly correlated feature parameter groups are processed to extract core influencing parameters. The interaction patterns of parameters within the strongly correlated feature parameter groups are determined, identifying the core dominant parameters and clarifying their dominant direction of quality control influence. In the strongly correlated parameter group of chest CT, tube voltage is the core dominant parameter, and its dominant direction of quality control influence is to directly determine image contrast and noise level by adjusting X-ray penetration. Around this dominant direction of quality control influence, tube voltage, directly correlated tube current, and affected image contrast are extracted as the dominant parameter set.

[0120] The fundamental impact of the dominant parameter set and weakly correlated feature parameters on medical image quality control results was assessed. A fundamental impact value was assigned to each parameter based on clinical diagnostic needs and equipment physical characteristics; for example, tube voltage had a fundamental impact of 0.9, tube current 0.75, image contrast 0.8, room temperature and humidity 0.3, and patient respiratory rate 0.25. Parameters with a fundamental impact value higher than a set threshold (e.g., 0.6) were selected as core impact parameters. In this case, tube voltage, tube current, and image contrast were identified as core impact parameters.

[0121] Based on the interaction patterns of strongly correlated feature parameter groups, the base influence values ​​of core influencing parameters are adjusted by superposition to obtain the adjusted influence values. Since there are positive linkages or negative inhibitions between strongly correlated parameters, directly using the base influence values ​​will lead to evaluation bias; therefore, superposition correction is necessary. For example, tube voltage and tube current have a positive linkage relationship; when both are abnormal simultaneously, their combined influence is greater than the simple sum of their individual base influence values. Image contrast is affected by both tube voltage and tube current, and its influence needs to deduct the portion that was calculated repeatedly.

[0122] The adjusted impact factor = base impact factor + Σ(base impact factor of related parameters × linkage coefficient), where the linkage coefficient reflects the strength of the correlation between parameters; it is positive for positive linkage and negative for negative suppression. Taking tube voltage as an example, its base impact factor is 0.9, the linkage coefficient of tube current is 0.4, and the base impact factor of tube current is 0.75. Substituting these values ​​into the formula, the adjusted impact factor of tube voltage is 0.9 + 0.75 × 0.4 = 1.2. Similarly, the base impact factor of image contrast is 0.8, the linkage coefficient of tube voltage is 0.5, and the linkage coefficient of tube current is 0.3; the adjusted impact factor is 0.8 + 0.9 × 0.5 + 0.75 × 0.3 = 1.575.

[0123] The target quality control coefficient is obtained by quantifying and integrating the target quality control characteristic parameters based on the adjusted impact value. The target quality control coefficient = Σ(adjusted impact × characteristic deviation rate). If the characteristic deviation rate of tube voltage is -100%, the adjusted impact is 1.2; the characteristic deviation rate of tube current is 50%, the adjusted impact is 0.75; and the characteristic deviation rate of image contrast is 30%, the adjusted impact is 1.575. Substituting these values ​​into the formula, the target quality control coefficient is 1.2 × (-1.0) + 0.75 × 0.5 + 1.575 × 0.3 = -0.3525. This negative value indicates that the overall quality control characteristics are biased towards the lower than the standard range, and its absolute value reflects the severity of the risk.

[0124] The quality control priority coefficient is obtained based on the target quality control coefficient, specifically including the following steps:

[0125] The actual deviation of the corresponding quality control feature is defined based on the target quality control coefficient, and the deviation definition result is obtained by clarifying the extent to which each target quality control coefficient deviates from the standard quality control feature range based on the actual deviation.

[0126] The assessment results of the transmission effect are obtained based on the deviation definition results.

[0127] The associated clinical diagnosis and treatment risk level is determined based on the results of the deviation transmission impact assessment.

[0128] The priority coefficient for quality control is determined based on the results of the associated clinical diagnosis and treatment risk levels.

