Medical equipment knowledge interaction method and device, computer equipment and storage medium

Abstract

The invention relates to a medical equipment knowledge interaction method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring user interaction information and a user portrait; screening a target knowledge set matched with the user portrait from a medical equipment knowledge set according to the user portrait; and screening target knowledge matched with user interaction information from the target knowledge set according to the user interaction information, and displaying the target knowledge. By the adoption of the method, stable training content output can be achieved, personalized target knowledge display can be provided for different users, and the effect of improving the matching degree of interaction content and specific crowds is achieved.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a medical device knowledge interaction method, apparatus, computer equipment and storage medium. Background Art

[0002] With the continuous iteration of medical equipment, the operation knowledge of medical equipment also needs to be learned in time so that the medical equipment can be put into use as soon as possible. Nowadays, operators often need to learn the relevant knowledge of medical equipment through face-to-face training.

[0003] However, the existing medical equipment training process has the following problems:

[0004] (1) Complex parameter settings: Taking magnetic resonance imaging equipment as an example, a large number of scanning parameters and control parameters need to be adjusted, and there are linkages between many different types of parameters, which makes the actual operation process complicated after the user is trained. At the same time, users tend to lose memory and forget information over time. There are low-frequency workflows in the training process, which may not be used for a long time after training, resulting in easy forgetting of operation details.

[0005] (2) High learning costs caused by version function updates: Medical scanning equipment and its software are frequently updated and upgraded. Each update may bring new parameters and workflows, which users need to relearn. Real-time problem solving requires rapid response. During the actual scanning process, users may encounter various unexpected problems, such as equipment failure and image quality problems. However, unified training alone cannot fully cover these problems encountered during actual operation. Improper operation may affect the safety of patients and operators.

[0006] (3) Users have personalized needs: Different hospitals and clinics may have different scanning requirements, and standard training courses may not meet all personalized needs.

[0007] It can be seen from this that the training content of face-to-face training is not stable and has a low degree of personalization. It may also lead to uneven training results among operators due to different learning progress and memory levels of different operators.

[0008] In summary, current medical device knowledge training still has the problem of low matching between interactive content and specific populations. Summary of the invention

[0009] Based on this, it is necessary to provide a medical equipment knowledge interaction method, device, computer equipment and storage medium that can improve the matching degree between the interactive content and a specific group of people in response to the above technical problems.

[0010] In a first aspect, this embodiment provides a medical device knowledge interaction method, the method comprising:

[0011] Obtain user interaction information and user portraits;

[0012] Filtering a target knowledge set matching the user profile from a medical device knowledge set according to the user profile;

[0013] According to the user interaction information, target knowledge matching the user interaction information is filtered from the target knowledge set, and the target knowledge is displayed.

[0014] In one embodiment, the obtaining of the user interaction information and the user portrait includes:

[0015] Obtaining the user's account history information; the account history information includes user identification, historical question and answer information, and historical learning records;

[0016] The user portrait is determined based on the user's account history information; the user portrait includes skill level information, interest areas and learning progress.

[0017] In one embodiment, determining the user portrait based on the user's account history information includes:

[0018] Obtaining user seniority information from the user identifier;

[0019] Obtaining answer text and preset standard text in the historical question and answer information;

[0020] Calculating the similarity between the answer text and the preset standard text;

[0021] Based on the user's seniority information and / or similarity, skill level information of the user is determined.

[0022] In one embodiment, selecting a target knowledge set matching the user profile from a medical device knowledge set according to the user profile includes:

[0023] Each piece of knowledge in the medical equipment knowledge set is correspondingly provided with level information;

[0024] Get the skill level information in the user profile;

[0025] According to the skill level information and the grade information, a target knowledge set matching the skill level information is selected from the medical equipment knowledge set.

[0026] In one embodiment, selecting target knowledge matching the user interaction information from the target knowledge set according to the user interaction information includes:

[0027] Extracting key fields from the user interaction information;

[0028] Performing vectorization processing on the key field to obtain a first vectorized feature of the user interaction information;

[0029] Performing vectorization processing on each piece of knowledge in the target knowledge set to obtain a second vectorized feature corresponding to each piece of knowledge;

[0030] Calculate similarity between the first vectorized feature and the plurality of the second vectorized features;

[0031] The knowledge corresponding to the second vectorized feature with the highest similarity is determined as the target knowledge.

[0032] In one embodiment, presenting the target knowledge includes:

[0033] Extracting key fields from the target knowledge;

[0034] Performing text expansion on the key fields based on the user interaction information to obtain target display information after natural language processing;

[0035] The target display information is displayed.

