Anesthesia ultrasound teaching evaluation system based on ability portrait

Abstract

The invention discloses an anesthesia ultrasound teaching evaluation system based on ability portraits, and relates to the technical field of medical education and medical information. According to the invention, the configurable capability dimension is pre-defined through the configuration module, and the scenarized weight is matched. The acquisition module integrates the ultrasonic image data, the anatomical structure recognition result and the synchronous operation behavior data. And the calculation module converts the multi-source data into quantitative parameters of multiple capability dimensions and combines the quantitative parameters to generate a target capability portrait which represents the comprehensive capability state of the learner and contains risk severity level marks. And the mapping module matches the ability portrait with an ability level model to obtain a target ability level of the learner. And the evaluation module outputs an evaluation report including the target capability portrait, the target capability level, the short plate dimension and the change trend, so that dynamic description of the capability can be realized, support is provided for personalized teaching, and the pertinence and the accuracy of anesthesia ultrasound teaching are improved.

Description

Technical Field

[0001] This application relates to the fields of medical education and medical information technology, specifically to an anesthesia ultrasound teaching assessment system based on competency profiles. Background Technology

[0002] Anesthesia ultrasound (including ultrasound-guided nerve blocks, vascular puncture, airway and gastric contents assessment, and spinal anesthesia-related ultrasound) places integrated demands on learners of "image recognition, spatial understanding, operational coordination, and safety decision-making." Learners need to identify key anatomical structures in two-dimensional dynamic ultrasound images and simultaneously plan the path and avoid risks while maintaining probe stability and coordinating needle insertion. With the development of ultrasound technology and perioperative ultrasound training, how to objectively, quantifiably, and reproducibly assess learners' multidimensional abilities has become a key issue in anesthesia ultrasound teaching and standardized residency assessment.

[0003] The assessment methods for teaching anesthesia ultrasound can be mainly divided into three categories.

[0004] The first type is subjective evaluation by teachers and a one-time assessment. This method involves the supervising teacher scoring based on experience at the bedside or in simulated training, or judging "pass / fail" through a one-time OSCE / graduation assessment. The advantage of this method is its low implementation cost and ease of integration with clinical work. However, this method is highly subjective, difficult to reproduce among different teachers and institutions, and difficult to break down into specific competency dimensions (e.g., it can only give a total score without explaining "where the mistakes / weaknesses were").

[0005] The second category is objective structured assessment based on checklists / scales. This method uses checklists and global rating scales to structurally score skills such as ultrasound-guided regional anesthesia, with some scales validated using the Delphi method or for reliability and validity. However, it is still limited to relying on observers' on-site scoring; it does not adequately characterize continuous data during the learning process (fluctuations in image quality, stability of probe trajectory, distribution of error types, and changes in ability over time), making it difficult to form a "long-term ability growth trajectory".

[0006] The third method is training records from ultrasound simulators / digital platforms. This method uses high-fidelity simulators or digital teaching platforms for practice, and the platform can record some training results (such as number of completions, accuracy, and duration). However, the recorded indicators are often "outcome-oriented" (number of completions / duration / level completion), making it difficult to form an interpretable multidimensional ability model; data from different training tasks cannot be unified into the same evaluation framework, making it difficult to use the evaluation results for personalized teaching path recommendations.

[0007] Therefore, traditional anesthesia teaching assessments are highly subjective, lack sufficient quantification, are difficult to reproduce consistently across teachers and institutions, and fail to reflect learners' differences in different ability dimensions (strong recognition ability but unstable operation, or stable operation but weak spatial understanding, etc.). Furthermore, they lack a dynamic portrayal of learners' abilities changing over training time, failing to create a "growth curve / ability trajectory," and thus the assessment results are difficult to directly drive teaching optimization. Summary of the Invention

[0008] The purpose of this application is to provide an anesthesia ultrasound teaching assessment system based on competency profiles to solve the technical problem that traditional anesthesia ultrasound teaching assessments cannot support precise teaching.

[0009] To achieve the above objectives, this application provides an anesthesia ultrasound teaching assessment system based on competency profiles, comprising: A configuration module is used to predefine multiple configurable capability dimensions for anesthesia ultrasound teaching, and the dimension weight of each capability dimension is determined based on the target anesthesia ultrasound teaching scenario; The acquisition module is used to acquire teaching data of learners performing anesthesia ultrasound teaching tasks in the target anesthesia ultrasound teaching scenario. The teaching data includes at least ultrasound image data of the target anatomical structure, anatomical structure recognition results, and operational behavior data synchronized with the ultrasound image data. The calculation module is used to calculate the quantitative parameters of the learner in each of the ability dimensions based on the teaching data; The profile generation module is used to combine the quantitative parameters of multiple ability dimensions to generate a target ability profile that represents the learner's comprehensive ability status, and the target ability profile includes a risk severity level label. The mapping module is used to match the ability profile with a preset ability level model to obtain the target ability level corresponding to the learner. The assessment module is used to output an assessment report that includes the target capability profile, the target capability level, the weakness dimension, and the trend of change.

[0010] The beneficial effects of this application are: This application addresses the issues of unclear dimensions and inconsistent evaluation standards by pre-defining configurable capability dimensions and matching them with scenario-based weights through a configuration module, making the assessment more aligned with teaching needs. The acquisition module integrates ultrasound imaging data, anatomical structure recognition results, and synchronous operational behavior data, providing objective evidence for the assessment and reducing subjective bias. The calculation module transforms multi-source data into quantitative parameters for multiple capability dimensions and combines them to generate a target capability profile representing the learner's comprehensive capability status, including risk severity level annotations, clearly revealing differences in the learner's capability structure and risk points. The mapping module matches the capability profile with a capability level model to obtain the learner's target capability level. The assessment module outputs an assessment report including the target capability profile, the target capability level, the weakness dimension, and the trend of change, enabling dynamic capability characterization, supporting personalized teaching, and thus improving the relevance and accuracy of anesthesia ultrasound teaching.

[0011] Other features and advantages of this application will be described in detail in the following detailed description section. Attached Figure Description

[0012] Figure 1 This is a schematic diagram illustrating an application scenario of an anesthesia ultrasound teaching assessment system based on competency profiling provided in this application embodiment; Figure 2 This is a schematic diagram of the structure of an anesthesia ultrasound teaching assessment system based on competency profiling provided in an embodiment of this application.

[0013] Explanation of reference numerals in the attached figures 1. Controller; 2. Ultrasound image acquisition device; 3. Human-computer interaction terminal; 4. Teaching database; 200. Anesthesia ultrasound teaching assessment system based on competency profile; 201. Configuration module; 202. Acquisition module; 203. Calculation module; 204. Profile generation module; 205. Mapping module; 206. Assessment module. Detailed Implementation

[0014] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0015] In the description of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "a plurality of" means two or more, unless otherwise explicitly specified. Details are set forth in the following description for illustrative purposes. It should be understood that those skilled in the art will recognize that this application can be implemented without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid unnecessarily obscuring the description of this application. Therefore, this application is not intended to be limited to the embodiments shown, but rather to be consistent with the broadest scope of the principles and features disclosed herein.

[0016] Figure 1 This is a schematic diagram illustrating an application scenario of an anesthesia ultrasound teaching assessment system based on competency profiling, as provided in this embodiment. The application scenario may include a controller 1, an ultrasound image acquisition device 2, a human-computer interaction terminal 3, and a teaching database 4. The controller 1 communicates with the ultrasound image acquisition device 2, the human-computer interaction terminal 3, and the teaching database 4, respectively.

[0017] Controller 1 integrates an anesthesia ultrasound teaching assessment system based on competency profiles. This system may include a memory and a processor. The memory is configured to store instructions and data, and the processor is configured to retrieve instructions and data from the memory and execute steps of clinical competency-based anesthesia ultrasound teaching when running instructions. Controller 1 is the carrier that integrates end-to-end data processing and teaching process control, realizing a closed-loop process of multimodal data quantification, learner competency assessment, and level unlocking or restriction.

[0018] The ultrasound image acquisition device 2 serves as the source of multimodal operational data for the learner and may include an ultrasound probe (with a built-in nine-axis inertial measurement unit (IMU) or optical markers), a video acquisition card, etc. The ultrasound probe can support simulated clinical anesthesia ultrasound operations, such as nerve blocks and vascular puncture scans. The IMU sensor synchronously acquires the probe's six degrees of freedom motion data, and the video acquisition card can capture high-resolution ultrasound video streams in real time.

