Medical scan protocol management system, medical imaging system and scan control method thereof

Abstract

The application relates to a medical scanning protocol management system, a medical imaging system and a scanning control method thereof. The medical scanning protocol management system comprises: an initial protocol package creation unit, containing a plurality of initial protocol packages, each initial protocol package containing scanning protocol parameter values and a corresponding scanning site; an analysis unit for analyzing the initial protocol package to generate a first natural language description, the first natural language description being a standardized description; a self-defined unit for setting a second natural language description for at least one initial protocol package in response to a user's input operation, the second natural language description being a self-defined description; and a target protocol generation unit for associating the first natural language description and / or the second natural language description with the initial protocol package to generate a plurality of target scanning protocols, the plurality of target scanning protocols being in a same-level parallel relationship. The embodiments of the application meet the retrieval of scanning protocols by the user's diversified text description, and improve the examination efficiency.

Description

Technical Field

[0001] This application belongs to the field of medical technology, and in particular relates to a medical scanning protocol management system, a medical imaging system, and a scanning control method for the medical imaging system. Background Technology

[0002] With the continuous development of medical imaging technology, medical scanning protocols play a crucial role in medical imaging equipment. A medical scanning protocol refers to a series of parameter settings and operational steps used during scanning on a medical imaging device. These parameters directly affect the quality and diagnostic value of the scanned images. Currently, medical imaging equipment typically has a large number of preset scanning protocols to adapt to different clinical needs and scanning sites. Magnetic Resonance Imaging (MRI) technology is an important tool in the field of modern medical imaging. Its core lies in generating high-resolution images of the internal structures of the human body through the phenomenon of nuclear magnetic resonance. MRI systems control the imaging process through complex pulse sequences and parameter settings; these sets of parameters are called MRI protocol packets. Protocol packets define the type of scanning sequence, such as T1-weighted imaging, T2-weighted imaging, or diffusion-weighted imaging (DWI); imaging parameters, such as repetition time (TR), echo time (TE), inversion time (TI), and other related settings. For example, in a brain scan, the protocol package might specify the use of T1-weighted imaging to highlight the anatomical structures of brain tissue, while in a liver examination, dynamic contrast-enhanced scanning might be used to detect lesions.

[0003] Currently, the management of MRI protocol packets mainly includes two types: hierarchical storage and DICOM metadata-based methods. Hierarchical storage is the more common management method in current MRI systems. It uses a three-level classification structure of "major protocol type - examination site - protocol packet name" to store and manage protocol packets according to a preset hierarchical relationship. For example, in brain scan protocol packets, users need to navigate through levels such as "manufacturer" - "site" - "specific protocol" to find the required protocol packet. This hierarchical structure is inefficient when there are many protocol packets, requiring users to expand the directory level by level, which is cumbersome and does not allow for cross-level searches. Furthermore, hierarchical storage lacks the ability to personalize protocol packets, failing to meet users' needs to add additional information. On the one hand, users often need to add their name, usage date, or other remarks to protocol packets, and the existing system cannot provide this flexibility. On the other hand, users can only understand the function and content of the protocol packet through its name, but different sites and different users have different naming rules for protocol packets, making it difficult for unfamiliar users to understand the specific function of the protocol packet. In specific clinical scanning scenarios, when a user defines a scanning protocol, other users may find it difficult to understand the specific purpose of the protocol package when they see the name, and they may not be able to select the protocol package for scanning. As the number of protocol packages increases and their complexity increases, this hierarchical structure becomes difficult to adapt to changes and fails to meet users' needs for rapid retrieval.

[0004] The DICOM metadata-based management method extracts technical parameters (such as TR, TE, etc.) from the DICOM file as search criteria. While this method can match the technical parameters of protocol packets, it primarily focuses on technical details and fails to identify functional characteristics such as the scanning site or clinical application of the protocol. Therefore, users find it difficult to quickly understand the actual application scenario when searching for protocol packets, resulting in unintuitive information and a poor user experience.

[0005] In summary, while the existing hierarchical storage method for MRI protocol packets achieves some level of packet management, it suffers from significant shortcomings in retrieval efficiency, information intuitiveness, and user customization, thus limiting the practicality and user experience of the MRI protocol management system. Similarly, the DICOM metadata-based management method also suffers from insufficient information intuitiveness and a poor user experience. Therefore, it is necessary to improve the existing MRI protocol packet management method to enhance the efficiency of MRI operations and the timeliness of diagnosis. Summary of the Invention

[0006] To address the issues of low retrieval efficiency and unintuitive information display in existing magnetic resonance protocol packet management systems, this application provides a medical scanning protocol management system, a medical imaging system, and a scanning control method thereof.

[0007] According to a first aspect of this application, a medical scanning protocol management system is provided, comprising: an initial protocol packet creation unit, a parsing unit, a customization unit, and a target protocol generation unit. The initial protocol packet creation unit contains multiple initial protocol packets, each containing scanning protocol parameter values ​​and a corresponding scanning site. The parsing unit parses the initial protocol packets to generate a first natural language description, which is a standardized description. The customization unit, in response to user input, sets a second natural language description for at least one initial protocol packet, which is a customized description. The target protocol generation unit associates the first and / or second natural language descriptions with the initial protocol packets to generate multiple target scanning protocols, which are parallel and at the same level.

[0008] Optionally, the first natural language description includes at least one of the following: the protocol type of the initial protocol packet, the scanner model, and the corresponding clinical diagnostic disease.

[0009] Optionally, the second natural language description includes at least one of the following: user name, creation date, and patient type.

[0010] Optionally, multiple target scanning protocols can be relatively independent of each other.

[0011] Optionally, at least two target scanning protocols contain the same scanning protocol parameter values ​​and the two target scanning protocols are associated with each other.

[0012] Optionally, the medical scanning protocol management system also includes an import unit for importing external data sources, which are used to update the scanning protocol parameter values ​​of the target scanning protocol.

[0013] According to a second aspect of this application, a medical imaging system is provided, comprising a scanner, a processor, and an interactive device. The scanner performs an imaging scan on a target object to acquire imaging data. The processor reconstructs the imaging data, and the interactive device is communicatively coupled to the processor, including a user interface. The processor is configured with executable instructions, which, when executed, cause the processor to: acquire keyword instructions; perform a retrieval operation on a database based on the keyword instructions to acquire a target scanning protocol, wherein the database contains multiple target scanning protocols with parallel relationships, each target scanning protocol contains multiple initial protocol packets and associated natural language descriptions, the initial protocol packets containing scanning protocol parameter values ​​and corresponding scanning sites, and the natural language descriptions containing standardized or custom descriptions; and control the scanner to execute the target scanning protocol to perform an imaging scan on the target object.

[0014] Optionally, before controlling the scanner to perform imaging scanning on the object to be detected based on the scanning protocol, the processor also performs: controlling the user interface to display at least one recommended target scanning protocol; and adding the target scanning protocol to the scan list in response to the user's decision operation on the user interface to determine the scan protocol.

[0015] Optionally, keyword instructions can be obtained by: responding to user input on the user interface to obtain user instructions; and inputting the user instructions into a natural language parsing model to obtain keyword instructions.

[0016] Optionally, keyword instructions can be obtained by: acquiring a scanned image of the application form; converting the scanned image of the application form into text information; and inputting the text information into a natural language processing model to obtain keyword instructions.