[0036] In one embodiment, the method further comprises:

[0037] Acquire the interaction time of the user interaction information;

[0038] If there is historical interaction information within the preset time range of the interaction time, obtaining the historical interaction information;

[0039] Filtering a target knowledge set matching the user profile from a medical device knowledge set according to the user profile;

[0040] According to the user interaction information and the historical interaction information, target knowledge matching the user interaction information is filtered from the target knowledge set, and the target knowledge is displayed.

[0041] In a second aspect, this embodiment provides a medical device knowledge interaction device, the device comprising:

[0042] Acquisition module, used to obtain user interaction information and user portraits;

[0043] A screening module, used for screening a target knowledge set matching the user portrait from the medical device knowledge set according to the user portrait;

[0044] The display module is used to filter target knowledge matching the user interaction information from the target knowledge set according to the user interaction information, and display the target knowledge.

[0045] In a third aspect, this embodiment provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0046] In a fourth aspect, this embodiment provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of the above method when executed by a processor.

[0047] The above-mentioned medical device knowledge interaction method, apparatus, computer equipment and storage medium obtain user interaction information and user portraits; filter the target knowledge set matching the user portrait from the medical device knowledge set according to the user portrait; filter the target knowledge matching the user interaction information from the target knowledge set according to the user interaction information, and display the target knowledge. Through the setting of the medical device knowledge set, stable training content output can be achieved. According to the user interaction information and user portraits, personalized target knowledge display can be provided to different users, thereby achieving the effect of improving the matching degree between the interactive content and a specific group of people. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 An application environment diagram of a medical device knowledge interaction method in one embodiment;

[0049] Figure 2 A schematic diagram of a process of a medical device knowledge interaction method in one embodiment;

[0050] Figure 3 A schematic diagram of a flow chart of a medical device knowledge interaction method in another embodiment;

[0051] Figure 4 is a structural block diagram of a medical equipment knowledge interaction device in one embodiment;

[0052] Figure 5 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0054] The medical device knowledge interaction method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other network servers. User interaction information and user portraits can be stored in the server 104 or the data storage system. The terminal 102 obtains user interaction information and user portraits from the server 104; according to the user portrait, the target knowledge set matching the user portrait is screened from the medical device knowledge set, and according to the user interaction information, the target knowledge matching the user interaction information is screened from the target knowledge set, and the target knowledge is displayed. Among them, the terminal 102 can be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, Internet of Things devices and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart car-mounted devices, etc. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The server 104 can be implemented with an independent server or a server cluster consisting of multiple servers. By deploying in the cloud or deploying separately, and supporting the above-mentioned multiple terminals, users can learn anytime and anywhere, and the server automatically records the learning progress. Users can synchronize learning content on different devices, thereby improving user learning efficiency.

[0055] In one embodiment, Figure 2 As shown, a medical device knowledge interaction method is provided, and the method is applied to Figure 1 The terminal 102 in the example is used as an example to illustrate, and the following steps are included:

[0056] Step S100, obtaining user interaction information and user portrait.

[0057] The user interaction information may be input information input by the user to the terminal 102. The input information may be questions or keywords input by the user for knowledge in a specific field that the user wishes to learn, and may be input in the form of natural language. Furthermore, the input information may be obtained based on text, image recognition, voice recognition, or other methods based on user behavior recognition.

[0058] User portraits can be user types obtained based on user account information and user behavior record analysis, and the number and name of types can be set based on actual needs. In order to provide users with personalized medical device knowledge, user portraits can be used to characterize the user's mastery of medical device knowledge or the efficiency of receiving new medical device knowledge.

[0059] Step S200, screening a target knowledge set matching the user portrait from a medical device knowledge set according to the user portrait.

[0060] Among them, it can be understood that the medical device knowledge set can be a plurality of medical device knowledge, wherein each medical device corresponds to one or more user portraits, and screening the target knowledge set matching the user portrait can be by selecting a plurality of medical device knowledge corresponding to the user portrait in the medical device knowledge set, and combining them to obtain the target knowledge set; there can also be multiple medical device knowledge sets, and screening the target knowledge set matching the user portrait can be by selecting medical device knowledge from multiple medical device knowledge sets; in addition, each medical device knowledge set can also correspond to one or more user portraits, and screening the target knowledge set matching the user portrait can also be by screening out one or more medical device knowledge sets corresponding to the user portrait.

[0061] The above-mentioned medical equipment knowledge set may be updated in real time, including but not limited to updates of medical scanning equipment and its software, corresponding new parameters, new workflows and usage tutorials, etc.

[0062] Step S300, according to the user interaction information, target knowledge matching the user interaction information is screened from the target knowledge set, and the target knowledge is displayed.

[0063] The target knowledge may be the target knowledge that matches the user interaction information. The target knowledge that matches the user interaction information may be selected by calculating the similarity of the text in the user interaction information with the text in the multiple medical device knowledge in the target knowledge set one by one, so as to determine the target knowledge that best matches the user interaction information. Furthermore, the attention mechanism may be used to select elements in the user interaction information and the medical device knowledge for similarity comparison, so as to further improve the accuracy of the target knowledge matching.