[0019] The human-computer interaction terminal 3 serves as a platform for operation and feedback interaction. It can be a visual interactive device for learners or a data retrieval device for administrators. The human-computer interaction terminal 3 may include a high-definition touchscreen display supporting multi-touch and integrating voice broadcasting and operation guidance functions. In one example, the human-computer interaction terminal 3 may support ultrasound parameter adjustment buttons, software-integrated teaching interactive interface, real-time display of ultrasound images, progress of levels, operation prompts, and achievement status, while simultaneously receiving learner operation commands to achieve real-time interaction between operation and feedback. In another example, the human-computer interaction terminal 3 may include a learning terminal for learners and a teaching management terminal for teachers. The learning terminal can display individual learner feedback data, while the teaching management terminal can display feedback data from a batch of learners.

[0020] Teaching Database 4 serves as the carrier for level resources and data storage. It employs a database server architecture and contains multiple sets of anesthesia ultrasound teaching level configuration trees, such as nerve block and vascular puncture series levels. The content stored in Teaching Database 4 can include standard ultrasound cross-sectional images, anatomical structure feature templates, progressive level teaching content, judgment thresholds for each level, learner historical operation data, and standard operation demonstration videos, forming the basis for initialization and access to teaching resources.

[0021] The ultrasound image acquisition device 2 serves as the data input terminal, and the teaching database 4 serves as the resource storage terminal. Both directly establish one-way / two-way data links with the controller 1. The human-computer interaction terminal 3 serves as the interactive feedback segment, forming a two-way interaction with the controller 1 for command issuance and data upload. The whole system constitutes a closed-loop architecture of acquisition, processing, storage, and interaction to ensure a smooth and efficient teaching process.

[0022] In one example, the ultrasound image acquisition device 2 is physically connected to the data input port of the controller 1 via a high-speed data transmission interface (such as a Universal Serial Bus (USB) 3.0, Thunderbolt interface, or wireless Wi-Fi module). The display screen of the human-machine interface terminal 3 is connected to the host computer where the controller 1 is located via a video output interface (High Definition Multimedia Interface (HDMI) / DisplayPort (DP), and the input devices of the human-machine interface terminal 3 are connected to the host computer via I / O interfaces. The ultrasound probe of the ultrasound image acquisition device 2 has a position sensor integrated inside or on its surface, such as a nine-axis IMU or optical markers, for synchronously acquiring the spatial motion data of the ultrasound probe.

[0023] Understandable Figure 1The electronic devices in the application scenario of the competency-based anesthesia ultrasound teaching assessment system shown do not constitute a limitation on the embodiments of this application. That is, the number and types of devices included in the application scenario of the competency-based anesthesia ultrasound teaching assessment system, or the number and types of devices included in each electronic device, do not affect the overall implementation of the technical solution in the embodiments of this application, and can all be considered as equivalent substitutions or derivatives of the technical solutions claimed in the embodiments of this application.

[0024] In this application embodiment, controller 1 can be an independent device, or a device network or device cluster composed of devices. For example, controller 1 described in this application embodiment includes, but is not limited to, a computer, a network host, a single network device, a set of multiple network devices, or a cloud device composed of multiple devices. Among them, the cloud device is composed of a large number of computers or network devices based on cloud computing.

[0025] Those skilled in the art will understand that Figure 1 The application scenarios shown are merely one application scenario corresponding to the technical solution of this application, and do not constitute a limitation on the application scenarios of the technical solution of this application. Other application scenarios may include more than one application scenario. Figure 1 The number of more or fewer electronic devices shown, or the network connections of electronic devices, for example Figure 1 Only one electronic device is shown in the image. It is understood that the scenario of this competency-based anesthesia ultrasound teaching assessment system may also include one or more other electronic devices, which are not specified here.

[0026] It should be noted that, Figure 1 The application scenario of the anesthesia ultrasound teaching assessment system based on competency profile shown is merely an example. The application scenario of the anesthesia ultrasound teaching assessment system based on competency profile described in this application is to more clearly illustrate the technical solution of this application embodiment and does not constitute a limitation on the technical solution provided in this application embodiment.

[0027] Based on the application scenarios of the competency-based anesthesia ultrasound teaching assessment system described above, an embodiment of the competency-based anesthesia ultrasound teaching assessment system is proposed. The modules and units in this embodiment can communicate with each other. A detailed description is provided below with reference to the accompanying drawings.

[0028] Figure 2 This is a schematic diagram of the structure of an anesthesia ultrasound teaching assessment system 200 based on competency profiling provided in an embodiment of this application. Figure 2 As shown, the anesthesia ultrasound teaching assessment system 200 based on competency profiles may include a configuration module 201, an acquisition module 202, a calculation module 203, a profile generation module 204, a mapping module 205, and an assessment module 206.

[0029] The configuration module 201 is used to predefine multiple configurable capability dimensions for anesthesia ultrasound teaching. Capability dimensions break down anesthesia ultrasound skills into quantifiable and interpretable specific competency directions, serving as core indicator categories for assessment. For example, at least four basic dimensions can be defined based on the capability requirements of anesthesia ultrasound impact recognition, spatial understanding, operational coordination, and safety decision-making, with support for the expansion of capability dimensions. Each capability dimension is configured with a dimension weight, which is assigned to each capability dimension based on teaching needs analysis for the target anesthesia ultrasound teaching scenario. The dimension weight of each dimension is determined based on the target anesthesia ultrasound teaching scenario, reflecting the importance of each capability dimension in a specific teaching scenario. This weighted calculation of comprehensive capability reduces the likelihood of a one-size-fits-all assessment.

[0030] The data acquisition module 202 is used to collect teaching data from learners performing anesthesia ultrasound teaching tasks in the target anesthesia ultrasound teaching scenario. The teaching data can include at least ultrasound image data of the target anatomical structure, anatomical structure recognition results, and operational behavior data synchronized with the ultrasound image data. Ultrasound image data refers to dynamic or static ultrasound images generated during anesthesia ultrasound teaching and is the basis for structure recognition. Anatomical structure recognition results are the recognition results and confidence levels of the target structure in the ultrasound image data through an algorithmic model. Synchronized operational behavior data refers to the learner's operational actions synchronized with the ultrasound image acquisition and can directly reflect operational proficiency. The collaborative acquisition of multiple data types can provide data support for multi-dimensional ability assessment.

[0031] The calculation module 203 is used to calculate the quantitative parameters of learners in each ability dimension based on teaching data. Quantitative parameters transform raw teaching data into calculable and comparable numerical indicators, forming the core of the ability profile. For example, it can calculate parameters related to recognition ability, operational stability, spatial comprehension, and operational consistency. This achieves the transformation of objective data into ability measurements, with clear calculation logic for each dimension, covering core points and providing data support for accurate profiling.

[0032] The profile generation module 204 combines quantitative parameters from multiple ability dimensions to generate a target ability profile representing the learner's comprehensive ability status. The target ability profile is a visualized / structured result representing the learner's comprehensive ability status, containing multi-dimensional quantitative parameters and risk annotations, which may include risk severity level annotations. Risk severity level annotations are levels based on the degree of risk of operational errors (e.g., misidentifying an artery as a vein is high-risk), used for safety teaching prompts. The quantitative parameters of each ability dimension are integrated according to a preset format, and then risk severity levels are annotated based on error event logs. The integrated parameters are transformed into visualizations such as radar charts or distribution maps to form the target ability profile, and the profile is automatically updated after each training session, forming an ability trajectory. The ability trajectory is a sequence of ability profiles, reflecting the learner's ability trends over time. In this way, the learner's strengths and weaknesses can be intuitively revealed, and risk annotations can strengthen the safety teaching orientation and provide early warnings of operational risks.

[0033] The mapping module 205 matches the ability profile with a pre-defined ability level model to obtain the learner's target ability level. The ability level model is a rule- or data-driven model that maps continuous quantified parameters to discrete levels, such as rule-based thresholding and machine learning clustering. The target ability level is a graded evaluation of the learner's comprehensive abilities, facilitating rapid assessment of ability levels. Through a pre-configured ability level model, the quantified parameters of the target ability profile can be compared with the boundary of the level model to determine the target ability level. Simultaneously, the level boundary can be optimized based on new training data to improve matching accuracy. Converting the ability profile into a target ability level facilitates teaching management and allows learners to quickly recognize their ability levels. The model iteration mechanism improves assessment adaptability, adapting to the ability characteristics of different groups, providing a clear basis for tiered instruction, and implementing graded teaching.