[0017] According to a third aspect of this application, a scanning control method for a medical imaging system is provided. The scanning control method includes: acquiring keyword instructions, which are obtained by parsing user input commands using natural language or by parsing scanned images from a request form; using the keyword instructions to search a database to obtain a target scanning protocol, wherein the database contains multiple target scanning protocols with parallel relationships, each target scanning protocol contains multiple initial protocol packets and associated natural language descriptions, the initial protocol packets contain scanning protocol parameter values ​​and corresponding scanning sites, and the natural language descriptions contain standardized or custom descriptions; and controlling the medical imaging system based on the target scanning protocol to perform imaging scanning on the object to be detected.

[0018] Optionally, the medical imaging system is at least one of an MR system, a CT system, a PET-MR system, or a PET-CT system.

[0019] The beneficial effects of this application are as follows: Each initial protocol packet contains scanning protocol parameter values ​​and corresponding scanning locations. Parsing the initial protocol packet can generate a standardized first natural language description. The standardized description covers key information such as the protocol's functions, usage scenarios, and applicable devices, enabling users to quickly understand the core functions and scope of application of the protocol packet, thus enhancing the intuitiveness of the information. A custom second natural language description can be set for the initial protocol packet through user input, allowing users to add personalized descriptions according to their needs, such as name, usage date, and remarks, meeting the diverse needs of different users and enhancing the system's flexibility and practicality. The first and second natural language descriptions are associated with the initial protocol packet to generate multiple target scans at the same level. The protocol description feature employs a hierarchical management structure, placing protocol packets at the same level. This avoids the complexity of traditional tree-like structures, allowing users to quickly locate target protocol packets directly through keyword searches without having to navigate through directories, significantly improving retrieval efficiency. When operating a medical imaging system loaded with multiple target scanning protocols, the feature provides a text-based keyword search function. The medical imaging system quickly matches and returns relevant results based on the text description, not only improving retrieval efficiency but also accurately locating protocol packets that meet user needs, resulting in a more intuitive information display. Furthermore, it supports multilingual descriptions and protocol packet version management, is compatible with existing protocol packet formats, and can adapt to the needs of different regions and application scenarios, improving protocol packet management efficiency and user experience. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this application or the conventional technology, the drawings used in the description of the embodiments or the conventional technology will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a block diagram of a medical scanning protocol management system provided in an embodiment of this application;

[0022] Figure 2 This is a schematic diagram of the structure of a medical imaging system provided in one embodiment of this application;

[0023] Figure 3 This is a schematic flowchart of a scanning control method for a medical imaging system provided in an embodiment of this application;

[0024] Figure 4 This is a schematic flowchart of a scanning control method for a medical imaging system provided in another embodiment of this application;

[0025] Figure 5 This is a schematic diagram of a database provided in an embodiment of this application;

[0026] Figure 6 This is a schematic diagram of database search results provided in an embodiment of this application;

[0027] Figure 7 This is a schematic diagram of a recommended scanning protocol method for a medical imaging system provided in an embodiment of this application;

[0028] Figure 8 This is a schematic diagram of a protocol recommendation process for head scanning provided in an embodiment of this application;

[0029] Figure 9 This is a schematic diagram illustrating the acquisition of a natural language parsing model according to an embodiment of this application;

[0030] Figure 10 This is a schematic diagram of a user interface provided in an embodiment of this application. Detailed Implementation

[0031] To facilitate understanding of this application, a more complete description will be provided below with reference to the accompanying drawings, which illustrate embodiments of the present application. However, the present application can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that the disclosure of this application will be thorough and complete.

[0032] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application.

[0033] Existing medical imaging protocol package management systems typically employ hierarchical structures or DICOM metadata-based management methods. While hierarchical structures can categorize protocol packages, users must navigate through levels of the directory, hindering rapid retrieval. DICOM metadata-based management, although capable of extracting some technical parameters, fails to identify the functional characteristics of medical imaging protocol packages, such as scan sites or clinical applications. To obtain diverse patient physiological information, users often perform MRI scans using multiple contrast parameters. However, due to individual patient differences, clinicians frequently combine various medical imaging protocol packages for direct use during scanning. Users group different contrast protocols into a single package, categorizing them by name. In complex scenarios, users struggle to quickly find the appropriate protocol package, especially with a large number of packages, resulting in low retrieval efficiency and a high risk of errors. Furthermore, existing medical imaging systems lack personalized descriptions for protocol packages, failing to meet users' needs for adding additional information. Moreover, the lack of standardized naming conventions for protocol packages leads to chaotic categorization and retrieval within the system, increasing the learning curve for new users and hindering system usability.

[0034] To address the aforementioned issues, this application improves upon existing management methods for medical imaging protocol packages by adding rich text descriptions to each package. These descriptions cover key information such as the protocol's function (scanning site, scanning purpose, etc.), usage scenarios (clinical diagnosis, research experiments, etc.), and applicable equipment. This approach allows users to quickly understand the core functions and scope of application of the protocol package, resolving the problem of unintuitive information in existing technologies. Users are also allowed to add personalized descriptions based on their needs, such as name, usage date, and remarks. This customization feature meets the diverse needs of different users, enhancing the system's flexibility and practicality. During actual clinical scanning, users can directly search for protocol packages by entering text keywords. The system quickly matches and returns relevant results based on the text descriptions. This text-based retrieval method not only improves retrieval efficiency but also accurately locates protocol packages that meet user needs. Multiple protocol packages are managed in a hierarchical structure, placing all packages at the same level. This avoids the complexity of traditional tree structures, allowing users to quickly locate the target protocol package directly through keyword search without having to navigate through levels of the directory, significantly improving retrieval efficiency. The technical solution of this application will be clearly and completely described below with reference to the accompanying drawings and embodiments.

[0035] This application proposes a medical scanning protocol management system for use in medical imaging systems. The medical scanning protocol management system can be a Picture Archiving and Communication System (PACS) or a Radiology Information System (RIS). The medical scanning protocol management system acts as a server for storing, managing, and transmitting medical image data. It receives, stores, and retrieves digital images generated by the aforementioned medical imaging scanning system, and transmits these images via a network to the workstations of doctors or other medical professionals for viewing and analysis. Users can quickly retrieve and use these images when needed, provided they have certain authorization. The medical scanning protocol management system also includes auxiliary diagnostic management functions. Figure 1 The diagram shown is a block diagram of a medical scanning protocol management system according to an embodiment of this application. The medical scanning protocol management system includes an initial protocol packet creation unit 102, a parsing unit 104, a customization unit 106, and a target protocol generation unit 108.

[0036] The initial protocol package creation unit 102 contains multiple initial protocol packages, each containing scanning protocol parameter values ​​and the corresponding scanning site. Initial protocol packages can be customized by the user in the medical scanning protocol management system or selected by the user from a public database, which can be stored locally or shared on a cloud platform. Each initial protocol package is specified by the user in terms of its technical parameters and applicable scanning scenarios. The technical parameters of the initial protocol package can include, for example, values ​​of scanning parameters such as repetition time (TR), echo time (TE), reversal time (TI), flip angle (FA), number of excitations (NEX), field of view (FOV), slice thickness, spacing, tube voltage, tube current, image cycle time, acquisition time, and contrast agent scheme. They can also include scanning protocol types such as T1-weighted imaging, T2-weighted imaging, and proton density-weighted imaging; scanning modes such as contrast-enhanced scanning, non-contrast scanning, and plain scanning; or scanner model parameters such as the field strength, number of detector rows, and manufacturer-defined product name of the medical imaging scanning system. The scanning scenarios can be classified according to organs or tissues, such as brain scans, head scans, heart scans, spine scans, and knee joint scans, and can also be used in scientific research, clinical settings, and other scenarios.