[0064] The target knowledge may be displayed through interactive devices such as display devices, audio and video playback devices, etc., to achieve interaction with users.

[0065] The present embodiment provides a medical device knowledge interaction method, which obtains user interaction information and user portraits; filters a target knowledge set matching the user portrait from a medical device knowledge set according to the user portrait; filters target knowledge matching the user interaction information from the target knowledge set according to the user interaction information, and displays the target knowledge. By setting the medical device knowledge set, stable training content output can be achieved. By providing personalized target knowledge display to different users based on the user interaction information and user portraits, the matching degree between the interactive content and a specific group of people can be improved.

[0066] In one of the embodiments, after obtaining the user interaction information and the user portrait, the step further includes: performing data cleaning on the user interaction information.

[0067] It is understandable that user interaction information may include incorrect punctuation, confusion between uppercase and lowercase input, repeated words in voice input, etc. Therefore, data cleaning of user interaction information can be performed, such as removing punctuation, converting to lowercase, etc., to achieve data optimization of user interaction information.

[0068] In one embodiment, the obtaining of the user interaction information and the user portrait includes:

[0069] Acquire the user's account history information; determine the user portrait based on the user's account history information; the user portrait includes skill level information, interest areas and learning progress.

[0070] Among them, the account history information includes user identification, historical question and answer information and historical learning records. The user identification can be the user's identity information, such as an identity code, etc., and can also include account information set by the user, such as the start time of learning, the total learning time, educational background, position, etc. The historical question and answer information can be the answers given by the user to a specific question, which is used to determine the user's understanding or mastery of the field knowledge corresponding to the specific question. The historical learning record can be a learning record generated when the user accesses the learning resources in the relevant field through the account corresponding to the user identification. The historical learning record can include access records of learning resources, and can also include the learning situation of each learning resource. In a specific embodiment, if the learning resource is an article page, the learning situation can be the browsing progress of the article page, and if the learning resource is a teaching video, the learning situation can be the playback progress of the teaching video.

[0071] The user portrait is determined based on the historical account information of the user, and the user portrait may be generated according to the user identifier, historical question and answer information, and historical learning records in the historical account information.

[0072] The skill level information may be obtained by setting different weight values ​​for the user identification, historical question and answer information, and historical learning records, and obtaining corresponding scores for the user identification, historical question and answer information, and historical learning records in turn according to preset rules. Based on the scores and the corresponding weight values, the characteristic value corresponding to the skill level information may be calculated, and the characteristic value may be used as the skill level information of the user. Furthermore, the skill level grade corresponding to the characteristic value may be determined according to whether the characteristic value falls into multiple preset skill level characteristic ranges, and the skill level grade may be used as the skill level information of the user.

[0073] The interest field can be a method of setting different weight values ​​for the user identification, historical question and answer information, and historical learning records. It can be a method of determining the user's self-set interest field based on the user identification, taking the field corresponding to the answer with a higher correct rate in the historical question and answer information as the interest field, and determining the interest field from large to small according to the learning resources accessed by the user's historical learning record. Based on the interest field information obtained from the above user identification, historical question and answer information, and historical learning records, and the corresponding weight values, comprehensive interest field information can be determined.

[0074] The learning progress can be achieved by setting different weight values ​​for user identification, historical question and answer information, and historical learning records. Specifically, the weight value of the historical learning record can be set to a higher weight. Among them, the approximate learning progress of the user can be determined based on the user identification information, and the approximate learning progress can be expressed as a larger threshold range; further learning progress can be calculated based on the accuracy of answers to questions in different fields in the historical question and answer information; or the learning progress can be calculated based on the learning situation corresponding to multiple learning resources in the historical learning records, as well as the proportion of multiple learning resources in all resources on the platform.

[0075] A medical device knowledge interaction method provided in this embodiment determines the user portrait based on the user's account history information; the user portrait includes skill level information, interest areas and learning progress, which can realize analysis and extraction of the user portrait. By obtaining the skill level information, interest areas and learning progress, it is possible to better screen the target knowledge set that matches the user, realize targeted medical device knowledge training, and achieve the effect of improving the matching degree of the interactive content with a specific group of people.

[0076] In one embodiment, determining the user portrait based on the user's account history information includes:

[0077] Obtaining user seniority information from the user identifier;

[0078] Obtaining answer text and preset standard text in the historical question and answer information;

[0079] Calculating the similarity between the answer text and the preset standard text;

[0080] Based on the user's seniority information and / or similarity, skill level information of the user is determined.

[0081] The user's seniority information may be the user's working hours, which may be provided by the user or obtained with the user's authorization. The user's skill level information is determined based on the user's seniority information, which may be determined based on the user's working hours.