[0034] The assessment module 206 outputs an assessment report containing a target competency profile, target competency level, weakness dimensions, and trends. Weakness dimensions refer to the learner's competency parameters that are below preset standard values, representing a key focus of personalized instruction. Trends reflect how the learner's competency changes with training frequency and duration, indicating teaching effectiveness. A standardized assessment report is generated by integrating the target competency profile (including risk labeling), target competency level, weakness dimensions (quantitative parameters below standard values), and competency trends (compared to historical profiles). Based on the comparison of quantitative parameters with standard values, the weaker dimensions requiring reinforcement are clearly identified. By comparing parameters in historical and current profiles, the report analyzes competency improvement / decline trends and indicates the rate of growth. The comprehensive report content provides precise targets for instructional optimization; weakness identification and trend analysis support personalized instruction and address the challenge of traditional assessments failing to drive instructional optimization. The standardized report format facilitates cross-institutional or cross-batch comparisons of teaching quality, supporting standardized training assessments and teaching management.

[0035] This application's embodiments address the issues of unclear dimensions and inconsistent evaluation standards by pre-defining configurable capability dimensions and matching them with scenario-based weights through a configuration module, making the evaluation more aligned with teaching needs. The acquisition module integrates ultrasound image data, anatomical structure recognition results, and synchronous operational behavior data, providing objective evidence for the evaluation and reducing subjective bias. The calculation module transforms multi-source data into quantitative parameters for multiple capability dimensions and combines them to generate a target capability profile representing the learner's comprehensive capability status, including risk severity level annotations, clearly revealing differences in the learner's capability structure and risk points. The mapping module matches the capability profile with a capability level model to obtain the learner's target capability level. The evaluation module outputs an evaluation report including the target capability profile, the target capability level, the weakness dimension, and the trend of change, enabling dynamic capability characterization, supporting personalized teaching, and thus improving the relevance and accuracy of anesthesia ultrasound teaching.

[0036] In this embodiment of the application, the configuration module 201 may include an identification unit, a calling unit, an adjustment unit, and a disabling unit.

[0037] The identification unit is used to identify the target anesthesia ultrasound teaching scenario corresponding to the current anesthesia ultrasound teaching task. The target anesthesia ultrasound teaching scenario refers to an anesthesia ultrasound application scenario directly corresponding to the current teaching task, with specific training objectives and operational requirements. It includes at least one of the following: ultrasound-guided nerve block scenario, vascular puncture scenario, airway assessment scenario, and spinal canal localization scenario. In one example, the core information of the current anesthesia ultrasound teaching task can be obtained, including the task label, training objectives, involved anatomical structures, and risk points. Then, the extracted task features are compared with a pre-set scenario feature library to match the corresponding target anesthesia ultrasound teaching scenario, generating a scenario identifier, which is synchronized to the calling unit, adjustment unit, and disabling unit as the basis for subsequent dimension configuration. The scenario feature library is a collection of core features of various teaching scenarios pre-stored by the system. It can contain key information such as the task type, anatomical structure, training focus, and risk points of each target anesthesia ultrasound teaching scenario for scenario matching and determination. The scenario identifier is a code or name used to uniquely identify the target teaching scenario, facilitating the synchronization of scenario information among various units of the system.

[0038] The invocation unit is used to match the capability dimensions that need to be activated in the target anesthesia ultrasound teaching scenario. Based on the scenario identifier output by the recognition unit, the system's preset scenario-dimensional association library can be queried to match the core capability dimensions that need to be emphasized in that scenario (e.g., the nerve block scenario is associated with the dimensions of "anatomical structure recognition, operational stability, and spatial understanding," while the vascular puncture scenario is additionally associated with the extended dimension of "risk awareness"). Then, the matched capability dimensions are sorted according to the scenario's training focus (e.g., the airway assessment scenario prioritizes "operational stability and anatomical structure recognition"). The list of matched capability dimensions and their priorities are output to the adjustment unit, triggering the dimension weight adjustment process to ensure that core dimensions are configured first. This enables scenario-based and precise invocation of capability dimensions, reducing dimension redundancy or missing dimensions. Matching based on the preset association library ensures the professionalism and consistency of dimension invocation.

[0039] The adjustment unit dynamically adjusts the weight allocation ratio of each activated capability dimension based on the target anesthesia ultrasound teaching scenario. For example, it retrieves the initial weight allocation scheme corresponding to the target scenario from the system's weight configuration library (e.g., initial weights for the nerve block scenario: anatomical structure recognition 40%, operational stability 30%, spatial understanding 20%, operational consistency 10%). The weight configuration library is a collection of pre-stored basic weight allocation schemes for various scenarios, which can be constructed based on the Delphi method, teaching practice data, and expert review, providing a benchmark for dynamic adjustment. Then, combining the scenario training focus, historical teaching data, and expert feedback, the weight ratio of each dimension is adjusted (e.g., for vascular puncture scenarios for novice learners, the weight of "anatomical structure recognition" is increased to 45%, and the weight of "operational consistency" is decreased to 5%; for experienced learners, the weight of "operational consistency" is increased to 15%). The adjusted weight scheme is verified against the scenario training objectives to ensure that the weight allocation matches the scenario focus (e.g., the weight of "operational stability" in the airway assessment scenario is not less than 35%). After confirmation, the final weight configuration is output. Through scenario-based and personalized dynamic adjustment of weights, a one-size-fits-all weight configuration can be reduced. Adjusting the weights based on learners' levels and training stages can improve the accuracy and fairness of the assessment.

[0040] The disabled unit is used to determine disabled capability dimensions based on the target anesthesia ultrasound teaching scenario. In one example, the necessity of matched capability dimensions can be analyzed based on the training objectives and operational requirements of the target scenario. Unnecessary capability dimensions refer to those with low relevance to the current scenario's training objectives, minimal impact on assessment results, or those that do not require assessment due to the learner's stage / scenario characteristics. For example, in the basic stage airway assessment scenario, the "operational consistency" dimension is deemed unnecessary because the learner has not yet formed stable operating habits; similarly, in the spinal canal positioning scenario, the "risk awareness" dimension is deemed unnecessary if it does not involve misjudging high-risk structures. Then, unnecessary capability dimensions are marked to generate a disabled dimension list (e.g., marking operational consistency as a disabled dimension in the basic stage airway assessment scenario). The disabled dimension list is output to the system to ensure that disabled dimensions do not participate in subsequent parameter calculations and profile generation, while recording the reasons for disabling (e.g., "Basic training stage, no need to assess operational consistency"), facilitating teaching review and dimension configuration optimization, and improving system traceability.

[0041] In this embodiment of the application, the acquisition module 202 may include a first acquisition unit and a second acquisition unit.

[0042] The first acquisition unit is used to acquire real-time ultrasound image streams from learners or retrieve pre-stored clinically labeled pathological images to obtain ultrasound image data. Ultrasound image acquisition equipment is hardware used to capture images relevant to anesthesia ultrasound teaching, including clinical ultrasound instruments and high-fidelity ultrasound simulators. Ultrasound image data can be acquired in real-time by the ultrasound image acquisition equipment to capture dynamic ultrasound image streams during the learner's operation, ensuring that the image frame rate meets the structural recognition requirements. Alternatively, ultrasound image data can also be retrieved from a clinically labeled pathological image library (such as images of typical cases of nerve blocks or abnormal anatomical images of vascular punctures) according to the needs of the target anesthesia ultrasound teaching scenario, ensuring that the images contain complete annotation information (target structure location, clinical diagnostic conclusion).

[0043] A synchronously triggered target detection algorithm identifies target anatomical structures in ultrasound image data, yielding anatomical structure recognition results. This algorithm transforms ultrasound image data into structured recognition results. It can process real-time image streams or pre-stored images frame-by-frame or in batches. The algorithm automatically locates target anatomical structures (such as the brachial plexus, veins, and pleura) and outputs the recognition results.

[0044] The anatomical structure identification results can include at least the category of the target anatomical structure, its confidence level, and its boundary overlap with the baseline anatomical atlas. The baseline anatomical atlas is a pre-defined standardized anatomical structure atlas (such as the human ultrasound anatomical standard atlas) that serves as a reference benchmark for measuring the accuracy of identification. Boundary overlap is the proportion of overlap between the structural region identified by the target detection algorithm and the corresponding structural region in the baseline atlas, ranging from 0 to 1, with higher values ​​indicating more accurate identification.