[0037] The parsing unit 104 is used to parse the initial protocol packet to generate a first natural language description, which is a standardized description. Parsing the initial protocol packet may include: keyword identification and extraction, relation extraction, and co-occurrence analysis. Magnetic resonance imaging (MRI) devices from different manufacturers may differ in hardware design, imaging parameters, and scanning protocols. Taking MRI sequences as an example, MRI sequence development is flexible, with differences between manufacturers and copyright issues related to sequence designers and institutions, resulting in a wide variety of MRI sequence names. Even for the same scanning sequence designed to address artifacts, different manufacturers may use different names. The parsing of the initial protocol packet first involves identifying and extracting keywords from the initial protocol packet. This process can employ heuristic rules, dictionary matching rules, machine learning, and other methods to identify and obtain keywords such as parameter types, scanning protocol parameter values, scanning protocol types, and scanning sites, and then extract the identified keyword information. Secondly, relation extraction is performed, extracting relational structures from the initial protocol packet or from the previously extracted keywords. Finally, co-occurrence analysis is performed. Based on the extracted keywords and relational structures, a semantic-level understanding of the initial protocol packet is obtained, establishing the association between the initial protocol packet and the standardized description. The first natural language description may include one or more combinations of the initial protocol packet's protocol type, scanner model, and corresponding clinical diagnostic disease.

[0038] The first natural language description (NLP) is a standardized description of the initial protocol package. Its format and content are predefined, ensuring consistency and standardization. For example, for an initial protocol package for a head CT scan, the parsing unit might generate the following NLP description: "Head CT scan protocol, tube voltage 120kV, tube current 250mA, slice thickness 5mm, scan time 0.5s / revolution, suitable for routine head examinations." This standardized description allows physicians and technicians to quickly understand the basic information and scope of application of the protocol.

[0039] The customization unit 106 is used to respond to user input operations and set a second natural language description for one or more initial protocol packets. The second natural language description is a customized description. Users can add various personalized descriptions such as name, usage date, and remarks according to their own needs. After the user inputs a sentence describing their needs, the customization unit 106 first analyzes the sentence structure, determines the sentence type based on the sentence structure, and extracts sentence keywords. The customization unit 106 further uses reasoning and context methods based on sentence keywords to obtain candidate sentences for the customized description; then, it performs structural analysis on the candidate sentences for the customized description, calculates sentence similarity, and removes duplicates or similarities based on sentence similarity; finally, it extracts the deduplicated customized description based on the sentence type requested by the user. In this embodiment, the second natural language description may include one or more combinations of user name, creation date, and patient type.

[0040] The target protocol generation unit 108 associates the first and second natural language descriptions with the initial protocol packet to generate multiple target scanning protocols. These multiple target scanning protocols are parallel and form a protocol database. The multiple target scanning protocols can be relatively independent of each other. Each target scanning protocol includes at least one parameter, such as scanning protocol name, setup time, slice thickness, image span, pitch coefficient, scan time, reconstruction interval, reconstruction parameters, patient position, and dose. The multiple target scanning protocols can also be interconnected. Two or more target scanning protocols may contain the same parameters. Sorting, modifying, or deleting parameters in one target scanning protocol will cause the same parameters in other associated target scanning protocols to be sorted, modified, or deleted synchronously.

[0041] The second natural language description is a user-defined description whose content and format are determined by the user, offering flexibility and personalization. For example, a user might add the following second natural language description to the initial protocol package for the aforementioned head CT scan: "Head CT scan protocol for Dr. Wang, neurologist, created on April 15, 2025, primarily for routine head examinations of elderly patients." This custom description allows users to add more useful information to the protocol based on their work habits and clinical needs.

[0042] This application embodiment generates standardized descriptions for parsing initial protocol packets, adding rich text descriptions to each packet. The content covers key information such as the protocol's function, usage scenarios, and applicable devices. This description method allows users to quickly understand the core functions and scope of application of the protocol packet, solving the problem of unintuitive information in existing technologies. Custom descriptions can be set for initial protocol packets according to user needs, allowing users to add personalized descriptions such as name, usage date, and remarks, meeting the diverse needs of different users and enhancing the system's flexibility and practicality. Target scanning protocols are arranged in a parallel hierarchy, and the database uses a horizontal management structure, placing all protocol packets at the same level. This avoids the complexity of traditional tree structures, allowing users to quickly locate target protocol packets directly through keyword searches without having to expand through directories, significantly improving retrieval efficiency.

[0043] In another embodiment, the medical scanning protocol management system may further include an import unit 110, which is used to import external data sources to update the scanning protocol parameter values ​​of the target scanning protocol. The import unit 110 is an extension component of the medical scanning protocol management system, and its main function is to interact with external systems or data sources to obtain the latest scanning protocol parameter values ​​and update the target scanning protocol in the system with these parameter values. External data sources may be parameter update packages provided by medical device manufacturers, containing the latest recommended device parameter values; best practice parameter sets published by medical research institutions, based on the latest research findings; scanning protocol parameters shared by other hospitals or medical institutions, which may reflect the clinical experience of different institutions; or protocol parameters stored in a cloud database, which may be jointly maintained and updated by multiple institutions.

[0044] In this embodiment, the import unit 110 establishes a connection with these external data sources through a preset interface to obtain the latest scanning protocol parameter values. The obtained parameter values ​​can directly update the target scanning protocol in the medical scanning protocol management system, or they can be verified and screened before being updated. This update mechanism enables the system to keep pace with the latest technologies and clinical practices, improving the accuracy and effectiveness of the scanning protocols. For example, the workflow of the import unit 110 may include: first, the import unit 110 periodically or according to user instructions establishes a connection with the preset external data source; the import unit 110 obtains the latest scanning protocol parameter values ​​from the external data source; the import unit 110 compares the obtained parameter values ​​with existing target scanning protocols in the database; if an update is found, the import unit 110 prompts the user whether to accept the update; if the user accepts the update, the import unit 110 applies the new parameter values ​​to the corresponding target scanning protocol; simultaneously, the import unit 110 records the update time, source, and content for subsequent tracking and management. In this embodiment, the introduction of the import unit 110 enables the medical scanning protocol management system to continuously absorb and apply the latest scanning technologies and parameters, improving the system's practicality and advancement.

[0045] The medical imaging system that includes the aforementioned medical scanning protocol management system can be an ultrasound (US) imaging system, an X-ray imaging system, a computed tomography (CT) imaging system, a magnetic resonance (MR) imaging system, a positron emission tomography (PET) imaging system, a single photon emission computed tomography (SPECT) imaging system, and a multimodal imaging system, such as a PET / CT, PET / MR, or SPECT / CT imaging system.

[0046] like Figure 2 The diagram shown is a schematic representation of a medical imaging system according to an embodiment of this application. The medical imaging system includes a scanner 202, an interactive device 204, and a processor 206. The scanner 202 and the interactive device 204 can be connected to the processor 206 via a network. Alternatively, the scanner 202 and the interactive device 204 can be directly connected to the processor 206.