[0082] The historical question and answer information includes the answer text, which may be the answer result given by the user based on a specific question. The similarity between the answer text and the preset standard text may be calculated by calculating the similarity between the answer result and the preset standard text corresponding to the specific question, i.e., the expected answer, to further determine the user's mastery of medical device knowledge. When there are multiple specific questions, the similarities between the user's answer texts to multiple specific questions and the corresponding preset standard texts may be combined to obtain a comprehensive similarity as one of the reference indicators for skill level information. Determining the user's skill level information based on the similarity may be determining the user's corresponding skill level information based on the interval in which the similarity value is located.

[0083] Based on the user's seniority information and similarity, the user's skill level information is determined. This can be done by vectorizing the user's seniority information and the similarity value, and determining the corresponding skill level information based on the vectorized features; or it can be done by multiplying the skill level information obtained respectively from the user's seniority information and the similarity with a preset weight value to obtain the final skill level information.

[0084] A medical device knowledge interaction method provided in this embodiment can make a more accurate judgment on the user's skill level by determining the user's skill level information based on the user's seniority information and / or similarity, thereby obtaining a more appropriate target knowledge set when screening medical device knowledge, thereby achieving the effect of improving the matching degree of the interactive content with a specific group of people.

[0085] In one embodiment, selecting a target knowledge set matching the user profile from a medical device knowledge set according to the user profile includes:

[0086] Each piece of knowledge in the medical equipment knowledge set is correspondingly provided with level information;

[0087] Get the skill level information in the user profile;

[0088] According to the skill level information and the grade information, a target knowledge set matching the skill level information is selected from the medical equipment knowledge set.

[0089] The grade information may correspond to the skill level information. For example, the grades may be divided into level one, level two and level three, corresponding to the novice, experienced user and technical expert in the skill level information, respectively. The grade information may be a label for the knowledge in the medical device knowledge set.

[0090] According to the skill level information and the grade information, a target knowledge set matching the skill level information is selected from the medical equipment knowledge set. This may be to select medical equipment knowledge in the medical equipment knowledge set whose grade information matches the skill level information to obtain a target knowledge set.

[0091] The present embodiment provides a medical device knowledge interaction method, which selects a target knowledge set that matches the skill level information from the medical device knowledge set according to the skill level information and the grade information, so that the medical device knowledge is matched with the user's skill level. This can achieve personalized medical device knowledge display based on the user's skill level, thereby achieving the effect of improving the matching degree of the interactive content with a specific group of people.

[0092] In one embodiment, selecting target knowledge matching the user interaction information from the target knowledge set according to the user interaction information includes:

[0093] Extracting key fields from the user interaction information;

[0094] Performing vectorization processing on the key field to obtain a first vectorized feature of the user interaction information;

[0095] Performing vectorization processing on each piece of knowledge in the target knowledge set to obtain a second vectorized feature corresponding to each piece of knowledge;

[0096] Calculate similarity between the first vectorized feature and the plurality of the second vectorized features;

[0097] The knowledge corresponding to the second vectorized feature with the highest similarity is determined as the target knowledge.

[0098] Among them, the key field can be a keyword or a key sentence in the user interaction information. The capture of the key field can be keyword extraction based on statistical features, keyword extraction based on a word graph model, or keyword extraction based on a topic model. In a specific embodiment, the user interaction information can be pre-processed by word segmentation to obtain the word segmentation result, and the frequency of occurrence of each word in the user interaction information or a preset corpus is calculated, and the words with lower frequency of occurrence, or the sentences composed of the above words and surrounding words, are used as key fields.

[0099] The key fields are vectorized by methods such as One-Hot Encoding or Word Embedding. Each piece of knowledge in the target knowledge set is vectorized by referring to the same vectorization method as the key fields to obtain a second vectorized feature corresponding to each piece of knowledge.

[0100] The similarity calculation between the first vectorized feature and the plurality of the second vectorized features may be performed by sequentially calculating the Euclidean distance, the Manhattan distance or the Chebyshev distance between the first vectorized feature and the second vectorized features, or by using other available similarity calculation methods.

[0101] It can be understood that the more the second vectorized feature with the highest similarity is calculated, the greater the possibility that the knowledge corresponding to it is the expected reply result of the user interaction information. Therefore, the knowledge corresponding to the second vectorized feature with the highest similarity can be determined as the target knowledge.

[0102] A medical device knowledge interaction method provided in this embodiment extracts key fields from the user interaction information; vectorizes the key fields to obtain a first vectorized feature of the user interaction information; vectorizes each knowledge in the target knowledge set to obtain a second vectorized feature corresponding to each knowledge; calculates the similarity between the first vectorized feature and multiple second vectorized features; and determines the knowledge corresponding to the second vectorized feature with the highest similarity as the target knowledge. This method can achieve the acquisition of corresponding target knowledge based on user interaction information, improve the accurate provision of target knowledge, and achieve the effect of improving the matching degree between interactive content and a specific group of people.