[0045] The second acquisition unit is used to synchronously acquire operational behavior data aligned with the timestamps of the ultrasound image data through the force sensor, electromagnetic tracking module, and inertial measurement unit built into the ultrasound probe. The operational behavior data includes at least three-dimensional coordinates, three-dimensional attitude angles, contact pressure values, scan trajectory coordinate sequences, and the time interval of continuous stable images. In one example, the electromagnetic tracking module can acquire the probe's X / Y / Z three-dimensional coordinates in space in real time, recording the probe's movement trajectory to obtain the three-dimensional coordinates. The inertial measurement unit acquires the probe's pitch, roll, and yaw angles to reflect the probe's tilt and rotation states. The force sensor acquires the contact pressure between the probe and the simulated tissue / patient surface, recording the pressure change curve. The three-dimensional coordinate data are concatenated according to the timestamps to form a complete probe scan trajectory. Combined with the image clarity detection results, continuous time segments where the ultrasound probe remains stable and the image has no significant jitter are marked, thus obtaining the continuous stable image time interval. The timestamps of the ultrasound image data are extracted, and the acquired operational behavior data are precisely aligned according to the same timestamp to ensure that each frame of image corresponds to unique operational status data. Multi-sensor collaborative acquisition comprehensively captures key information such as the position, attitude, and force of the probe operation, completely reconstructing the operation process. Precise alignment of data and image timestamps supports the correlation analysis between operational actions and image results, ensuring rigorous evaluation logic for capability dimensions such as operational stability and spatial understanding. Refined acquisition parameters (such as continuous stable image intervals) directly serve the calculation of capability parameters, enhancing the quantification of operational capability assessments.

[0046] In this embodiment of the application, the acquisition module 202 may further include a third acquisition unit, a fourth acquisition unit, and a fifth acquisition unit.

[0047] The third data collection unit is used to collect learners' identity identifiers and link them to their historical training data. The identity identifier is a unique information carrier that distinguishes learners; it can be an account, number, biometrics, etc., and serves as the core index for individual training data. Historical training data comprises all data and evaluation results generated from the learner's past participation in anesthesia ultrasound teaching and training, forming the basis for analyzing their skill development trajectory. The system database can be queried based on the identity identifier to retrieve the learner's historical training data (including past ultrasound images, operational behaviors, skill profiles, and evaluation reports), establishing a binding relationship between the identity identifier and current training data, forming a personalized training data file, and ensuring data continuity.

[0048] The fourth acquisition unit is used to synchronously record task information when anesthesia ultrasound teaching tasks are initiated. Task information is a set of information describing the core attributes of anesthesia ultrasound teaching tasks and is an important basis for scenario adaptation, risk assessment, and capability dimension configuration. Task information can include at least the target anesthesia ultrasound teaching scenario corresponding to the task, a list of target anatomical structures, and a preset risk severity level. For example, when anesthesia ultrasound teaching tasks are initiated (after the learner confirms the start of training), the core task information is automatically recorded; the target anesthesia ultrasound teaching scenario (e.g., "ultrasound-guided central venous puncture") and the list of target anatomical structures (e.g., "internal jugular vein, carotid artery, pleura") are clearly recorded; and the preset risk severity level standards for the task are retrieved (e.g., "artery misdiagnosed as vein" is high risk, "slight probe jitter" is low risk). The task information is stored in a preset format (e.g., JSON format) and associated with the current training data to ensure that each batch of data includes task background information. Task information backtracking is supported, facilitating verification of training objectives and risk standards during subsequent teaching reviews. Synchronous recording of task background information provides a clear basis for subsequent scenario-based assessments (e.g., dimension weight adaptation, risk determination). Standardized storage of task information facilitates data classification, statistics, and batch analysis (such as summarizing training results by scenario). Pre-defined risk level standards ensure consistency in risk assessment and reduce subjective judgment bias.

[0049] The fifth acquisition unit is used to monitor learners' operational behaviors and recognition results in real time during training, and to mark the type of erroneous operation when operations or recognition results that do not conform to clinical standards are detected. Clinical standard operating procedures (SOPs) are based on the operating guidelines and teaching syllabus for anesthesia ultrasound clinical practice, clearly defining the boundaries of correct operation. The unit compares learners' operational behavior data with clinical SOPs (such as probe movement range and contact pressure thresholds) in real time, and compares anatomical structure recognition results with the target structure list. Monitoring trigger conditions are defined; when operational behavior exceeds the standard range, or recognition results are incorrect or have low confidence, error marking is triggered. Then, an error log is simultaneously recorded, consisting of the ultrasound image frame location where the erroneous operation occurred, the corresponding operational behavior data, and a timestamp. The error log is structured data recording learners' violations or recognition errors during training, including error type, occurrence scenario, and related data; it is the core data source for risk assessment. The error log can be precisely correlated with the ultrasound image data and operational behavior data according to timestamps, ensuring traceability of error events. Real-time monitoring of erroneous behaviors and recognition results allows for timely identification of skill gaps and safety risks, providing direct evidence for risk labeling. Error logs contain multi-dimensional correlated data, facilitating instructors' review of error causes and providing precise direction for personalized guidance. The error data is linked to real-time training data, enriching the dimensions of capability assessment, improving the interpretability of assessment results, and supporting subsequent compensatory training recommendations.

[0050] In this embodiment of the application, the calculation module 203 may include an association unit, an adaptation unit, and a calculation unit.

[0051] The association unit is used to establish the mapping relationship between each capability dimension and the reference data source. The reference data source refers to various types of raw data supporting the capability dimension assessment, including ultrasound image data, anatomical structure recognition results, operational behavior data, error logs, and baseline template data. Specifically, the assessment logic of each core capability dimension is analyzed one by one (e.g., anatomical structure recognition capability needs to be based on image recognition results, and operational stability needs to be based on probe operation data). Then, a mapping relationship table is established. For example, anatomical structure recognition capability dimension ↔ ultrasound image data, anatomical structure recognition results; operational stability capability dimension ↔ operational behavior data (position / angle / pressure), image clarity data; spatial understanding capability dimension ↔ scan trajectory data, baseline path template, and relative structural position data; operational consistency capability dimension ↔ multi-round training data / cross-task transfer data. In addition, priorities can be configured for the mapping relationships to ensure the effectiveness of data retrieval. Based on the teaching scenario adaptation results (e.g., adding a risk awareness dimension to the vascular puncture scenario), the completeness of the mapping relationship is verified, and the mapping between the new dimensions and their corresponding data sources is supplemented (e.g., risk awareness dimension ↔ error logs, high-risk structure recognition results). Clearly define the data dependencies for each capability dimension to reduce data retrieval chaos and ensure the logical rigor of quantitative calculations. Standardize mapping relationships to provide a clear basis for algorithm adaptation in subsequent calculation units and reduce data redundancy. Simultaneously, it supports flexible expansion of dimensions and data sources, improving the system's adaptability to changes in teaching scenarios.

[0052] The adaptation unit is used to set clinical qualification thresholds for the calculated indicators of each capability dimension based on clinical operation guidelines for anesthesia ultrasound, clinical big data, and expert advice. Calculated indicators are specific parameters used to quantify a capability dimension (such as average confidence level, positional variation, etc.), and are the core decomposition items of the capability dimension. The clinical qualification threshold is a critical value for the calculated indicator set based on clinical standards and teaching experience, used to determine whether the indicator has reached the qualification level. Statistical analysis is performed on the calculated indicators for each capability dimension based on a reference dataset. Then, clinical qualification thresholds are set, and the thresholds are adjusted according to the differences in teaching scenarios to form a scenario-based threshold library. The scenario-based threshold library is a set of qualification thresholds categorized and stored according to different teaching scenarios (such as novice / expert, nerve block / vascular puncture), supporting flexible threshold retrieval.

[0053] The calculation unit is used to calculate the initial results for each capability dimension based on a preset algorithm, and to calibrate and correct the initial results to obtain the quantitative parameters for each capability dimension. First, based on the data source determined by the association unit and the preset algorithm (such as statistical analysis or path matching algorithm), the raw values ​​of the corresponding indicators for each capability dimension (such as the raw value of the average confidence score and the raw value of the location change) are calculated. Then, combined with the clinical qualification threshold set by the adaptation unit, the raw values ​​are standardized (e.g., converting the raw values ​​to a [0-100] score scale) to eliminate dimensional differences between different indicators. Referring to the associated data in the error log (such as the impact of high-risk errors on capabilities), the calibrated results are corrected (e.g., the anatomical structure recognition capability score is reduced by 10% due to a high-risk error of "artery misidentified as vein"). The corrected scores of each indicator are integrated and weighted according to the dimension weights to form the final quantitative parameters for each capability dimension (e.g., the anatomical structure recognition capability quantitative parameter = 85 points).

[0054] In this embodiment, the capability dimension may include an anatomical structure recognition capability dimension, operational stability capability dimension, spatial understanding capability dimension, and operational consistency capability dimension. The calculation module 203 can be used for the following steps.