[0129] First, the actual deviation degree of the corresponding quality control feature is defined based on the target quality control coefficient, clarifying the range of deviation of each target quality control coefficient from the standard quality control feature range, thus obtaining the deviation definition result. Taking chest CT examination as an example, if the target quality control coefficient is -0.3525, it is compared with the preset deviation range. For example, the coefficient is divided into severe deviation, moderate deviation, and mild deviation. Among them, a coefficient with an absolute value greater than 0.5 is severe deviation, between 0.2 and 0.5 is moderate deviation, and less than 0.2 is mild deviation. In this case, the absolute value of the target quality control coefficient is 0.3525, which is in the range of 0.2 to 0.5 and is defined as moderate deviation.

[0130] Based on the deviation definition results, the propagation impact of the deviation in the quality control feature chain is assessed, resulting in a deviation propagation impact assessment. In cases of abnormal tube voltage in chest CT scans, moderate deviations in tube voltage can propagate and affect tube current output and image contrast, leading to decreased clarity of lung nodule boundaries. The length of the propagation path and the number of features involved are determined. For example, if the deviation only propagates to tube current and image contrast without triggering a wider cascading equipment abnormality, it is assessed as a local propagation impact. If the deviation causes multi-module failures in the equipment and results in abnormalities in multiple imaging parameters, it is assessed as a widespread propagation impact.

[0131] The clinical risk level was determined based on the assessment of the effects of deviation transmission. In the case of chest CT, moderate deviation with local transmission effects led to decreased identification of pulmonary nodules, thus increasing the risk of missed diagnoses, but did not directly endanger the patient's life; therefore, it was classified as intermediate clinical risk. If the deviation was severe and had widespread transmission effects, such as causing equipment malfunction leading to scan interruption or rendering the images completely unusable for diagnosis, it was classified as high clinical risk. If the deviation was mild and had no transmission effects, such as only a slight increase in background noise that did not affect lesion observation, it was classified as low clinical risk.

[0132] Quality control priority coefficients are determined based on the correlation of clinical diagnosis and treatment risk levels. Different risk levels are assigned corresponding priority coefficients; for example, high risk corresponds to a priority coefficient of 3, medium risk to a priority coefficient of 2, and low risk to a priority coefficient of 1. For instance, a priority coefficient of 2 corresponds to a medium clinical risk. This value serves as the basis for prioritizing the processing of this quality control characteristic, ensuring that high-risk quality control issues are addressed first.

[0133] The coverage of conduction effects is calculated as follows: (Number of features affected by conduction / Total number of quality control features) × 100%. For example, in a chest CT case, the features affected by conduction are tube current and image contrast, totaling 2 items. Since the total number of quality control features is 5, the coverage of conduction effects is calculated as 2 / 5 × 100% = 40%. This value reflects the extent of deviation conduction and provides a quantitative basis for determining clinical risk levels.

[0134] The determination of clinical diagnosis and treatment risk level can combine the deviation amplitude and the coverage of transmission impact. The risk score is calculated using the following formula: Risk Score = Deviation Amplitude Weight × Deviation Level Coefficient + Transmission Impact Weight × Transmission Impact Coverage. The deviation amplitude weight and transmission impact weight can be set according to clinical needs, for example, 0.6 and 0.4 respectively. The deviation level coefficient corresponds to values ​​of 3, 2, and 1 for severe, moderate, and mild deviations, respectively. In the chest CT case, the deviation level coefficient is 2, and the transmission impact coverage is 40%. Substituting these values ​​into the formula, we get the risk score = 0.6 × 2 + 0.4 × 0.4 = 1.2 + 0.16 = 1.36. Risk levels are classified based on the risk score; for example, a score greater than 2 is high risk, between 1 and 2 is medium risk, and less than 1 is low risk. Therefore, this case is classified as medium risk.

[0135] The quality control priority order of standard quality control features is obtained based on the quality control priority coefficient, which specifically includes the following steps:

[0136] Several hierarchical intervals are determined based on the numerical distribution of the quality control priority coefficient, and each hierarchical interval corresponds to a unique priority level;

[0137] The correlation results are obtained by determining the correlation degree of standard quality control features with priority levels.

[0138] The correlation between the standard quality control characteristics and the target quality control characteristics is confirmed based on the correlation results and the target quality control characterization characteristics.

