Multidisciplinary diagnosis and treatment and science popularization method, system, equipment and product based on language model

Abstract

The invention provides a multidisciplinary diagnosis and treatment and science popularization method, system, device and product based on a language model.The multidisciplinary diagnosis and treatment and science popularization method based on the language model.The multidisciplinary diagnosis and treatment and science popularization method based on the language model.The multidisciplinary diagnosis and treatment and science popularization method based on the language model.The multidisciplinary diagnosis and treatment and science popularization method comprises the steps that based on a received illness state inquiry request from a user terminal, a preset illness state inquiry problem list of a multidisciplinary Generating medical history information based on the received illness state description information sent by the user terminal; according to the medical history information, obtaining a department corresponding to an illness state consulted by the user terminal; sending the medical history information to a diagnosis terminal corresponding to the department; generating comprehensive diagnosis and treatment information based on the received single-department diagnosis and treatment information sent by the diagnosis terminal; sending the comprehensive diagnosis and treatment information to the user terminal; and performing science popularization pushing to the user terminal based on the medical history information and the comprehensive diagnosis and treatment information. Doctor time management can be optimized, communication efficiency among doctors is improved, doctor-patient communication efficiency is enhanced, and patient participation degree is improved.

Description

Background Art

[0002] With the development of medicine, multidisciplinary diagnosis and treatment (Multidisciplinary Team, MDT) has gradually become an important method for solving the diagnosis and treatment of complex diseases. The core concept of multidisciplinary diagnosis and treatment is to improve the accuracy of diagnosis and treatment effect by bringing together doctors from different specialties to conduct comprehensive evaluation and comprehensive treatment of patients. When patients face specific symptoms, they often need to transfer between different departments. For example, headaches may involve neurology, neurosurgery, orthopedics or psychiatry, while abdominal pain may require evaluation by multiple departments such as gastroenterology, urology, and nephrology. This phenomenon reflects the complexity of the disease and the necessity of multidisciplinary cooperation.

[0003] However, although multidisciplinary diagnosis and treatment can theoretically provide patients with more comprehensive medical services, it still faces many challenges in practical applications. At present, patients often encounter difficulties when registering, the appointment quotas of various departments are tight, and the test results often need to be referred repeatedly, resulting in poor medical experience for patients and even fear of medical treatment. In addition, MDT is usually based on an appointment system. Doctors are often pressed for time during the diagnosis and treatment process and find it difficult to communicate in depth with patients. Patients and their families often just act as bystanders in the process and lack a sense of active participation. Moreover, even if the doctor has plenty of time, due to the professionalism of medicine, patients and their families lack basic popular science information, and it is difficult to understand the doctor's professional diagnosis and treatment recommendations in a short period of time.

[0004] Currently, the scarcity and high cost of multidisciplinary diagnosis and treatment (MDT) clinics significantly restrict their promotion and popularization. Due to the limited number of MDT clinics, patients often have difficulty in obtaining timely services. In addition, the high cost makes it impossible for many patients to enjoy this medical model economically and efficiently. This situation makes it impossible to normalize MDT clinics, resulting in limited efforts to optimize physician time management, improve communication efficiency between physicians, strengthen doctor-patient communication efficiency, and improve patient engagement. In general, these shortcomings have prevented multidisciplinary diagnosis and treatment from fully realizing its potential advantages, affecting patients' medical experience and the overall quality of medical services. Summary of the invention

[0005] In view of this, the present invention provides a multidisciplinary diagnosis and treatment and popular science method, system, equipment and product based on a language model, so as to at least solve the problems of low doctor time utilization, low communication efficiency between doctors, low doctor-patient communication efficiency and low patient participation in the existing multidisciplinary diagnosis and treatment methods.

[0006] In one aspect, an embodiment of the present invention provides a multidisciplinary diagnosis and treatment and popular science method based on a language model, comprising:

[0007] Based on receiving a medical condition inquiry request from a user terminal, sending a medical condition inquiry question list preset by a multidisciplinary diagnosis and treatment mechanism to the user terminal;

[0008] Generate medical history information based on the condition description information received from the user terminal;

[0009] Obtain the department corresponding to the condition consulted by the user terminal according to the medical history information;

[0010] Send medical history information to the corresponding diagnosis terminal of the department;

[0011] Generate comprehensive diagnosis and treatment information based on the diagnosis and treatment information of a single department sent by the diagnosis terminal;

[0012] Sending comprehensive diagnosis and treatment information to user terminals;

[0013] Based on medical history information and comprehensive diagnosis and treatment information, popular science information is pushed to user terminals.