[0055] For the anatomical structure recognition capability dimension, based on the recognition results of target anatomical structures from ultrasound image data, an effective time window is defined during the training process. The effective time window represents the operational phase directly related to capability assessment during training; invalid time segments are eliminated to ensure computational accuracy. A confidence sequence of the target anatomical structure within the effective time window is extracted. This confidence sequence is a set of target structure recognition confidence scores arranged chronologically, reflecting the dynamic changes in recognition accuracy. Within the effective time window, the recognition confidence of the target anatomical structure in each frame can be extracted by timestamp, forming a confidence sequence. The average recognition confidence (sum of sequences / number of frames) and the confidence fluctuation variance (reflecting confidence stability) of the sequence are calculated. The number of misidentifications within the effective time window (the number of frames where the identified category is inconsistent with the target structure) is counted, and the misidentification rate (number of misidentifications / number of effective frames) is calculated. The average confidence, confidence fluctuation variance, and misidentification rate are weighted and calculated to form the first quantitative parameter of the anatomical structure recognition capability dimension.

[0056] For the operational stability capability dimension, the position and angle time-series data of the ultrasound probe can be extracted from the operational behavior data and the data can be segmented by unit time. The position change and angle change per unit time are calculated, and combined with the clarity assessment results of the ultrasound image data, the time required to achieve stable images and the ratio of the duration of stable images to the total training time are determined as the second quantitative parameter of the operational stability capability dimension. Stable images are ultrasound image data with a clarity greater than the set clarity.

[0057] For the spatial understanding dimension, preset anesthesia ultrasound baseline scanning paths (such as the standard scanning trajectory for supraclavicular brachial plexus block) and corresponding baseline structural relative position templates (such as the standard relative coordinates of the brachial plexus and subclavian artery) are invoked. The learner's scanning path data is compared with the baseline scanning path to calculate the path matching degree. Simultaneously, the actual relative position of the target structure in the ultrasound image is analyzed, and the relative position deviation value is obtained by comparing it with the baseline template. The consistency of the structural localization results in dynamic scenarios (such as during probe movement) is statistically analyzed. The relative position deviation value of the target anatomical structure in the ultrasound image data is analyzed. Combined with the structural localization results in dynamic scenarios, a localization consistency index is statistically calculated to form the third quantitative parameter of the spatial understanding dimension.

[0058] For the operational consistency ability dimension, based on learners' multiple training data from the same anesthesia ultrasound teaching task or transfer training data from different anesthesia ultrasound teaching tasks, the mean and variance of the anatomical structure recognition ability dimension, operational stability ability dimension, and spatial understanding ability parameters are calculated. For example, multiple training data of learners on the same task (e.g., 3 nerve block training sessions) or cross-task transfer data (e.g., nerve block and vascular puncture training) are collected, and the mean (reflecting the overall level) and variance (reflecting the degree of fluctuation) of the first, second, and third quantitative parameters in multiple training sessions / cross-tasks are calculated. The repetition pass rate of the same anesthesia ultrasound teaching task (the number of times the quantitative parameters reached the qualified threshold in multiple training sessions / the total number of training sessions) and the cross-task transfer coefficient of different anesthesia ultrasound teaching tasks (i.e., the correlation coefficient of quantitative parameters between different tasks, reflecting the degree of correlation of learners' ability performance in different teaching tasks) are statistically analyzed to obtain the fourth quantitative parameter of the operational consistency ability dimension. Specific calculation logic is designed for each ability dimension to ensure that the quantitative parameters accurately reflect the corresponding ability level, solve the ambiguity problem of traditional assessment dimensions, integrate multi-indicator weighted calculation, take into account both core and auxiliary abilities, and improve the comprehensiveness of quantitative results.

[0059] In this embodiment, the calculation module 203 may further include a filtering unit and a dynamic calibration unit.

[0060] The filtering unit is used to remove invalid data types from the teaching data, calculate the percentage of valid data, and trigger a re-acquisition command or issue an invalidity alert if the percentage of valid data is less than a set percentage. Invalid data refers to data that cannot support capability assessment due to equipment failure, improper operation, or abnormal data transmission. The scope of invalid data can be clearly defined, including: ultrasound image data (blurred frames, frames without target structure, segments with a frame rate lower than 10 frames / second), operational behavior data (jump values ​​caused by sensor anomalies, data without timestamps), and recognition results (abnormal values ​​with a confidence level consistently of 0 or 1). A data cleaning algorithm is used to identify and remove the above invalid data, and the percentage of valid data to total data is calculated. In one example, a threshold for the percentage of valid data is set. If the actual percentage of valid data is greater than or equal to the set percentage, subsequent calculations continue. If the actual percentage of valid data is less than the set percentage, two optional processing mechanisms can be triggered: one is to send a re-acquisition command to the acquisition module (applicable to real-time training scenarios), and the other is to output an invalidity alert to the system (including the invalid data type and percentage, applicable to pre-stored image training scenarios). Removing invalid data avoids abnormal data interfering with the quantitative results, ensuring the accuracy and reliability of the assessment.

[0061] The dynamic calibration unit is used to build a parameter benchmark library based on reference operation data. At set intervals, newly added clinical case data is incorporated to update the qualified thresholds of the parameter benchmark library. Reference operation data refers to standardized operation data from skilled operators, serving as the core basis for constructing the parameter benchmark library. The parameter benchmark library is a database storing qualified and excellent thresholds for quantitative indicators of each competency dimension, forming the foundation for threshold setting and dynamic updates. Dynamic calibration is based on newly added clinical and teaching data, periodically updating threshold standards to ensure that thresholds iterate synchronously with clinical practice and teaching needs. In one example, clinical operation data and standardized training data from skilled operators can be collected to build an initial parameter benchmark library, containing qualified and excellent thresholds for quantitative parameters of each competency dimension. Then, an update cycle is set (e.g., every 3 months), and newly added clinical case data and multi-center teaching data are collected, filtering for high-quality data (valid data percentage ≥ 90%). Then, based on the newly added data, the distribution characteristics of each quantitative indicator (e.g., mean, standard deviation) are recalculated, and the qualified thresholds are adjusted in conjunction with expert feedback (e.g., if the original average confidence level qualified threshold was 80%, and newly added data shows a mean of 85% for skilled operators, the threshold is adjusted to 82%). The updated thresholds are synchronized to the contextualized threshold library of the adaptation unit to ensure that subsequent calculations use the latest standards. New clinical data is dynamically incorporated to ensure that the quantified thresholds always align with actual clinical standards, improving the clinical relevance of the assessment.

[0062] In this embodiment, the image generation module 204 may include a standardization unit, a modeling unit, a visualization unit, and a trajectory generation unit.

[0063] The standardized unit is used to normalize the quantitative parameters of each capability dimension according to clinical fit weights, resulting in a standardized score. Clinical fit weights are the proportions of importance of each capability dimension set by the registration and module based on the target anesthesia ultrasound teaching scenario, and serve as the weighting basis for the standardized score. Normalization is the process of mapping quantitative parameters of different dimensions and ranges to a fixed interval, eliminating incomparability between indicators. For example, the clinical fit weights of the target anesthesia ultrasound teaching scenario are retrieved to ensure consistency with the current teaching scenario. The min-max normalization algorithm is used to map the quantitative parameters of each capability dimension to a [0-100] score scale, resulting in a standardized score. The standardized score is a unified score obtained after normalization and weight calibration of the quantitative parameters, and is the core data for capability profile modeling.

[0064] The modeling unit is used to construct a target vector for a competency profile based on standardized scores. The target vector is a data carrier representing a learner's comprehensive competency and risk status in a structured vector format, containing competency scores and risk indicators. The dimensions of the target vector correspond one-to-one with the competency dimensions, and target risk dimension indicators are embedded within the target vector. For example, the weighted standardized scores of each dimension output by the standardization unit can be sequentially filled into the vector to form an initial competency vector. Then, target risk dimension information is extracted from error logs, and the risk level is encoded as an additional dimension of the vector, embedded at the end of the initial vector to form the target vector. Embedding risk dimension indicators allows the competency profile to simultaneously include competency level and risk status, addressing the lack of risk correlation in traditional assessments. The standardized vector format provides a data foundation for cross-scenario and cross-batch profile comparisons.