[0103] In one embodiment, the step of selecting a target knowledge set matching the user portrait from a medical device knowledge set according to the user portrait; selecting target knowledge matching the user interaction information from the target knowledge set according to the user interaction information, and displaying the target knowledge comprises:

[0104] Initialize the database. The initialization may include initialization of magnetic resonance-related parameters, workflows, safety specifications and other data. The magnetic resonance-related parameters, workflows, safety specifications and other data may be obtained based on matching with the user profile, thereby initializing a target knowledge set that matches the user profile.

[0105] Perform natural language understanding on user interaction information. For example, text classification or entity recognition can be performed using, for example, a BERT model or other pre-trained language model. The BERT model refers to a language representation model, representing a bidirectional encoder representation from a Transformer.

[0106] Contextual short-term memory based on user interaction information. For example, the Transformer model can be used to store and encode historical interaction information between the user and the system. Each question and answer pair is regarded as a sequence unit, and these units are input into the Transformer model in chronological order. The Transformer model encodes this information and generates a context vector. When there is new input, the context vector can be used to generate a more accurate and relevant response.

[0107] Query the database items that are related to the user interaction information but not completely matched. Among them, it can be an algorithm that combines TF-IDF (term frequency-inverse document frequency) with cosine similarity. Specifically, each item in the database can be converted into a vector using the TF-IDF algorithm, that is, when a query is received, it is also converted into a vector using TF-IDF; then, the cosine similarity algorithm is used to find the database vector that is most similar to the query vector, and specifically the similarity between the two vectors is evaluated by calculating the cosine value of the angle between the two vectors.

[0108] The database items are sorted and screened. Each database item related to the query may be weighted according to a set weight; the item with the highest weighted score will be returned first. The set weight may be, for example, accuracy, relevance, novelty, etc.

[0109] The target knowledge can be displayed by combining the context vector of the Transformer model and the above search results to generate the final response.

[0110] In one embodiment, presenting the target knowledge includes:

[0111] Extracting key fields from the target knowledge;

[0112] Performing text expansion on the key fields based on the user interaction information to obtain target display information after natural language processing;

[0113] The target display information is displayed.

[0114] It is understandable that when the user interaction information is a question raised by the user, the focus of the question may be different, and the target knowledge stored in the medical device knowledge set is often in the original format, so it is necessary to perform text expansion for the question raised by the user. Exemplarily, the text expansion of key fields based on user interaction information can be based on the question and answer scenario of the user interaction information and the corresponding preset answer text, and input into the pre-trained dialogue generation model, and through one or more text processing methods such as sentence rewriting and synonym conversion, the expanded text of the key field is obtained as the target display information.

[0115] A medical device knowledge interaction method provided in this embodiment can convert target knowledge into answer text for user interaction information by text expansion of key fields in the target knowledge, making the target knowledge easier to understand and achieving the effect of improving the matching degree of the interaction content with a specific group of people.

[0116] In one embodiment, the method further comprises:

[0117] Acquire the interaction time of the user interaction information;

[0118] If there is historical interaction information within the preset time range of the interaction time, obtaining the historical interaction information;

[0119] Filtering a target knowledge set matching the user profile from a medical device knowledge set according to the user profile;

[0120] According to the user interaction information and the historical interaction information, target knowledge matching the user interaction information is filtered from the target knowledge set, and the target knowledge is displayed.

[0121] The interaction time may be the time when the user interaction information is input into the terminal. The preset time range of the interaction time may be a preset time interval starting from the interaction time and extending in the direction of the previous moment. The preset time range may be determined by analyzing the dialogue update interval of the user in the historical interaction, or may be set according to actual needs.

[0122] It is understandable that the interaction process between the user and the terminal is continuous, and may be a dialogue with multiple inputs and multiple outputs, so there may be a situation where historical interaction information exists within a preset time range of the current user interaction information. Obtaining historical interaction information may be obtaining historical user interaction information.

[0123] According to the user interaction information and the historical interaction information, the target knowledge matching the user interaction information is screened from the target knowledge set, and the matching target knowledge is screened by combining the user interaction information and the historical interaction information. In a specific embodiment, the key fields in the user interaction information and the historical interaction information are analyzed separately for matching. Furthermore, different weights may be set for the user interaction information and the historical interaction information in the process of matching the target knowledge, and the key fields of the historical interaction information may be used as an attention mechanism to achieve preliminary screening of the target knowledge, and the screened target knowledge is further matched and screened by the user interaction information.