[0047] Scanner 202 is used to perform imaging scans on the object to be detected to obtain imaging data. In this embodiment, scanner 202 is described as an MR scanner. The MR scanner includes a magnet assembly, a gradient coil assembly, and a radio frequency coil assembly. The magnet assembly generates a main magnetic field. The magnet assembly is a superconducting magnet, which forms a main magnetic field in the detection area after excitation.

[0048] The gradient coil assembly, located inside the magnet assembly, generates gradient field components Gx, Gy, and Gz in different directions. These gradient fields are superimposed on the main magnetic field generated by the magnet assembly, allowing the magnetic orientation of the object's protons to vary according to their position within the gradient field, thereby encoding spatial information into an MR signal generated by the imaged object region. The gradient coil assembly may include an X-coil for generating the gradient field component Gx in the x-direction; a Y-coil for generating the gradient field component Gy in the y-direction; and a Z-coil for generating the gradient field component Gz in the z-direction. In some embodiments, the Z-coil may be designed based on a Maxwell coil, while the X and Y coils may be designed based on a Golay coil configuration. The three sets of coils execute parameters related to the gradient pulse in the scanning protocol to generate three distinct magnetic field components for position encoding. The gradient coil assembly enables spatial encoding of the MR signal.

[0049] An RF coil assembly may include multiple RF coils. Each RF coil may include one or more RF transmitting coils and / or one or more RF receiving coils, with each RF coil comprising one or more RF receiving coil units. The RF transmitting coil can transmit RF pulses toward an object. Under the combined action of a main magnetic field and a gradient magnetic field, the applied pulse sequence generates an object-related MR signal. The RF receiving coil can acquire the MR signal from the object. The pulse sequence can be defined by scan protocol parameters and settings associated with the scan protocol parameters in the time series. Scan protocol parameters may include parameters related to the RF pulses transmitted by the RF coils, such as the number of excitations (NEX), bandwidth, etc.; parameters related to the gradient field generated by the gradient coils (e.g., gradient direction, duration of gradient application, etc.); and parameters related to the MR signal, such as echo time (TE), echo sequence length (ETL), etc. The pulse sequence can be defined by one or more time-related parameters, such as repetition time (TR), acquisition time (TA), etc.

[0050] One or more RF coils can both transmit RF pulses and receive MR signals at different times. In some embodiments, the function, size, type, geometry, location, number, and / or amplitude of the RF coils can be determined or changed according to one or more specific conditions. For example, RF coils can be classified as volumetric coils and local coils depending on their function and size. In some embodiments, the RF receiving coil unit can correspond to a channel used for acquiring MR signals.

[0051] Interactive device 204 is communicatively coupled to processor 206 and includes a user interface. Interactive device 204 may include a mobile device, tablet, laptop, or similar device, or a combination thereof. Optionally, interactive device 204 may include input devices, output devices, etc. Optionally, interactive device 204 is part of scanner 202. Interactive device 204 may include input devices such as a keyboard, mouse, pen, voice input device, touchscreen, etc. It may also include output devices such as display devices, speakers, printers, etc. The user interface of interactive device 204 may display information about recommended target scanning protocols in the form of lists, cards, or detail pages. The displayed information may include the protocol's name, description, parameter settings, scope of application, etc., helping users understand the characteristics and applicable situations of each protocol. The user interface may also provide sorting, filtering, and comparison functions to help users browse and select protocols more conveniently.

[0052] The processor 206 is configured with one or more executable instructions, which are computer program code propagation data signals. When executed, the executable instructions cause the processor 206 to: acquire keyword instructions; perform a retrieval operation on a database based on the keyword instructions to acquire a target scanning protocol, wherein the database contains multiple target scanning protocols with parallel relationships at the same level, each target scanning protocol contains multiple initial protocol packets and associated natural language descriptions, wherein the initial protocol packets contain scanning protocol parameter values ​​and corresponding scanning areas, and the natural language descriptions contain standardized or custom descriptions; and control the scanner to execute the target scanning protocol to perform imaging scanning on the object to be detected.

[0053] Keyword commands can be obtained by parsing text or voice information. For example, when a user performs input operations such as text, mouse clicks, selections, touch, or voice commands on a user interface, user commands can be obtained. Inputting these user commands into a natural language processing (NLP) model yields keyword commands. Keyword commands can also be obtained by parsing scanned application forms. For example, a user obtains a scanned image of an application form, which may be filled out by a clinician or the patient. The scanned image is converted into text information. Inputting this text information into a NLP model yields keyword commands. Specifically, to ensure the accuracy of automatic keyword command acquisition, the NLP model processes the text information. Users can define the scope of keyword commands, and the NLP model can identify keywords within the user-specified scope and exclude keywords outside of it.

[0054] To recognize the diverse expressions of different users and hospitals, the acquisition of a natural language parsing model involves three stages: selecting a base model, continuous pre-training, and post-training.

[0055] In the base model selection stage, a general large model is chosen as the basic framework.

[0056] In the incremental pre-training stage, medical knowledge is injected into the large model to address the gaps in its medical knowledge. A medical corpus is constructed using multi-source data, and a masked language model (MLM) is used for self-supervised learning. This enables the large model to learn semantic associations such as "disease-alias", "scanning protocol-manufacturer-custom name", and "functional scan-manufacturer-custom feature name", thus solving the problem that the large model "cannot understand medical terminology".

[0057] In the post-training phase, supervised fine-tuning is performed using tens of thousands of "concept variant-standard expressions" (e.g., "DM→diabetes") annotated by medical experts. Multi-task learning is combined to optimize concept standardization and keyword extraction. Subsequently, keyword weights are adjusted based on user feedback (e.g., keywords preferred by doctors for diagnosis, keywords preferred by technicians for scanning protocols), achieving a complete parsing process from "domain knowledge injection → semantic alignment → preference adaptation." For example, "DM" is standardized to "diabetes," and "diabetic foot" is mapped to "diabetic foot," with weighted keywords output according to role requirements, improving the accuracy and personalization of medical text processing.

[0058] Multiple target scanning protocols can be relatively independent of each other. Each target scanning protocol includes at least one parameter, such as scanning protocol name, setup time, slice thickness, image span, pitch coefficient, scan time, reconstruction interval, reconstruction parameters, patient position, and dose. Multiple target scanning protocols can also be interconnected, containing the same parameters. Sort, modify, or delete parameters in one target scanning protocol, and this will trigger simultaneous sorting, modification, or deletion of the same parameters in other related target scanning protocols.

[0059] The database included in the medical scan protocol management system is illustrated using magnetic resonance pulse sequences (also known as magnetic resonance imaging protocols) as an example. Protocol types for magnetic resonance pulse sequences may include spin echo (SE) pulse sequences, gradient focused echo (GRE) pulse sequences, inversion recovery (IR) pulse sequences, multi-echo magnetic resonance pulse sequences, T1ρ preprocessed pulse sequences, T2 preprocessed pulse sequences, diffusion-weighted imaging (DWI) pulse sequences, etc.