[0065] The visualization unit synchronously outputs vectors for teaching management and radar charts for learners' self-assessment, marking ability dimensions below the clinical safety threshold and target risk dimensions on the radar charts. In other words, the teaching management terminal can generate raw data for the target vectors, summarizing and displaying it by class or scenario, supporting batch comparison of learner vectors. The learning terminal can generate radar charts based on the ability dimension scores of the target vectors. The radar chart uses ability dimensions as axes, with scores corresponding to the scale on the axes. Ability dimensions below the clinical safety threshold (adaptive to unit settings) can be marked in red on the radar chart (e.g., spatial understanding score 55 < threshold 60, marked in red). Target risk dimensions are marked with special icons (e.g., high-risk indicators correspond to red exclamation marks), and the legend explains the risk type (e.g., arterial misjudgment risk). Learners can click on the radar chart axes to view detailed score composition and error log data for that dimension, facilitating self-assessment of weaknesses. The dual-mode output adapts to the different needs of teaching administrators and learners; the management end facilitates batch analysis, while the learner end facilitates self-assessment. Radar charts visually present strengths and weaknesses, while risk markers highlight key safety areas, addressing the issue of uninterpretable traditional assessment results. Interactive features provide detailed traceability, helping learners identify the causes of weaknesses and providing direction for self-training.

[0066] The trajectory generation unit is used to generate target ability profiles associated with training batches, constructing time-series ability trajectory curves. These time-series ability trajectory curves are visualized curves that sequentially string together multiple training ability profiles, reflecting the dynamic changes in a learner's comprehensive abilities. For example, by training batch (e.g., "Batch 202508: Neural Block Training," "Batch 202509: Vascular Puncture Training"), the historical target ability profile and current profile for the same ability dimension of a learner can be associated. The time-series ability trajectory curve can simultaneously record the score change trend and clinical fit change for each ability dimension. For example, with training time (or batch) as the horizontal axis and the standardized score of each ability dimension as the vertical axis, a time-series ability trajectory curve can be plotted, simultaneously annotating the clinical fit at each time point (e.g., "Batch 202508: Adapted to 80% of the clinical standard for neural block scenarios") and key events (e.g., "Batch 202509: Completed intensive training"). It also supports extracting learner growth sub-trajectories separately by ability dimension. Clinical fit is the degree of conformity between a learner's ability score and clinical standards, reflecting the clinical applicability of the ability. A growth sub-trajectory is a score change trajectory for a single ability dimension, allowing focus on the growth process of a particular ability. Constructing ability trajectory curves enables dynamic characterization of ability growth, addressing the lack of long-term tracking in traditional assessments. Adding clinical fit and key events makes the growth trajectory more interpretable, facilitating educators' analysis of training effectiveness. Support for sub-trajectory extraction accurately presents the improvement / decline trend of individual abilities, providing data support for personalized reinforcement training.

[0067] In this embodiment of the application, the mapping module 205 may include a first mapping unit, a second mapping unit, a level determination unit, an update unit, and a promotion unit.

[0068] The first mapping unit is used for clinical operation standards based on anesthesia ultrasound teaching. It sets multi-level thresholds for the baseline parameters of each capability dimension, and calculates an evaluation value by weighting the values ​​according to the configured dimension weights. The evaluation value is then matched with the corresponding first capability level. The multi-level thresholds divide each capability dimension parameter into multiple intervals based on clinical standards, with each interval corresponding to a capability level. This is the core basis for standardized level matching. For example, based on the anesthesia ultrasound clinical operation standards and the clinical qualification thresholds of the adaptation unit, multi-level thresholds can be defined for the baseline parameters of each capability dimension (e.g., L1-Beginner: <60 points, L2-Intermediate: 60-75 points, L3-Proficient: 75-90 points, L4-Expert: ≥90 points), clearly defining the parameter ranges for each level. Then, the dimension weights are retrieved, and the standardized scores of each capability dimension are summed according to their weights to obtain a comprehensive evaluation value. The comprehensive evaluation value is compared with the preset multi-level threshold intervals to match the corresponding level and output the first capability level. The first capability level is the capability level obtained through rule-based thresholds and weighted calculations, and has a clear clinical standard basis. The grading matching logic aligns with clinical operating standards, ensuring the clinical relevance of the assessment results. Based on preset rules and weight calculations, the results are interpretable and reproducible, avoiding subjective bias. Multi-level threshold division refines the ability levels, meeting the grading needs of different teaching stages.

[0069] The second mapping unit uses clinical reference competency profile data as a baseline sample and matches it with the level boundaries defined by the baseline sample using a clustering algorithm to obtain the second competency level. Clinical reference competency profile data is a standardized set of competency profiles of learners at different competency levels that has been clinically validated, forming the basis for data-driven level classification. In one example, a large amount of competency profile data of qualified clinical learners (such as standardized profiles of skilled operators and resident trainees) can be collected as a baseline sample library for clinical reference competency profiles. Unsupervised clustering algorithms (such as K-means) are used to perform cluster analysis on the baseline samples, naturally dividing the level boundaries from L1 to L4 according to competency levels. The level boundaries are the critical values ​​of the vector intervals between different competency levels after clustering, reflecting the natural distribution of competency levels in actual teaching scenarios, such as the profile vector interval corresponding to the cluster center. Then, the target competency profile vector of the current learner is compared with the level boundaries obtained from clustering, and vector similarity (such as cosine similarity) is calculated. The best-fitting level is matched, and the second competency level is output. The second competency level is the competency level obtained through data clustering and similarity matching, which conforms to the competency distribution pattern of the actual learning group. Then, extract the first capability level of the first mapping unit and the second capability level of the second mapping unit, and compare whether the two are consistent.

[0070] The level determination unit is used to determine the target ability level when the matching results of the first ability level and the second ability level are consistent, and to associate the corresponding ability description with the first ability level. The target ability level is the final graded result representing the learner's comprehensive ability level after double matching verification. Conversely, if the two are inconsistent, a secondary verification is triggered (such as recalculating the assessment value or supplementing the benchmark sample clustering) to ensure that the level determination is unbiased before outputting the result. The use of a rule-based and data-driven dual verification mechanism significantly improves the accuracy and reliability of ability level determination.

[0071] The update unit initiates a capability level model update every time a set number of new training data points are accumulated. For example, a data update threshold is set (e.g., 1000 new training data points), and the amount of valid training data collected by the system is counted in real time. Once the threshold is reached, the capability level model update process is initiated. Based on the parameter distribution of qualified clinical operations in the new training data, the grading rule thresholds for each capability dimension are adjusted. For example, qualified clinical operation data (quantitative parameters reaching the qualified threshold and no high-risk errors) is selected from the new training data to ensure the quality of the updated data. Based on the parameter distribution of the qualified data, the mean and standard deviation of each capability dimension are recalculated, and multi-level thresholds are adjusted. Additionally, the newly incorporated training data can be re-clustered to optimize the level boundaries, ensuring that the boundaries closely match the latest capability distribution.

[0072] The enhancement unit is used to correlate clinical adverse event data. Clinical adverse event data refers to records of events that occur during the clinical application or teaching training of anesthesia ultrasound, violate clinical safety protocols, and may lead to adverse consequences. Examples include vascular injury and nerve stimulation complications caused by ultrasound-guided puncture. If a capability dimension has a correlation with clinical adverse event data exceeding a set proportion, its weight in the capability level model is increased. The set proportion is a threshold for determining the correlation (i.e., the degree of impact) of the occurrence of clinical adverse events. When the correlation proportion of a capability dimension exceeds the set threshold, the weight of that capability dimension in the capability level model is automatically increased, and the dimension weights in the configuration module are updated synchronously, strengthening the assessment priority of that capability dimension. Dynamically adjusting the weights of high-risk dimensions to adapt to changes in clinical safety needs can improve the practicality of the assessment.

[0073] In this embodiment of the application, the evaluation module 206 may include an integration unit and a bottleneck matching unit.

[0074] The integration unit combines standardized scores, clinical fit, weakness dimensions, and stage-wise change trends of each capability dimension in the target capability profile, and links them to the target risk error log to generate an assessment report including a three-dimensional assessment dataset. The three-dimensional assessment dataset is a structured data set with capability score, risk level, and growth slope as its core dimensions, comprehensively representing the learner's capability level, safety status, and growth trend. Clinical fit is the degree to which the learner's capability score matches clinical standards, reflecting the adaptability of capability to clinical practice. The growth slope is the rate of change of capability score with training batches; positive numbers indicate capability improvement, negative numbers indicate decline, and the absolute value reflects the rate of growth / decline. The target risk error log contains high / medium risk error records marked during training, including error type, occurrence scenario, and related data. Integrating the three-dimensional data solves the problems of single information and lack of correlation in traditional assessment reports. Clarifying the correlation between weakness dimensions and risk errors can provide precise targets for subsequent personalized teaching. The standardized report structure can accommodate both learner self-assessment and instructor evaluation, improving the report's practicality and readability.