[0124] A medical device knowledge interaction method provided in this embodiment obtains the interaction time of the user interaction information, and filters the target knowledge matching the user interaction information from the target knowledge set according to the user interaction information and the historical interaction information. This method can guide the display of the target knowledge by combining the historical interaction information between the user and the terminal, so that the target knowledge displayed by the user interaction information within a preset time range will not deviate from the topic of the conversation, thereby achieving the effect of improving the matching degree between the interaction content and the specific group of people and the interaction topic.

[0125] In order to more clearly illustrate the technical solution of the present application, the present application also provides a detailed embodiment. A medical device knowledge interaction method of the present embodiment can be applied to Figure 3 The application environment shown in the figure includes user, interface interaction and question-answering interface, user portrait module, personalized recommendation module, natural language processing module, user information database and intelligent training module. Among them:

[0126] The user inputs user interaction information through the interface interaction and question-and-answer interface. For example, "Please teach me how to perform a brain MRI scan." The interface interaction and question-and-answer interface sends it to the user portrait module, and displays the target display information returned by the user portrait module.

[0127] The user portrait module may be used to transmit user interaction information and obtain the user's account history information; the account history information includes user identification, historical question and answer information, and historical learning records; the user portrait is determined based on the user's account history information; the skill level information, interest areas, and learning progress included in the user portrait are used to help the personalized recommendation module understand the user's needs and provide personalized training content. Specifically, it may be to obtain the user's seniority information in the user identification; obtain the answer text and the preset standard text in the historical question and answer information; calculate the similarity between the answer text and the preset standard text; determine the user's skill level information based on the user's seniority information and / or similarity. Exemplarily, in the user portrait generated by the user portrait module, it may be determined that the user is a novice, indicating that the user does not know much about magnetic resonance scanning.

[0128] The personalized recommendation module may be used to recommend suitable training content to users based on their historical learning records and ability levels, thereby helping users learn and improve their skills more efficiently. Specifically, it may be to screen a target knowledge set that matches the user profile from the medical device knowledge set based on the user profile. Furthermore, each piece of knowledge in the medical device knowledge set is provided with level information; it may be to obtain skill level information in the user profile; and based on the skill level information and the level information, select a target knowledge set that matches the skill level information from the medical device knowledge set.

[0129] The intelligent training module can be used to store, update and retrieve knowledge and skills related to medical equipment, such as magnetic resonance imaging. The natural language processing module can obtain the required information by communicating with this module. Specifically, it can be used to filter target knowledge that matches the user interaction information from the target knowledge set based on the user interaction information. Specifically, it can be used to extract key fields in the user interaction information; vectorize the key fields to obtain the first vectorized features of the user interaction information; vectorize each knowledge in the target knowledge set to obtain the second vectorized features corresponding to each knowledge; calculate the similarity between the first vectorized feature and multiple second vectorized features; and determine the knowledge corresponding to the second vectorized feature with the highest similarity as the target knowledge. Exemplarily, the basic steps and precautions for brain magnetic resonance imaging can be selected as answers.

[0130] Furthermore, the intelligent training module is also used to obtain the interaction time of the user interaction information; if there is historical interaction information within the preset time range of the interaction time, the historical interaction information is obtained, and based on the user interaction information and the historical interaction information, the target knowledge matching the user interaction information is screened from the target knowledge set.

[0131] The natural language processing module may be responsible for converting the natural language queries input by the user into structured requests so that the intelligent training model can understand and process them. At the same time, the module also converts the answers provided by the intelligent training model into natural language so that the user can easily understand them. When the natural language processing module obtains the target knowledge output by the intelligent training module, it may extract the key fields in the target knowledge; perform text expansion on the key fields based on the user interaction information to obtain the target display information after natural language processing. Exemplarily, the target display information after natural language processing may be "The basic steps of brain magnetic resonance imaging include: 1. Ensure patient safety, 2. Set appropriate scanning parameters, 3. Select a suitable scanning sequence, 4. Observe and analyze the results. During the scan, please make sure the patient's head is fixed to avoid motion artifacts. If you have any questions, please feel free to ask me."

[0132] The user information database may be used to store user identification and corresponding historical question and answer information and historical learning records.

[0133] The medical device knowledge interaction method provided in this embodiment can provide more accurate training content by automatically analyzing the user portrait based on the user interaction information and the user's account history information. Based on the user portrait, skill level and learning goals, the system can recommend more suitable training courses, case studies and practical guidance for users, and improve the matching degree of interactive content with specific groups of people; through the conversion of natural language and programming language, the system can understand the user's questions and provide clear and concise answers, reducing the learning difficulty of the system and the user; this embodiment can also regularly update the medical device knowledge set to ensure that the training content keeps pace with the times; the medical device knowledge interaction method of this embodiment can also select personalized learning paths from the knowledge set based on the user's learning progress, skill level information and goals, so as to improve the learning efficiency of users.