[0060] The aforementioned magnetic resonance pulse sequence can be defined by one or more parameters, such as the application time of the magnetic resonance pulse sequence, the duration of the magnetic resonance pulse sequence, parameters related to the radio frequency pulses in the magnetic resonance pulse sequence, the count of radio frequency pulses in the magnetic resonance pulse sequence, repetition time (TR), number of repetitions, invert time (TI), b-value, T1ρ preparation duration, T2 preparation duration, echo sequence length, echo interval, velocity coding (VENC) value, and the count of average values. TR, as used in this specification, can refer to the time interval between two repetitive and consecutive radio frequency pulses in the magnetic resonance pulse sequence (e.g., the time interval between two consecutive excitation radio frequency pulses in an SE pulse sequence, or the time interval between two consecutive 180° invert pulses in an IR pulse sequence). The number of repetitions can refer to the count of repetitions in the magnetic resonance pulse sequence. TI can refer to the time interval between a 180° invert pulse and a subsequent 90° excitation pulse in an IR pulse sequence. The b-value can refer to a factor reflecting the intensity and time of the diffusion sensitization gradient in the DWI pulse sequence. The T1ρ preparation duration can refer to the duration of the spin-locked pulse in the T1ρ preparation pulse sequence. The T2 preparation duration can refer to the duration of the T2 preparation pulse in the T2 preparation pulse sequence.

[0061] The medical scan protocol management system's database contains multiple user-created scan protocols. Each user-created scan protocol includes a first natural language description and a second natural language description. The first natural language description is a standardized description. However, considering that different hospitals and users may use different methods to express the same medical concepts, making it difficult to extract medical terminology using a single rule, the second natural language description is a customized description. Furthermore, multiple user-created scan protocols are listed at the same level, allowing the system to adapt to different user input requirements and obtain the target scan protocol. For example, one user-created scan protocol is a protocol package for brain T1-weighted imaging, with TR set to 2000ms and TE to 20ms. The medical scan protocol management system generates a first natural language description based on the protocol package's technical parameters and scanning scenario: "For brain T1WI, suitable for clinical diagnosis, suitable for MR equipment with a 3.0T scanner." The user adds a personalized description to the protocol package based on their needs: "Created by Zhang San, for a brain scan project in April 2025, suitable for the diagnosis of Alzheimer's disease patients."

[0062] Optionally, the same scanning protocol in the database can have different versions, and users can view different versions of the protocol package and its description. For example, in the MRI department of a hospital, Dr. Li needs to regularly update and view different versions of the brain T1-weighted imaging protocol package. The system supports version management of protocol packages, allowing users to view and select different versions. Specifically, Dr. Li can view a list of all brain T1-weighted imaging protocol packages:

[0063] Version 1.0: Applicable to T1-weighted imaging of the brain, suitable for clinical diagnosis, and compatible with 1.5T MRI scanners (TR=2000ms, TE=20ms).

[0064] Version 2.0: Suitable for T1-weighted imaging of the brain, applicable to clinical diagnosis, and compatible with MRI scanners of the 3.0T model (TR=1900ms, TE=18ms). Note: Optimized noise reduction technology improves image quality.

[0065] Dr. Li can click on version 2.0, and the system will display detailed information about that version, including technical parameters (TR=1900ms, TE=18ms), functional description (optimized noise reduction technology, improved image quality), and custom description (created by Dr. Wang of this hospital for a brain scan project in February 2025). This application can support real-time version management of protocol packages, allowing users to easily view and select different versions of protocol packages, ensuring the accuracy and efficiency of the scanning process.

[0066] The scanning protocol can be one or more of the scanning protocols created by the user. It should be understood that the user currently operating the medical imaging system and the user creating the database can be the same person. Alternatively, the user currently operating the medical imaging system and the user creating the database can be different people. For example, the user creating the database could be a radiologist from a top-tier hospital, a researcher from a research institution, or a research and development personnel from a medical device manufacturer, while the user currently operating the medical imaging system could be a radiologist or doctor from a county-level hospital. The user currently operating the medical imaging system and the user creating the database can come from the same institution or different institutions.

[0067] The target scanning protocol is one of the scanning protocols in the database. Before controlling the scanner 202 to perform imaging scanning on the object to be detected based on the target scanning protocol, the processor 206 also executes the following: controlling the user interface of the interactive device 204 to display recommended target scanning protocols; and in response to the user's decision operation on the user interface, adding the target scanning protocol to the list of protocols to be scanned to determine the target scanning protocol. The user's decision operation on the user interface may be, for example, operating the mouse of the interactive device 204, touching the screen of the interactive device 204, or using the shortcut keys of the interactive device 204 to accept recommended information.

[0068] During magnetic resonance imaging (MRI) scanning, selecting an appropriate protocol package is crucial for ensuring image quality and diagnostic accuracy. This application embodiment utilizes a medical scan protocol management system, which can help users quickly and accurately select the appropriate protocol package. For example... Figure 3 The diagram illustrates a scanning control method for a medical imaging system according to an embodiment of this application, comprising the following steps:

[0069] Step S301: Obtain keyword instructions. Keyword instructions are obtained through natural language parsing of user input or through image-text parsing of the scanned image of the application form. The user enters keywords in the system's search module. Keywords can be scanned areas, scan types, clinical diagnoses, etc. In this embodiment, the user enters "brain T1-weighted imaging" or "brain scan" on the user interface of the interactive device 204. The medical imaging system will match these keywords against the text descriptions of protocol packets contained in the database of the medical scan protocol management system.

[0070] Natural language parsing of user-input commands can include processes such as identifying key entities and relationships, extracting scan-related keywords and parameters, image semantic disambiguation, and unstructured text parsing. The acquisition of scan-related keywords and parameters can be based on machine learning analysis of the vocabulary, grammar, and semantics of user commands to identify key information such as scan site, clinical department, scan type, patient information, and clinical purpose.

[0071] The image semantic disambiguation process in this embodiment includes:

[0072] First, an imaging standard library is established, comprising standard scanning protocol names and clinical imaging image representation keywords as the target set of the imaging standard library. This target set is used to disambiguate user-input text commands. The target set can be a correspondence between specific scanning protocols and abbreviations, such as FSE (Fast Spin Echo) or TSE (Turbo Spin Echo) corresponding to fast spin echo, T1DM corresponding to type 1 diabetes, T2DM corresponding to type 2 diabetes, T1WI corresponding to T1-weighted imaging, and T2WI corresponding to T2-weighted imaging; it can also be a correspondence between clinical imaging image representations and abbreviations, such as "lacunar infarction" corresponding to "lacunar stroke"; or it can be a correspondence between intelligent light shuttle imaging and fast scanning, uVision... TM Corresponding to intelligent positioning, easyscan TM It corresponds to a series of functional protocol combinations such as intelligent scanning.

[0073] Secondly, similarity calculation and mapping are performed. Specifically, natural language processing text embedding technology is used to convert the scanned keywords and parameters into vectors and perform similarity calculation. The cosine similarity between the text embedding vectors is calculated to determine whether the extracted different expressions point to the same target set.

[0074] Finally, the sliding window method is used to take into account the contextual information of the user's instructions. Based on the sentences before and after scanning related keywords and parameters, the relationship between keywords and parameters is extracted, and the target set is further filtered based on the relationship between keywords and parameters to obtain keyword instructions.

[0075] Step S303: Use keyword commands to search the database to obtain the target scanning protocol. The database contains multiple target scanning protocols with the same level and parallel relationship. The target scanning protocol contains multiple initial protocol packets and associated natural language descriptions. The initial protocol packets contain scanning protocol parameter values ​​and corresponding scanning parts. The natural language descriptions contain standardized descriptions or custom descriptions.