[0075] The bottleneck matching unit retrieves teaching resources from the anesthesia ultrasound teaching resource library that match the type of bottleneck dimension based on the type and target risk error characteristics. The learning terminal then displays a radar chart and a textual assessment report. For example, it can analyze bottleneck dimension types (e.g., anatomical structure recognition bottlenecks and operational stability bottlenecks). It extracts target risk error characteristics (e.g., artery misjudgment corresponding to structure recognition errors, probe jitter corresponding to insufficient operational stability). Based on a pre-defined association library of bottleneck type-risk characteristics-teaching resources, it retrieves matching resources from the anesthesia ultrasound teaching resource library (e.g., structural recognition bottlenecks correspond to anatomical structure ultrasound atlas interpretation and high-risk structure recognition enhancement videos; operational stability bottlenecks correspond to probe grip standard animations and stable operation simulation training levels). The learner's terminal simultaneously displays a visual radar chart (including red highlighting of bottleneck dimensions and risk indicators) and textual interpretation. The textual interpretation can include bottleneck cause analysis (e.g., insufficient operational stability due to uneven probe pressure control), personalized training lists (e.g., tasks ordered from basic consolidation to advanced enhancement to practical simulation), and a full-cycle competency growth curve (including growth trajectories and changes in clinical adaptability for each dimension). The teaching management terminal outputs a statistical report on common weaknesses of the same group of learners, a heatmap of target risk error distribution, and feedback data on the effectiveness of personalized teaching resource delivery. The target risk error distribution heatmap presents the distribution of risk errors in a specific batch / scenario in a visual heatmap format, intuitively reflecting teaching weaknesses. For example, it presents the frequency of errors by scenario / error type, with darker colors indicating higher frequency. The feedback data records learners' usage and corresponding ability changes after the delivery of teaching resources, used to optimize resource delivery strategies, such as resource clicks, completion rates, and changes in growth rates for corresponding weakness dimensions. This solves the problem of blindly pushing traditional teaching resources and improves the targeting of personalized teaching. Textual interpretations and growth curves on the learner's terminal help learners clarify the causes of weaknesses and improvement paths, enhancing self-learning efficiency. The common weakness statistics and heatmap on the teaching management terminal provide data support for optimizing teaching plans and adjusting courses. The feedback data on delivery effectiveness forms a closed-loop teaching system, continuously optimizing the resource library and delivery logic to improve overall teaching quality.

[0076] The following example, using ultrasound-guided supraclavicular brachial plexus block as a teaching assessment, can include the following steps.

[0077] (1) Task settings: The system loads the "supraclavicular brachial plexus" training task, with preset target structures including the subclavian artery, brachial plexus nerve cluster, pleural reflex, etc., and configures the weights of the ability dimensions (such as recognition ability 40%, stability 30%, spatial understanding 20%, consistency 10%).

[0078] (2) Data acquisition: Real-time ultrasound image stream is acquired during training; the structure recognition module outputs the recognition results and confidence level of the target structure; the operation acquisition module records the probe position, angle changes and image stabilization time.

[0079] (3) Calculation of capability parameters: Recognition capability: Calculate the average confidence and false recognition rate of each target structure in the effective frames; Stability: Calculate the percentage and position / angle change range for continuous and stable display for ≥5 seconds; Spatial understanding: Calculate the relative positional deviation between the brachial plexus cluster and the subclavian artery, and the matching degree of the scanning path; Consistency: Calculate the variance of the above indicators over three consecutive exercises.

[0080] (4) Capability profile generation: Generate a four-dimensional capability profile vector and present it as a radar chart.

[0081] (5) Ability level mapping: The ability profile is mapped to L1-L4 levels according to the preset threshold.

[0082] (6) Results output: Output an evaluation report (including the dimensions and trends of shortcomings) and recommend compensatory training tasks such as "Pleural recognition enhancement cases" and "Probe stability training checkpoints".

[0083] Let's take ultrasound-guided vascular puncture teaching assessment as an example to illustrate this further.

[0084] The system uses the "central venous puncture / peripheral venous puncture" task as an example. The target structure includes veins, arteries, and surrounding tissues; risk-related structures can be used to expand dimensions (such as risk awareness). In addition to outputting a competency profile, the system can also output "high-risk misjudgment statistics" (such as the number of times an artery was misjudged as a vein), which can be used for teaching safety tips and training strategy adjustments.

[0085] For competency tracking and teaching quality management, the system can accumulate competency profiles for the same learner on a weekly or training batch basis, forming a competency trajectory curve. The teaching management terminal can summarize the distribution of various dimensions by class / year of residency training, identify common weaknesses (such as weak overall spatial comprehension), and optimize course arrangements or add specialized training modules accordingly.

[0086] It should be noted that operational behavior data can come from built-in sensors on the probe, external optical / electromagnetic tracking, IMU, touchscreen / handheld trajectory, etc.; image data can be real-time ultrasound streams, recorded case files, or images synthesized from simulators. Without changing the "capability profile," dimensions can be expanded, such as the "risk awareness dimension" (related to misjudging high-risk structures) and the "decision consistency dimension" (the rationality of needle insertion path selection). Recognition capability can be assessed using average confidence, Top-k accuracy, or structural boundary overlap. Stability can be assessed using probe trajectory length, angular velocity, and the proportion of stable images. Spatial understanding can be assessed using relative structural position error or scan path matching; consistency can be assessed using cross-session variance or learning curve slope. The capability level model can use rule thresholds, statistical quantiles, machine learning classification / clustering, or milestone mapping based on a competency framework. Besides anesthesia ultrasound, this technology can also be transferred to other ultrasound training scenarios requiring "image recognition + operational coordination" (such as emergency POCUS, critical care ultrasound, etc.), without affecting the core technical points of this invention.

[0087] Compared with traditional technical solutions, the embodiments of this application have the following beneficial effects.

[0088] 1. Enables multi-dimensional and quantifiable assessment of anesthesia ultrasound competence: By parametrically calculating ultrasound image recognition results and operational behavior data, learner competence is transformed into measurable technical indicators, reducing the bias of traditional subjective scoring.

[0089] 2. It can generate interpretable ability profiles and reveal differences in ability structure: The ability profile can simultaneously present dimensions such as recognition, stability, spatial understanding, and consistency, so that the "total score" can be broken down into interpretable weakness positioning, providing a basis for precision teaching.

[0090] 3. Supports dynamic assessment of abilities over time: Ability profiles can be updated after each training session, forming an ability trajectory and growth curve, which facilitates the assessment of training effectiveness and differences in learning speed.

[0091] 4. Provides a technical foundation for tiered teaching and personalized training recommendations: Through ability level mapping and weakness identification, compensatory training or resource push can be automatically triggered, improving the automation and consistency of teaching path optimization.

[0092] 5. Facilitates the formation of a standardized evaluation system and supports large-scale promotion: A unified capability dimension and parameter calculation framework helps to form reproducible evaluation standards across tasks and institutions, supporting standardized training assessment and teaching quality management.

[0093] Those skilled in the art will understand that all or part of the functions of the various methods in the above embodiments can be implemented by hardware or by computer programs. When all or part of the functions in the above embodiments are implemented by computer programs, the program can be stored in a computer-readable storage medium, which may include: read-only memory, random access memory, disk, optical disk, hard disk, etc., and the program is executed by a computer to achieve the above functions. For example, the program can be stored in the memory of a device, and when the program in the memory is executed by the processor, all or part of the above functions can be achieved. In addition, when all or part of the functions in the above embodiments are implemented by computer programs, the program can also be stored in a server, another computer, disk, optical disk, flash drive, or external hard drive, etc., and can be downloaded or copied to the memory of a local device, or the system of the local device can be updated. When the program in the memory is executed by the processor, all or part of the functions in the above embodiments can be achieved.

[0094] The above examples illustrate this application only to aid understanding and are not intended to limit its scope. Those skilled in the art to which this application pertains can make various simple deductions, modifications, or substitutions based on the ideas presented.

Claims

1. An anesthesia ultrasound teaching assessment system based on competency profiles, characterized in that, include: A configuration module is used to predefine multiple configurable capability dimensions for anesthesia ultrasound teaching, and the dimension weight of each capability dimension is determined based on the target anesthesia ultrasound teaching scenario; The acquisition module is used to acquire teaching data of learners performing anesthesia ultrasound teaching tasks in the target anesthesia ultrasound teaching scenario. The teaching data includes at least ultrasound image data of the target anatomical structure, anatomical structure recognition results, and operational behavior data synchronized with the ultrasound image data. The calculation module is used to calculate the quantitative parameters of the learner in each of the ability dimensions based on the teaching data; The profile generation module is used to combine the quantitative parameters of multiple ability dimensions to generate a target ability profile that represents the learner's comprehensive ability status, and the target ability profile includes a risk severity level label. The mapping module is used to match the ability profile with a preset ability level model to obtain the target ability level corresponding to the learner. The assessment module is used to output an assessment report that includes the target capability profile, the target capability level, the weakness dimension, and the trend of change.