[0139] The quality control features are sorted according to their priority levels based on their degree of correlation to form a quality control feature quality control order.

[0140] First, several hierarchical intervals are determined based on the numerical distribution of the quality control priority coefficients, with each interval corresponding to a unique priority level. Taking the quality control scenario of chest CT examination as an example, if the quality control priority coefficients of all current quality control features are distributed between 1 and 3, they are divided into three hierarchical intervals: for example, 1.0 to 1.5 corresponds to low priority, 1.5 to 2.5 corresponds to medium priority, and 2.5 to 3.0 corresponds to high priority. Each hierarchical interval is assigned a clear priority label.

[0141] The correlation degree of standard quality control features at the same priority level is determined to obtain the correlation results. The medium-priority feature group in chest CT includes abnormal tube voltage, tube current fluctuation, and abnormal image contrast. The logical relationships between these features are determined; for example, abnormal tube voltage and tube current fluctuation have a device-linked relationship, abnormal tube voltage and abnormal image contrast have a causal relationship, and tube current fluctuation and abnormal image contrast also have an indirect relationship. By calculating the frequency and intensity of the correlations between features, the correlation degree of each feature with other features is obtained. For example, the correlation degree between abnormal tube voltage and tube current fluctuation is 0.8, the correlation degree between abnormal tube voltage and abnormal image contrast is 0.9, and the correlation degree between tube current fluctuation and abnormal image contrast is 0.6.

[0142] After determining the correlation, the degree of correlation between the standard quality control features and the target quality control features was confirmed based on the correlation results and the target quality control characterization characteristics. In the chest CT case, the target quality control characterization characteristics included the causes of equipment calibration cycle delays and the correlation between tube voltage and image contrast. Matching the standard quality control features with this characterization information revealed that, for example, abnormal tube voltage was directly related to the causes of equipment calibration cycle delays (correlation strength 0.9); tube current fluctuations were indirectly related to equipment calibration cycle delays (correlation strength 0.7); and abnormal image contrast was relatively indirectly related to equipment calibration cycle delays (correlation strength 0.5).

[0143] Standard quality control features of the same priority level are ranked according to their correlation, thus establishing a quality control feature order. In the medium-priority feature group of chest CT, tube voltage abnormalities have a correlation of 0.9, ranking first; tube current fluctuations have a correlation of 0.7, ranking second; and image contrast abnormalities have a correlation of 0.5, ranking third. This ranking clarifies the quality control execution order within the same priority, prioritizing features more closely related to the root cause.

[0144] To quantify the degree of correlation, the correlation coefficient can be calculated using the following formula: Correlation Coefficient = Feature Correlation Degree × Representation Matching Degree. Here, the feature correlation degree is the numerical correlation between the standard quality control feature and other features, and the representation matching degree is the degree of matching between the standard quality control feature and the target quality control representation feature, with values ​​ranging from 0 to 1. In the chest CT case, the feature correlation degree for abnormal tube voltage is 0.85, and the representation matching degree is 0.9, resulting in a correlation coefficient of 0.85 × 0.9 = 0.765. The feature correlation degree for tube current fluctuation is 0.7, and the representation matching degree is 0.7, resulting in a correlation coefficient of 0.7 × 0.7 = 0.49. The feature correlation degree for abnormal image contrast is 0.6, and the representation matching degree is 0.5, resulting in a correlation coefficient of 0.6 × 0.5 = 0.3. The features are then sorted according to their correlation coefficients, with higher coefficients ranking higher.

[0145] Based on the quality control characteristics, quality control order, and target quality control representation characteristics, output multi-dimensional medical image quality control processing instructions for the medical images to be controlled, specifically including the following steps:

[0146] Based on the quality control characteristics and quality control order, the target quality control characterization characteristics are hierarchically sorted out, and the quality control characterization characteristics corresponding to each priority are selected.

[0147] Determine the type and impact of quality control anomalies based on the quality control characteristics and the quality control sequence to determine the processing priority of each anomaly type.