[0076] In this embodiment, the medical imaging system quickly matches matching protocol packages based on the input keywords and returns a list containing information such as the protocol package name, functional description, and usage scenario. For example, when the user enters "brain scan," the system returns all protocol packages related to brain scans and displays their functional descriptions, such as "suitable for T1-weighted brain imaging, suitable for clinical diagnosis" or "suitable for T2-weighted brain imaging, suitable for brain lesion detection." Users can click on each protocol package in the search results to view its detailed information, including technical parameters (such as TR, TE, sequence type) and custom descriptions (such as user name, creation date, patient type, etc.). This step helps users further confirm the applicability and accuracy of the protocol packages.

[0077] Step S305: Control the medical imaging system to perform an imaging scan on the object to be detected based on the scan protocol. The medical imaging system executes the scan protocol automatically or with manual user intervention. Optionally, if an abnormality is detected based on the partially scanned images during the execution of the scan protocol, the medical imaging system can automatically update the scan protocol parameter values. For example, if the partially acquired images indicate the location of a suspicious lesion at the scanned site, the scan protocol parameter values ​​can be updated by changing the scan area, changing the protocol type, or changing the scan angle. As another example, if the partially acquired images indicate that the scanned site is shifted due to respiratory motion, the scan protocol parameter values ​​can be updated by changing the protocol type to increase insensitivity to motion.

[0078] This application embodiment stores all magnetic resonance protocol packets in a single-level database, avoiding the complexity of traditional hierarchical structures and ensuring rapid retrieval of protocol packets. Users can directly search for protocol packets by entering text keywords, and the medical imaging system will quickly match and return relevant results based on the text description, improving retrieval efficiency. The medical imaging system adds a standardized text description to each protocol packet, covering key information such as the protocol's function, usage scenarios, and applicable equipment. Users can add personalized descriptions according to their own needs, further enhancing the information richness and flexibility of the protocol packets.

[0079] like Figure 4 The image shown is a scanning control method for a medical imaging system provided in another embodiment of this application:

[0080] Step S401: Determine if the patient is ready for examination. Patient preparation may include: removing metal objects worn by the patient, removing items that pose a safety hazard, administering a sedative to the patient beforehand, ensuring the medical imaging system's examination bed carries the patient into the central scanning area, and confirming that the patient is currently ready for scanning.

[0081] Step S403: Obtain the keyword command input by the user. The keyword command can be text information, voice information, etc. The doctor, acting as the user, logs into the medical scan protocol management system. The medical scan protocol management system includes a protocol manager, where the user can enter text in the search bar.

[0082] To recognize the diverse expressions from different users and hospitals, user text information is input into a natural language processing (NLP) model to obtain keyword commands. The acquisition of the NLP model involves three stages: selecting a base model, incremental pre-training, and post-training.

[0083] In the base model selection stage, a general large model is chosen as the basic framework.

[0084] In the incremental pre-training stage, medical knowledge is injected into the large model to address the gaps in its medical knowledge. A medical corpus is built using multi-source data, and a masked language model is used for self-supervised learning. This enables the large model to learn semantic associations such as "disease-alias", "scanning protocol-manufacturer-custom name", and "functional scan-manufacturer-custom feature name", thus solving the problem that the large model "cannot understand medical terminology".

[0085] In the post-training phase, supervised fine-tuning is performed using tens of thousands of "concept variant-standard expressions" (e.g., "DM→diabetes") annotated by medical experts. Multi-task learning is combined to optimize concept standardization and keyword extraction. Subsequently, keyword weights are adjusted based on user feedback (e.g., keywords preferred by doctors for diagnosis, keywords preferred by technicians for scanning protocols), achieving a complete parsing process from "domain knowledge injection → semantic alignment → preference adaptation." For example, "DM" is standardized to "diabetes," and "diabetic foot" is mapped to "diabetic foot," with weighted keywords output according to role requirements, improving the accuracy and personalization of medical text processing.

[0086] Step S405: A keyword command is used to search the database to determine if candidate scanning protocols exist. The database contains multiple scanning protocols with parallel relationships. Each scanning protocol contains multiple initial protocol packets and associated natural language descriptions. The initial protocol packet contains scanning protocol parameter values ​​and corresponding scanning sites. The natural language description includes standardized or custom descriptions. In this embodiment, the database protocol packets can include various types, such as: Type 1, a target scanning protocol containing only a first natural language description, suitable for standardized clinical scenarios; Type 2, a target scanning protocol containing only a second natural language description, suitable for scenarios with specific user needs; Type 3, a target scanning protocol containing both first and second natural language descriptions, containing both standardized and user-defined information, suitable for scenarios requiring a balance between standardization and personalization.

[0087] like Figure 5 The diagram shows the scanning protocol packages included in the database of this application embodiment. All scanning protocol packages have standardized descriptions such as scanning protocol package name, applicable body part, and positioning, and are arranged side-by-side at the same level. Some scanning protocol packages include custom descriptions. Specifically, HFP (head first-Prone) represents head-first, prone position; HFS (head first-Supine) represents head-first, supine position; FFP (Feet First-Prone) represents feet-first, prone position; and FFS (Feet First-Supine) represents feet-first, supine position. The corresponding scanning body part refers to the human body part to which the initial protocol package is applicable, such as the head, chest, abdomen, and limbs. In this embodiment, the initial protocol packages stored in the initial protocol package creation unit can cover all major scanning body parts. Each body part may have multiple initial protocol packages with different parameter settings to adapt to different clinical needs.

[0088] Custom descriptions can be annotations made by the current user or other users in their history. In this embodiment, the custom description of the "general" protocol packet is "Head plain scan protocol packet, used to scan patients with head-related diseases and complaints such as headache and dizziness"; the custom description of the "easyscan" protocol packet is "Head auto-localization plain scan protocol packet, which includes auto-localization functionality compared to the normal plain scan protocol packet"; the custom description of the "ACS2D" protocol packet is "Head ACS2D plain scan protocol packet, which adds ACS functionality compared to the normal plain scan protocol packet, ACS functionality..."; the custom description of the "ACS 3D" protocol packet is "Head ACS 3D plain scan protocol packet, which adds ACS functionality compared to the normal 3D scan protocol packet, ACS functionality..."; the custom description of the "Qscan" protocol packet is "Head silent plain scan protocol packet, which adds a silent function compared to the normal scan protocol packet, reducing noise interference...". Other protocols only have standardized descriptions and no custom descriptions.

[0089] like Figure 6 The diagram shows a database search result illustration of an embodiment of this application. Users can enter "loop" in the search bar of the protocol manager to filter for the "head_easyscout_Qscan" and "head_easyscout" protocol packets. The custom description of the "head_easyscout_Qscan" protocol packet is "Header automatically segmented mute protocol packet, which includes the function of automatically segmenting the preceding and following loops, and can be used during scanning...". The custom description of the "head_easyscout" protocol packet is "Header automatically segmented protocol packet, which includes the function of automatically segmenting the preceding and following loops, and can be used during scanning...". In addition to the above custom descriptions, each protocol packet also displays various information such as historical scan time and protocol storage path. The historical scan time for the "head_easyscout_Qscan" protocol packet is 23 seconds, and the historical scan time for the "head_easyscout" protocol packet is 13 seconds.