[0134] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0135] Based on the same inventive concept, the embodiment of the present application also provides a medical device intelligent training system for implementing the above-mentioned medical device knowledge interaction method. The medical device intelligent training system includes a natural context communication module (Natural Context Communication Module, NCCM) and a professional database matching module (Professional Database Matching Module, PDBM). Among them, the natural context communication module is responsible for processing and understanding user input and providing appropriate responses; the professional database matching module is connected to professional databases related to medical scanning technology.

[0136] In one embodiment, taking the application of the intelligent training system in a magnetic resonance device as an example, the natural context communication module includes a preprocessing unit, a natural language understanding unit, a context short-term memory unit, and a response generation unit. Specifically:

[0137] Preprocessing unit, where users clean input text, such as removing punctuation marks, converting to lowercase, etc.

[0138] The natural language understanding unit is used to understand the user input intention. The algorithm it adopts may include using BERT or other pre-trained language models for text classification or entity recognition. Among them, the BERT model refers to a language representation model that represents the bidirectional encoder representation from the Transformer.

[0139] Contextual short-term memory units are used to store and encode historical interaction information between users and the system. They may include a Transformer model, which treats each question and answer pair as a sequence unit. These units are input into the Transformer model in chronological order. The Transformer model encodes this information and generates a context vector. When there is new input, the context vector can be used to generate a more accurate and relevant response.

[0140] The response generation unit is used to generate a response based on the understood intent and context information. Optionally, GPT-3 or other applicable generation models can be selected.

[0141] The professional database matching module may include: an initialization module, a search module, a sorting and screening module, and an output module, wherein:

[0142] The initialization module is used for database initialization. In this embodiment, the initialization of the initialization module may include initialization of magnetic resonance related parameters, workflow, safety specifications, etc.

[0143] In the search module, the user uses a fuzzy search algorithm to find database items that are related to the user query but do not completely match it. In this embodiment, the search module uses an algorithm that combines TF-IDF (term frequency-inverse document frequency) with cosine similarity. The search module uses the TF-IDF algorithm to convert each item in the database into a vector, that is, when a query is received, it is also converted into a vector using TF-IDF; then, the cosine similarity algorithm is used to find the database vector that is most similar to the query vector, specifically by calculating the cosine value of the angle between the two vectors to evaluate the similarity between the two.

[0144] The sorting and filtering module weights each database item related to the query according to the set weight (such as accuracy, relevance, novelty, etc.); the item with the highest weighted score will be returned first.

[0145] The output module is used to combine the context vector of the natural context communication module and the search results of the professional database matching module to generate the final response.

[0146] Based on the same inventive concept, the embodiment of the present application also provides a medical device knowledge interaction device for implementing the medical device knowledge interaction method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more medical device knowledge interaction device embodiments provided below can refer to the limitations of the medical device knowledge interaction method above, and will not be repeated here.

[0147] In one embodiment, Figure 4 As shown, a medical device knowledge interaction device is provided, comprising: an acquisition module 100, a screening module 200 and a display module 300, wherein:

[0148] An acquisition module 100 is used to acquire user interaction information and user portraits;

[0149] A screening module 200, configured to screen a target knowledge set matching the user portrait from the medical device knowledge set according to the user portrait;

[0150] The display module 300 is used to filter target knowledge matching the user interaction information from the target knowledge set according to the user interaction information, and display the target knowledge.

[0151] In one embodiment, the medical device knowledge interaction device further includes a portrait determination module, which is used to:

[0152] Obtaining the user's account history information; the account history information includes user identification, historical question and answer information, and historical learning records;

[0153] The user portrait is determined based on the user's account history information; the user portrait includes skill level information, interest areas and learning progress.

[0154] In one embodiment, the portrait determination module is further used to:

[0155] Obtaining user seniority information from the user identifier;

[0156] Obtaining answer text and preset standard text in the historical question and answer information;

[0157] Calculating the similarity between the answer text and the preset standard text;

[0158] Based on the user's seniority information and / or similarity, skill level information of the user is determined.

[0159] In one embodiment, the screening module 200 is further used to:

[0160] Each piece of knowledge in the medical equipment knowledge set is correspondingly provided with level information;

[0161] Get the skill level information in the user profile;

[0162] According to the skill level information and the grade information, a target knowledge set matching the skill level information is selected from the medical equipment knowledge set.

[0163] In one embodiment, the display module 300 is further used for:

[0164] Extracting key fields from the user interaction information;

[0165] Performing vectorization processing on the key field to obtain a first vectorized feature of the user interaction information;

[0166] Performing vectorization processing on each piece of knowledge in the target knowledge set to obtain a second vectorized feature corresponding to each piece of knowledge;

[0167] Calculate similarity between the first vectorized feature and the plurality of the second vectorized features;

[0168] The knowledge corresponding to the second vectorized feature with the highest similarity is determined as the target knowledge.