[0148] By processing the types, impact levels, and handling priorities of quality control anomalies, a multi-dimensional collaborative processing framework is obtained, which specifically includes the following steps:

[0149] Generate basic quality control processing content at the corresponding level based on the type and degree of impact of quality control anomalies;

[0150] By combining the basic quality control processing content, the correlation between different processing priorities is determined, forming a multi-dimensional collaborative processing framework;

[0151] Based on the multi-dimensional collaborative processing framework, the adjustment rules for processing content in different quality control scenarios are clarified, resulting in multi-dimensional medical image quality control processing instructions.

[0152] First, the target quality control characteristics were stratified and sorted according to the quality control sequence, and the quality control characteristics corresponding to each priority were selected. Taking the quality control scenario of chest CT examination as an example, the high-priority characteristic group includes tube voltage drop caused by X-ray tube overheating, the medium-priority characteristic group includes tube current fluctuation and abnormal image contrast, and the low-priority characteristic group includes room temperature and humidity fluctuation and patient respiratory artifacts. The target quality control characteristics were divided into this hierarchy, and the characterization information corresponding to high, medium, and low priorities was extracted respectively. For example, high priority corresponds to X-ray tube temperature exceeding the standard and automatic protection mechanism triggering of the equipment, medium priority corresponds to tube current calibration deviation and contrast imbalance, and low priority corresponds to high room humidity and unstable respiratory rhythm.

[0153] The types and severity of quality control anomalies at each level were determined, and the processing priority for each anomaly type was determined based on the quality control feature sequence. In the high-priority feature group, the anomaly type was equipment hardware failure, with a severe impact, directly leading to scan interruption or complete image failure; this had the highest processing priority. In the medium-priority feature group, the anomaly type was equipment parameter drift, with a moderate impact, which would reduce lesion identification but the image could still be used for auxiliary diagnosis; this had the second-highest processing priority. In the low-priority feature group, the anomaly type was environmental and patient-related interference, with a mild impact, only increasing image noise but not affecting the observation of core lesions; this had the moderate processing priority.

[0154] By integrating the types, impact levels, and handling priorities of quality control anomalies, a multi-dimensional collaborative handling framework is constructed. Basic quality control procedures are established for different anomaly types. For example, hardware failures require immediate shutdown for repair and replacement of the X-ray tube with a spare; equipment parameter drift requires recalibrating the tube current and voltage; and environmental and patient-related interference requires adjusting the room temperature and humidity and instructing patients on breath-holding. Based on this, the relationships between different handling priorities are determined. For example, equipment repair takes precedence over parameter calibration, and parameter calibration takes precedence over environmental adjustments. This clarifies the sequence of handling procedures, forming a multi-dimensional collaborative handling framework encompassing anomaly type, impact assessment, handling measures, and execution order.

[0155] Based on a multi-dimensional collaborative processing framework, adjustment rules for processing content under different quality control scenarios are clearly defined, generating multi-dimensional medical image quality control processing instructions. For example, in emergency scenarios, to ensure patient treatment efficiency, the processing rules for low-priority anomalies are adjusted, non-urgent steps such as adjusting environmental temperature and humidity are simplified, and high- and medium-priority equipment failures and parameter drift are prioritized. In routine physical examination scenarios, all processing steps are strictly executed according to the framework to ensure that image quality meets diagnostic standards. These adjustment rules are integrated into specific instructions. For example, the instruction for high-priority anomalies is to immediately stop the machine to repair the X-ray tube and start the backup equipment; the instruction for medium-priority anomalies is to calibrate the equipment parameters after the current patient's scan is completed; and the instruction for low-priority anomalies is to adjust the room temperature and humidity after the day's scanning tasks are completed.