2. The anesthesia ultrasound teaching assessment system based on competency profiles according to claim 1, characterized in that, The configuration module includes: The identification unit is used to identify the target anesthesia ultrasound teaching scenario corresponding to the current anesthesia ultrasound teaching task. The target anesthesia ultrasound teaching scenario includes at least one of the following: ultrasound-guided nerve block scenario, vascular puncture scenario, airway assessment scenario, and spinal canal localization scenario. The calling unit is used to match the capability dimensions that need to be activated in the target anesthesia ultrasound teaching scenario; An adjustment unit is used to dynamically adjust the weight allocation ratio of each activated capability dimension based on the target anesthesia ultrasound teaching scenario. The disable unit is used to determine the disabled capability dimensions based on the target anesthesia ultrasound teaching scenario.

3. The anesthesia ultrasound teaching assessment system based on competency profiles according to claim 1, characterized in that, The acquisition module includes: The first acquisition unit is used to acquire the real-time ultrasound image stream of the learner or retrieve the pre-stored pathological images with clinical annotations through an ultrasound image acquisition device to obtain the ultrasound image data. Simultaneously, it triggers a target detection algorithm to identify the target anatomical structure in the ultrasound image data and obtains the anatomical structure identification result. The anatomical structure identification result includes at least the category, confidence level, and boundary overlap degree of the target anatomical structure with the reference anatomical atlas. The second acquisition unit is used to synchronously acquire the operation behavior data aligned with the timestamp of the ultrasound image data through the force sensor, electromagnetic tracking module and inertial measurement unit built into the ultrasound probe. The operation behavior data includes at least three-dimensional coordinates, three-dimensional attitude angles, contact pressure values, scanning trajectory coordinate sequences and time intervals of continuous stable images.

4. The anesthesia ultrasound teaching assessment system based on competency profiles according to claim 3, characterized in that, The acquisition module also includes: The third acquisition unit is used to acquire the learner's identity identifier and associate and bind the identity identifier with the learner's historical training data; The fourth acquisition unit is used to synchronously record task information when the anesthesia ultrasound teaching task is started. The task information includes at least the target anesthesia ultrasound teaching scene corresponding to the anesthesia ultrasound teaching task, the target anatomical structure list, and the preset risk severity level. The fifth acquisition unit is used to monitor the learner's operation behavior and recognition results in real time during the training process. When an operation or recognition result that does not meet clinical standards is detected, the type of erroneous operation is marked, and an error log consisting of the ultrasound image frame position where the erroneous operation occurred, the corresponding operation behavior data, and the timestamp is recorded synchronously. The error log is then accurately associated with the ultrasound image data and the operation behavior data according to the timestamp.

5. The anesthesia ultrasound teaching assessment system based on competency profiling according to claim 1, characterized in that, The computing module includes: The association unit is used to establish the mapping relationship between each capability dimension and the reference data source; The adaptation unit is used to set clinical qualification thresholds for the calculation indicators of each capability dimension based on the clinical operation specifications of anesthesia ultrasound, clinical big data, and expert advice data. The calculation unit is used to calculate the initial result of each capability dimension based on a preset algorithm, and to calibrate and correct the initial result to obtain the quantization parameter of each capability dimension.

6. The anesthesia ultrasound teaching assessment system based on competency profiling according to claim 5, characterized in that, The capability dimensions include anatomical structure recognition capability dimension, operational stability capability dimension, spatial understanding capability dimension, and operational consistency capability dimension. The computing unit is used for: Regarding the anatomical structure recognition ability dimension, based on the recognition results of the target anatomical structure in the ultrasound image data, an effective time window is defined in the training process, the confidence sequence of the target anatomical structure in the effective time window is extracted, the average recognition confidence and confidence fluctuation variance are calculated, and the false recognition rate is extracted to form the first quantitative parameter of the anatomical structure recognition ability dimension. For the operational stability capability dimension, the position and angle time sequence data of the ultrasound probe are extracted from the operational behavior data, the position change and angle change per unit time are calculated, and combined with the clarity evaluation results of the ultrasound image data, the time required to achieve stable images and the ratio of the duration of stable images to the total training time are determined as the second quantitative parameter of the operational stability capability dimension. The stable image is ultrasound image data with a clarity greater than the set clarity. For the spatial understanding ability dimension, a preset anesthesia ultrasound baseline scanning path and corresponding baseline structure relative position template are invoked. The learner's scanning path data is compared with the baseline scanning path to calculate the path matching degree. At the same time, the relative position deviation value of the target anatomical structure in the ultrasound image data is analyzed. Combined with the structural positioning results in the dynamic scene, the positioning consistency index is statistically analyzed to form the third quantitative parameter of the spatial understanding ability dimension. Regarding the operational consistency ability dimension, based on the learner's multiple training data in the same anesthesia ultrasound teaching task or the transfer training data in different anesthesia ultrasound teaching tasks, the mean and variance of the anatomical structure recognition ability dimension, the operational stability ability dimension, and the spatial understanding ability parameter are calculated. The repeated achievement rate of the same anesthesia ultrasound teaching task and the cross-task transfer coefficient of different anesthesia ultrasound teaching tasks are statistically analyzed to obtain the fourth quantitative parameter of the operational consistency ability dimension.

7. The anesthesia ultrasound teaching assessment system based on competency profiling according to claim 5, characterized in that, The computing module also includes: The filtering unit is used to remove invalid data types from the teaching data, calculate the percentage of valid data, and trigger a re-collection command or issue an invalid reminder if the percentage of valid data is less than a set percentage. The dynamic calibration unit is used to build a parameter benchmark library based on reference operating data, and to update the qualified threshold of the parameter benchmark library by incorporating newly added clinical case data at set intervals.

8. The anesthesia ultrasound teaching assessment system based on competency profiling according to claim 1, characterized in that, The portrait generation module includes: A standardized unit is used to normalize and transform the quantitative parameters of each capability dimension according to clinical fit weights to obtain a standardized score. A modeling unit is used to construct a target vector for a capability profile based on the standardized score. The dimensions of the target vector correspond one-to-one with the capability dimensions, and a target risk dimension identifier is embedded in the target vector. A visualization unit is used to simultaneously output vectors for teaching management and radar charts for learners' self-assessment, and the radar charts are marked with the ability dimension and the target risk dimension that are below the clinical safety threshold. The trajectory generation unit is used to generate the target ability profile according to the training batch, construct a time series ability trajectory curve, the time series ability trajectory curve synchronously records the score change trend and clinical fit change of each ability dimension, and supports the extraction of the learner's growth sub-trajectory by the ability dimension.

9. The anesthesia ultrasound teaching assessment system based on competency profiling according to claim 1, characterized in that, The mapping module includes: The first mapping unit is used to set multi-level thresholds for the baseline parameters of each capability dimension based on the clinical operation specifications of the anesthesia ultrasound teaching, perform weighted calculations according to the configured dimension weights to obtain an evaluation value, and match the first capability level corresponding to the evaluation value. The second mapping unit is used to match the clinical reference capability profile data with the level boundaries of the benchmark sample through a clustering algorithm to obtain the second capability level. A level determination unit is used to take the first capability level as the target capability level when the matching results of the first capability level and the second capability level are consistent. The update unit is used to initiate the capability level model update when a set number of new training data are accumulated, and adjust the grading rule threshold of each capability dimension based on the parameter distribution of clinical qualified operations in the new training data, and re-cluster the incorporated new training data to optimize the level boundary. An enhancement unit is used to correlate clinical adverse event data. If there is a capability dimension whose correlation with the clinical adverse event data is greater than a set proportion, the weight of the capability dimension in the capability level model is increased.

10. The anesthesia ultrasound teaching assessment system based on competency profiling according to claim 1, characterized in that, The evaluation module includes: The integration unit is used to integrate the standardized scores, clinical fit, weakness dimensions, and stage change trends of each capability dimension in the target capability profile, and associate them with the target risk error log to generate an assessment report including a three-dimensional assessment dataset, which includes capability scores, risk levels, and growth slopes. The weakness matching unit is used to retrieve teaching resources from the anesthesia ultrasound teaching resource library that match the type of the weakness dimension based on the type of the weakness dimension and the characteristics of the target risk error. The learning terminal of the learning unit presents an assessment report with a radar chart and text interpretation. The text interpretation includes weakness cause analysis, personalized training list and full-cycle ability growth curve. The teaching management terminal outputs a statistical report of common weaknesses of the same group of learners, a heat map of target risk error distribution and feedback data on the effect of personalized teaching resource push.

Images