[0169] In one embodiment, the display module 300 is further used for:

[0170] Extracting key fields from the target knowledge;

[0171] Performing text expansion on the key fields based on the user interaction information to obtain target display information after natural language processing;

[0172] The target display information is displayed.

[0173] In one embodiment, the medical device knowledge interaction device further includes a memory matching module for:

[0174] Acquire the interaction time of the user interaction information;

[0175] If there is historical interaction information within the preset time range of the interaction time, obtaining the historical interaction information;

[0176] Filtering a target knowledge set matching the user profile from a medical device knowledge set according to the user profile;

[0177] According to the user interaction information and the historical interaction information, target knowledge matching the user interaction information is filtered from the target knowledge set, and the target knowledge is displayed.

[0178] Each module in the above-mentioned medical device knowledge interaction device can be implemented in whole or in part by software, hardware and their combination. Each of the above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the corresponding operations of each of the above modules.

[0179] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a medical device knowledge interaction method is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a button, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0180] Those skilled in the art will understand that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0181] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the medical device knowledge interaction method of any of the above embodiments is implemented:

[0182] Obtain user interaction information and user portraits;

[0183] Filtering a target knowledge set matching the user profile from a medical device knowledge set according to the user profile;

[0184] According to the user interaction information, target knowledge matching the user interaction information is filtered from the target knowledge set, and the target knowledge is displayed.

[0185] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the medical device knowledge interaction method of any of the above embodiments is implemented:

[0186] Obtain user interaction information and user portraits;

[0187] Filtering a target knowledge set matching the user profile from a medical device knowledge set according to the user profile;

[0188] According to the user interaction information, target knowledge matching the user interaction information is filtered from the target knowledge set, and the target knowledge is displayed.

[0189] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0190] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0191] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0192] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A medical device knowledge interaction method, characterized in that: The method comprises: Obtain user interaction information and user portraits; Filtering a target knowledge set matching the user profile from a medical device knowledge set according to the user profile; According to the user interaction information, target knowledge matching the user interaction information is filtered from the target knowledge set, and the target knowledge is displayed.

2. The method according to claim 1, characterized in that The obtaining of user interaction information and user portrait includes: Obtaining the user's account history information; the account history information includes user identification, historical question and answer information, and historical learning records; The user portrait is determined based on the user's account history information; the user portrait includes skill level information, interest areas and learning progress.

3. The medical device knowledge interaction method according to claim 2, characterized in that: Determining the user portrait based on the user's account history information includes: Obtaining user seniority information from the user identifier; Obtaining answer text and preset standard text in the historical question and answer information; Calculating the similarity between the answer text and the preset standard text; Based on the user's seniority information and / or similarity, skill level information of the user is determined.

4. The medical device knowledge interaction method according to claim 1 or 3, characterized in that: The step of selecting a target knowledge set matching the user portrait from the medical device knowledge set according to the user portrait includes: Each piece of knowledge in the medical equipment knowledge set is correspondingly provided with level information; Get the skill level information in the user profile; According to the skill level information and the grade information, a target knowledge set matching the skill level information is selected from the medical equipment knowledge set.

5. The medical device knowledge interaction method according to claim 4, characterized in that: Filtering target knowledge matching the user interaction information from the target knowledge set according to the user interaction information includes: Extracting key fields from the user interaction information; Performing vectorization processing on the key field to obtain a first vectorized feature of the user interaction information; Performing vectorization processing on each piece of knowledge in the target knowledge set to obtain a second vectorized feature corresponding to each piece of knowledge; Calculate similarity between the first vectorized feature and the plurality of the second vectorized features; The knowledge corresponding to the second vectorized feature with the highest similarity is determined as the target knowledge.

6. The medical device knowledge interaction method according to claim 1, characterized in that: The presenting of the target knowledge comprises: Extracting key fields from the target knowledge; Performing text expansion on the key fields based on the user interaction information to obtain target display information after natural language processing; The target display information is displayed.

7. The medical device knowledge interaction method according to claim 1, characterized in that: The method further comprises: Acquire the interaction time of the user interaction information; If there is historical interaction information within the preset time range of the interaction time, obtaining the historical interaction information; Filtering a target knowledge set matching the user profile from a medical device knowledge set according to the user profile; According to the user interaction information and the historical interaction information, target knowledge matching the user interaction information is filtered from the target knowledge set, and the target knowledge is displayed.

8. A medical equipment knowledge interaction device, characterized in that: The device comprises: Acquisition module, used to obtain user interaction information and user portraits; A screening module, used for screening a target knowledge set matching the user portrait from the medical device knowledge set according to the user portrait; The display module is used to filter target knowledge matching the user interaction information from the target knowledge set according to the user interaction information, and display the target knowledge.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images