[0156] To quantify the adjustment range in different scenarios, the scenario adjustment coefficient can be calculated using the following formula: Scenario Adjustment Coefficient = Scenario Urgency × (1 - Abnormal Impact Weight). The scenario urgency ranges from 0 to 1, with 0.9 for emergency scenarios and 0.5 for routine scenarios. The abnormal impact weight is the proportion of the impact of low-priority abnormalities to the total impact. For example, if the impact of low-priority abnormalities is 0.2 and the total impact is 1.0, then the weight is 0.2. Substituting these values ​​into the formula for an emergency scenario, the scenario adjustment coefficient is 0.9 × (1 - 0.2) = 0.72. This coefficient reflects the simplification of low-priority abnormality handling; a higher coefficient indicates a greater degree of simplification.

[0157] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0158] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0159] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A domain-driven medical image quality control architecture, characterized in that, include: Extraction module: Extracts the quality control feature parameters of the medical images to be quality controlled; The comparison and generation module compares the feature parameters to be controlled with the standard quality control feature range to obtain the target quality control features, and performs attribute classification and correlation judgment on the target quality control features to generate a quality control feature set; Acquisition module: Acquires the feature association parameters of the target quality control feature to which the quality control feature set belongs, as well as the target quality control feature parameters; Analysis module: Verifies and analyzes the feature correlation parameters to obtain the target quality control characterization features; analyzes the target quality control feature parameters to obtain the target quality control coefficient; Processing module: Obtains quality control priority coefficient based on target quality control coefficient, and obtains quality control feature order of standard quality control features based on quality control priority coefficient; Output module: Outputs multi-dimensional medical image quality control processing instructions for the medical images to be controlled based on the quality control characteristics, quality control order, and target quality control characterization characteristics.

2. The domain-driven medical image quality control architecture according to claim 1, characterized in that, The target quality control feature is obtained by comparing the feature parameter to be controlled with the standard quality control feature range. This includes the following steps: Extract the parameter generation time series information of the corresponding examination process of the medical image to be quality controlled, and combine the parameter generation time series information to determine the original feature parameters; Retrieve the standard quality control feature range adapted to the clinical scenario, compare the original feature parameters with the standard quality control feature range, filter out the original deviation feature parameters that deviate from the standard quality control feature range and the original qualified feature parameters that conform to the standard quality control feature range, and obtain the time-series node information of the original deviation feature parameters. Based on the time-series node information of the original deviation characteristic parameters, the subordinate characteristic parameters of the corresponding time-series node information are located. The subordinate characteristic parameters are compared with the standard quality control characteristic range, and subordinate deviation characteristic parameters that deviate from the standard quality control characteristic range and subordinate qualified characteristic parameters that conform to the standard quality control characteristic range are selected. The transmission relationship between the original deviation characteristic parameter and the subordinate deviation characteristic parameter is determined, and the transmission determination result is obtained by judging whether the subordinate deviation characteristic parameter transmits to the original deviation characteristic parameter. A temporary quality control feature set is constructed by integrating the original qualified feature parameters and the subordinate qualified feature parameters; The target quality control features are obtained by regularizing the temporary quality control feature set based on the transmission judgment results.

3. The domain-driven medical image quality control architecture according to claim 1, characterized in that, The target quality control features are divided into attributes and their correlation is determined to generate a quality control feature set. This process includes the following steps: After classifying the target quality control features by attributes to obtain compliant target quality control features and abnormal target quality control features, a preliminary screening feature set is formed. The correlation between the quality control features of compliant targets and the quality control features of abnormal targets in the initial screening feature set is determined, and the quality control features of compliant targets and the quality control features of abnormal targets that have mutual influence are marked as feature groups; Based on the degree of influence of the target quality control features on medical image diagnosis, the feature group is divided into primary features and secondary features to generate a quality control feature set.

4. The domain-driven medical image quality control architecture according to claim 1, characterized in that, The target quality control characterization features are obtained by verifying and analyzing the feature association parameters, specifically including the following steps: The feature association parameters are validated to form a valid set of association parameters. By exploring the mutual constraints among parameters in the effective set of related parameters, and clarifying the positive linkages and negative inhibitions between different parameters based on these mutual constraints, a parameter constraint relationship map is formed. Based on the parameter constraint relationship map, locate the target-related parameters that have a dominant influence on the target quality control characteristics; Tracing the location results of the causes corresponding to the associated parameters of the target; Based on the results of the cause localization and the parameter constraint relationship map, the influence of the target quality control characteristics is determined and the influence status of the characteristics is formed. The characteristics affecting the system are integrated to form the target quality control characterization features.