[0090] By double-clicking or right-clicking in the protocol package fields, you can view the scan protocol parameter values. These parameters may include, but are not limited to, the following: scan time, scan angle, scan slice thickness, scan interval, scan voltage, scan current, scan dose, and scan sequence type. These parameter values ​​vary depending on the scanning equipment and the purpose of the scan. For example, for CT scanning equipment, scan protocol parameter values ​​may include tube voltage (e.g., 120kV), tube current (e.g., 250mA), slice thickness (e.g., 5mm), and scan time (e.g., 0.5s / revolution); for MRI scanning equipment, scan protocol parameter values ​​may include TR value (e.g., 2000ms), TE value (e.g., 30ms), slice thickness (e.g., 3mm), and FOV (e.g., 240mm×240mm).

[0091] In step S407, in response to the existence of candidate scanning protocols, the protocol to be scanned is obtained according to the candidate scanning protocols, and imaging scanning is performed. The user selects the most suitable protocol package for scanning based on the returned description information and recommendation list. During the selection process, the medical imaging system can provide a preview function, allowing the user to preview the specific settings of the protocol package before selection to ensure it meets actual needs.

[0092] Step S409: In response to the absence of a candidate scanning protocol, determine whether an initial protocol packet exists.

[0093] Step S411: In response to the existence of an initial protocol packet, add a custom description to the initial protocol packet, generate the protocol to be scanned, and perform imaging scanning.

[0094] In step S413, in response to the absence of an initial protocol packet, the user creates an initial protocol packet and adds a custom description, generates the protocol to be scanned, and performs an imaging scan.

[0095] In this embodiment, the protocol package of the protocol manager has a customizable text description, making the content intuitive and helping users quickly understand the core functions and scope of application. Users can accurately search for protocol packages using keywords and select the most suitable protocol package based on the patient's condition. Multiple protocol packages are managed at the same level, avoiding complex tree structures. Keyword search or frequently used options can be set to improve search efficiency. User-defined annotation and editing functions are supported, allowing different venues and users to quickly understand the functions of the protocol packages, with strong compatibility and scalability.

[0096] The embodiments of this application can be used for recommending scanning protocols, saving technicians from the complex process of selecting scanning protocols. For example... Figure 7 The image shown illustrates a recommended scanning protocol method for a medical imaging system in one embodiment of this application:

[0097] Step S701: Obtain the keyword instruction input by the user. The keyword instruction can be obtained manually: the user performs an input operation on the user interface to obtain the user instruction; the user instruction is processed by a natural language processing model to obtain the keyword instruction. Alternatively, the keyword instruction can be obtained by scanning an application form: the user scans the application form image; the scanned image is converted into text information; the text information is processed by a natural language processing model to obtain the keyword instruction. Or, the keyword instruction can be obtained through the RIS system: the RIS system contains patient registration information and clinician's medical orders; the patient registration information and clinician's medical orders are processed by a natural language processing model to obtain the keyword instruction.

[0098] In this embodiment, such as Figure 8As shown, the information in the scanned image of the application form exists in image form. It needs to be recognized by an Optical Character Recognition (OCR) algorithm to identify the text, symbols, and tables in the image and convert them into structured text data. Optionally, the OCR algorithm can be based on traditional image processing techniques or modern OCR algorithms based on deep learning, such as Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs). By introducing image recognition functionality, the medical imaging system of this embodiment can directly process paper or electronic examination application forms, automatically extracting key information and reducing the workload and error rate of manual input. This approach not only improves the system's efficiency but also enhances the accuracy and completeness of information extraction, helping the system to more accurately select the appropriate scanning protocol.

[0099] Optionally, the user interface can provide multiple input methods, such as text input boxes, voice input, and preset template selection, allowing users to express their needs in the most natural and convenient way. User input can be short keywords or detailed descriptive text.

[0100] User-inputted natural language commands typically contain a wealth of information and context, requiring the extraction of key information points through a natural language processing (NLP) model. The NLP model analyzes the semantic structure of the command, identifies key entities and relationships, and extracts scan-related keywords and parameters. In this embodiment, the NLP model can be machine learning-based, analyzing the vocabulary, grammar, and semantics of the user command to identify key information such as the scan site ("head"), clinical department (endocrinology and metabolism), patient information ("diabetes, grade 3 hypertension (very high risk), 63 years old"), and clinical purpose ("lacunar infarction"), and organizing this information into structured keyword commands. By introducing natural language processing capabilities, the medical imaging system in this embodiment can understand and process the user's natural language input, enabling the user to express their scanning needs in a more natural and intuitive way. This approach not only improves the system's usability and user-friendliness but also enhances the accuracy and completeness of the expressed needs, helping the system to more accurately select the appropriate scanning protocol. In this embodiment, addressing the core pain points of general-purpose models lacking medical domain expertise and being unable to accurately identify variations in medical concepts and exhibiting significant semantic comprehension biases when directly applied, the construction of the natural language processing model adopts a domain knowledge enhancement approach. like Figure 9 The diagram shown is a schematic representation of the natural language parsing model acquisition method according to an embodiment of this application. The process of acquiring the natural language parsing model may include:

[0101] In the base model selection stage, a general large model is chosen as the basic framework.

[0102] In the incremental pre-training stage, medical knowledge is injected into the large model to address the gaps in its medical knowledge. A medical corpus is built using multi-source data, and a masked language model is used for self-supervised learning. This enables the large model to learn semantic associations such as "disease-alias", "scanning protocol-manufacturer-custom name", and "functional scan-manufacturer-custom platform name", thus solving the problem that the large model "cannot understand medical terminology".

[0103] The post-training phase includes two processes: supervised fine-tuning and direct preference optimization. In supervised fine-tuning, 100,000 "concept variant-standard expressions" (e.g., "DM→diabetes") annotated by medical experts are used to fine-tune the text, combining multi-task learning to optimize concept standardization and keyword extraction. In direct preference optimization, keyword weights are adjusted based on user feedback (e.g., keywords for doctors' diagnoses, keywords for technicians' scanning protocols), achieving a complete parsing process from "domain knowledge injection → semantic alignment → preference adaptation." Weighted keywords are output according to user needs, improving the accuracy and personalization of medical text processing.

[0104] This application's embodiments address the gaps in medical knowledge through incremental pre-training, combined with supervised learning and preference optimization, systematically solving the problems of "knowledge blind spots" and "insufficient adaptability" of general-purpose large models in the medical field.

[0105] Step S703: Provide a protocol manager. The database of the protocol manager contains protocol packets and text descriptions. The database of the protocol manager contains multiple scanning protocols (protocol packets), which are listed at the same level. Each scanning protocol contains multiple initial protocol packets and associated natural language descriptions. The initial protocol packets contain scanning protocol parameter values ​​and corresponding scanning locations. The natural language descriptions include standardized descriptions or custom descriptions.

[0106] Custom descriptions may include at least one of the following: user name, creation date, and patient type. For example, the user name refers to the name or identifier of the user who created or modified the protocol. Recording the user name helps track the origin and responsible party of the protocol, facilitating subsequent management and communication. The creation date refers to the date the protocol was created or last modified, helping to understand its timeliness and update status, and facilitating the determination of whether the protocol needs updating or verification. The patient type refers to the characteristics of the patient group to which the protocol applies, such as "elderly patients," "child patients," or "pregnant women." Different types of patients may require different scanning parameters and precautions; recording the patient type helps in selecting the most suitable protocol for a specific patient. Custom descriptions may also include the protocol's functionality and applicable scanning scenarios. Figure 8As shown, for a plain scan protocol packet, the custom description is "This protocol packet is a plain scan protocol packet, containing T1, T2, DWI, and FLAIR sequences, using acceleration and automatic localization technology, suitable for patients with head or neurological diseases." For an MRA (MR angiography) protocol packet, the custom description is "This protocol packet is an MRA protocol packet, containing TOF (Time of Flight) sequences, using ACS (Artificial Intelligence Compressed Sensing) acceleration technology, suitable for patients with cerebrovascular diseases and aneurysms, with embedded automatic localization and automatic post-processing in the sequence packet."