5. The domain-driven medical image quality control architecture according to claim 4, characterized in that, The target quality control coefficient is obtained by analyzing the target quality control characteristic parameters, specifically including the following steps: Based on the dependency and constraint attributes of the target quality control feature parameters, the strongly correlated feature parameter group and the weakly correlated feature parameter group are distinguished. The core influence parameters are obtained by processing the strongly correlated feature parameter group and the weakly correlated feature parameter group. Based on the interaction law of strongly correlated feature parameter groups, the basic influence values ​​of the core influence parameters are superimposed and adjusted to obtain the adjusted influence values. The target quality control characteristic parameters are quantified and integrated based on the adjusted influence values ​​to obtain the target quality control coefficient.

6. The domain-driven medical image quality control architecture according to claim 5, characterized in that, The core influencing parameters are obtained by processing the strongly correlated feature parameter set and the weakly correlated feature parameter set. The specific steps include: The core dominant parameters are obtained by judging the interaction law of parameters within a strongly correlated feature parameter group, the dominant direction of the quality control influence of the core dominant parameters is clarified, and the dominant parameter set of the core dominant parameters is extracted based on the dominant direction of the quality control influence. Assess the basic influence of the dominant parameter set and weakly correlated feature parameters on the quality control results of medical images, and determine the core influencing parameters based on the basic influence.

7. The domain-driven medical image quality control architecture according to claim 6, characterized in that, The quality control priority coefficient is obtained based on the target quality control coefficient, specifically including the following steps: The actual deviation of the corresponding quality control feature is defined based on the target quality control coefficient, and the deviation definition result is obtained by clarifying the extent to which each target quality control coefficient deviates from the standard quality control feature range based on the actual deviation. The assessment results of the transmission effect are obtained based on the deviation definition results. The associated clinical diagnosis and treatment risk level is determined based on the results of the deviation transmission impact assessment. The priority coefficient for quality control is determined based on the results of the associated clinical diagnosis and treatment risk levels.

8. The domain-driven medical image quality control architecture according to claim 1, characterized in that, The quality control priority order of standard quality control features is obtained based on the quality control priority coefficient, which specifically includes the following steps: Several hierarchical intervals are determined based on the numerical distribution of the quality control priority coefficient, and each hierarchical interval corresponds to a unique priority level; The correlation results are obtained by determining the correlation degree of standard quality control features with priority levels. The correlation between the standard quality control characteristics and the target quality control characteristics is confirmed based on the correlation results and the target quality control characterization characteristics. The quality control features are sorted according to their priority levels based on their degree of correlation to form a quality control feature quality control order.

9. The domain-driven medical image quality control architecture according to claim 8, characterized in that, Based on the quality control characteristics, quality control order, and target quality control representation characteristics, output multi-dimensional medical image quality control processing instructions for the medical images to be controlled, specifically including the following steps: Based on the quality control characteristics and quality control order, the target quality control characterization characteristics are hierarchically sorted out, and the quality control characterization characteristics corresponding to each priority are selected. Determine the type and impact of quality control anomalies based on the quality control characteristics and the quality control sequence to determine the processing priority of each anomaly type. A multi-dimensional collaborative processing framework is obtained by processing the types, impact levels, and processing priorities of quality control anomalies. Based on the multi-dimensional collaborative processing framework, the adjustment rules for processing content in different quality control scenarios are clarified, resulting in multi-dimensional medical image quality control processing instructions.

10. The domain-driven medical image quality control architecture according to claim 9, characterized in that, By processing the types, impact levels, and handling priorities of quality control anomalies, a multi-dimensional collaborative processing framework is obtained, which specifically includes the following steps: Generate basic quality control processing content at the corresponding level based on the type and degree of impact of quality control anomalies; By combining the basic quality control processing content to determine the correlation between different processing priorities, a multi-dimensional collaborative processing framework is formed.

Images