[0107] Associations can be established between different scanning protocols in the database: scanning protocols can share parameters or have functional dependencies, based on factors such as the same scanning area, similar scanning purpose, or the same scanner model. Associated target scanning protocols typically share some parameter settings but also have their own unique parameters and functions. Setting associations reduces parameter duplication and redundancy, improving system efficiency and consistency; it also enables batch updates and maintenance of protocols, allowing simultaneous modification of common parameters across multiple associated protocols. Associated scanning protocols maintain their individual integrity and functional characteristics while forming an organic protocol network through shared parameters and association markers, facilitating user selection and use.

[0108] Step S705: Based on the keyword instruction, determine whether a protocol packet mapping exists in the protocol manager.

[0109] In step S707, in response to the existence of a protocol packet that satisfies the mapping rules, the protocol packet that satisfies the mapping rules is selected as the protocol to be scanned, and the medical imaging system is controlled to perform the scan. Optionally, the medical imaging system can automatically execute the scanning process without any manual operation by the user. To ensure the safety of medical scanning, the user can select one or more target scanning protocols and add them to the list to be scanned by clicking, dragging, checking, etc., and the medical imaging system will automatically perform the scan.

[0110] Of course, users can also adjust the order of protocols in the list, set protocol priorities or execution conditions, etc. Based on these user operations, the protocol manager updates the list to be scanned, and after user confirmation, uses the protocols in the list as the final protocols to be scanned, controlling the scanner to perform the scan. By introducing user interaction functions, the medical imaging system in this embodiment, while maintaining automation and intelligence, also gives users sufficient control and decision-making power. Users can select the scanning protocol most suitable for the current detection object and clinical needs based on their professional judgment and clinical experience, improving the targeting and effectiveness of the scan. At the same time, the system's recommendation function also provides users with references and suggestions, reducing the user's decision-making burden and improving work efficiency.

[0111] Step S709: In response to the absence of a protocol packet satisfying the mapping rule, a recommended protocol packet is generated. After user confirmation, the recommended protocol packet allows the medical imaging system to perform a scan. The most suitable protocol packet can be automatically recommended based on the user's usage history and preferences. For example, if a user frequently uses a certain type of brain scan protocol packet, the system will prioritize recommending these protocol packets after the user enters keywords. For instance, in the MRI department of a hospital, Dr. Zhang frequently undergoes brain T1-weighted imaging and cardiac imaging scans. The medical imaging system records the frequency of Dr. Zhang's historical scan protocol usage. In the presence of multiple recommended protocols, the system automatically recommends frequently used protocol packets based on the frequency of historical scan protocols. Figure 8 As shown, the recommended protocol packets are the plain scan packet + MRA packet, and the user adds the applicable scanning scenario to the recommended protocol packets this time.

[0112] like Figure 10 The diagram shows a recommended protocol package displayed on the user interface of this application embodiment. The user interface includes patient information keywords, through which users can learn about the patient's basic information, condition information, and clinical department information. The recommended protocol package includes the package name, its clinical scanning functions, application scenarios, and technical details involved in the scanning process. After clicking the confirmation button on the user interface, the user can perform a one-click scan on the current patient.

[0113] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0114] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for various specific applications, but such implementations should not be considered beyond the scope of this application.

[0115] In the embodiments provided in this application, it should be understood that the disclosed apparatus / device and method can be implemented in other ways. For example, the apparatus / device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0116] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0117] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0118] If an integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0119] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A medical scanning protocol management system, characterized in that, The medical scanning protocol management system includes: An initial protocol packet creation unit contains multiple initial protocol packets, each of which contains scan protocol parameter values ​​and a corresponding scan region; A parsing unit is configured to parse the initial protocol packet to generate a first natural language description, wherein the first natural language description is a standardized description; A custom unit is configured to, in response to user input, set a second natural language description for at least one of the initial protocol packets, wherein the second natural language description is a custom description; The target protocol generation unit associates the first natural language description and / or the second natural language description with the initial protocol packet to generate multiple target scanning protocols, wherein the multiple target scanning protocols are parallel to each other at the same level.

2. The medical scanning protocol management system according to claim 1, characterized in that, The first natural language description includes at least one of the following: the protocol type of the initial protocol packet, the scanner model, and the corresponding clinical diagnostic disease.

3. The medical scanning protocol management system according to claim 1, characterized in that, The second natural language description includes at least one of the following: user name, creation date, and patient type.

4. The medical scanning protocol management system according to any one of claims 1-3, characterized in that, At least two of the target scanning protocols contain the same scanning protocol parameter value, and the two target scanning protocols are associated with each other.

5. The medical scanning protocol management system according to claim 1, characterized in that, The medical scanning protocol management system also includes: An import unit is used to import an external data source, which is used to update the scanning protocol parameter values ​​of the target scanning protocol.

6. A medical imaging system, characterized in that, The medical imaging system includes: A scanner is used to perform imaging scans on an object to obtain imaging data; A processor for reconstructing the imaging data; An interactive device, which is communicatively coupled to the processor, includes a user interface. The processor is configured with executable instructions, which, when executed, cause the processor to: Get keyword instructions; The database is searched according to the keyword instruction to obtain the protocol to be scanned. The database contains multiple target scanning protocols with the same level and parallel relationship. The target scanning protocol contains multiple initial protocol packets and associated natural language descriptions. The initial protocol packets contain scanning protocol parameter values ​​and corresponding scanning parts. The natural language descriptions include standardized descriptions or custom descriptions. The scanner is controlled to execute the scanning protocol to perform an imaging scan on the object to be detected.

7. The medical imaging system according to claim 6, characterized in that, Before controlling the scanner to perform imaging scanning on the object to be detected based on the scanning protocol, the processor also performs: Control the user interface to display at least one recommended target scanning protocol; In response to the user's decision operation on the user interface, the target scanning protocol is added to the scan list to determine the scan protocol.

8. The medical imaging system according to claim 7, characterized in that, The keyword instruction is obtained in the following way: In response to user input on the user interface, user instructions are obtained; The user command is input into a natural language parsing model to obtain the keyword command; or, Obtain a scanned image of the application form; Convert the scanned image of the application form into text information; The text information is input into a natural language processing model to obtain the keyword instruction.

9. The medical scanning protocol management system according to claim 7, characterized in that, The multiple target scanning protocols are relatively independent of each other; or, At least two of the target scanning protocols contain the same scanning protocol parameter value, and the two target scanning protocols are associated with each other.

10. A scanning control method for a medical imaging system, characterized in that, The scanning control method includes: Get keyword instructions; The database is searched using the keyword command to obtain the target scanning protocol. The database contains multiple target scanning protocols with the same level and parallel relationship. Each target scanning protocol contains multiple initial protocol packets and associated natural language descriptions. The initial protocol packets contain scanning protocol parameter values ​​and corresponding scanning parts. The natural language descriptions include standardized descriptions or custom descriptions. The medical imaging system is controlled based on the scanning protocol to perform an imaging scan on the object to be detected.

Images