[0014] In some embodiments, based on the medical history information and comprehensive diagnosis and treatment information, popular science information is pushed to the user terminal, including:

[0015] Based on a preset disease keyword list, keyword matching is performed on the medical history information and the comprehensive diagnosis and treatment information; wherein the medical history information and the comprehensive diagnosis and treatment information are in text form;

[0016] Based on the successful matching of the condition keywords, the successfully matched condition keywords are obtained;

[0017] Based on the successfully matched disease keywords, popular science articles and videos are retrieved through language models;

[0018] Based on the preset screening mechanism, obtain the filtered popular science articles and videos;

[0019] Send the selected popular science articles and videos to the user terminal.

[0020] In some embodiments, generating medical history information based on the condition description information received from the user terminal includes:

[0021] Receiving condition description information sent from a user terminal; the condition description information may be in the form of text, voice, image, and video, and the language used in the condition description information may be Chinese, Tibetan, and English;

[0022] Based on the language model program, the disease description information in the form of voice, image and video is converted into text form, and the disease description information in a language different from the preset language is converted into the preset language; wherein the language model program has the ability to process text, voice, image and video;

[0023] Based on the language model program, the converted condition description information is modularized according to at least the following categories: the patient's basic information, chief complaint information, past medical history and medication information, and the medical history information is obtained.

[0024] In some embodiments, obtaining the department corresponding to the condition consulted by the user terminal according to the medical history information includes:

[0025] Based on the preset disease keyword list, keyword matching is performed on the text-based medical history information;

[0026] Based on the successful matching of the condition keyword, the department corresponding to the successfully matched condition keyword is obtained.

[0027] In some embodiments, based on the single-department diagnosis and treatment information received from the diagnosis terminal, comprehensive diagnosis and treatment information is generated, including:

[0028] Receiving single-department diagnosis and treatment information sent from a diagnosis terminal; wherein the single-department diagnosis and treatment information includes text and voice;

[0029] Convert the diagnosis and treatment information of a single department in voice form into text form;

[0030] Semantically segment each department's diagnosis and treatment information in text form into a preliminary diagnosis module, a further examination module, and a treatment recommendation module;

[0031] Semantic segmentation is continued for each preliminary diagnosis module, further examination module, and treatment suggestion module, and each module is divided into multiple sub-items;

[0032] Semantic vector conversion is performed on each sub-item of the preliminary diagnosis module, further examination module and treatment suggestion module respectively, and two sub-items with the difference of the semantic vector within a preset range are merged as sub-items of the preliminary diagnosis module, further examination module and treatment suggestion module of the comprehensive diagnosis and treatment information respectively.

[0033] In some embodiments, semantic vector conversion is performed on each sub-item of the preliminary diagnosis module, the further examination module, and the treatment suggestion module, and the difference of the semantic vectors is within a preset range. The two sub-items are merged as the sub-items of the preliminary diagnosis module, the further examination module, and the treatment suggestion module of the comprehensive diagnosis and treatment information, respectively. After that, it also includes:

[0034] Sending the preliminary diagnosis module, further examination module and treatment suggestion module of the comprehensive diagnosis and treatment information, as well as the popular rewriting instructions to the language model program;

[0035] The popular rewritten text data generated based on the popular rewriting instructions returned by the language model program is obtained as comprehensive diagnosis and treatment information; wherein the text data returned by the language model program includes a preliminary diagnosis module, a further examination module and a treatment suggestion module.

[0036] In some embodiments, after sending comprehensive diagnosis and treatment information to the user terminal, the method further includes:

[0037] In response to receiving the inquiry information sent by the user terminal, keyword matching is performed on the inquiry information, the department corresponding to the keyword successfully matched is obtained, and the inquiry information is sent to the diagnosis terminal corresponding to the department; wherein the inquiry information includes the inspection result information corresponding to the further inspection module;

[0038] In response to receiving the reply information from the diagnosis terminal, the reply information is sent to the user terminal.

[0039] On the other hand, an embodiment of the present invention further provides a multidisciplinary diagnosis and treatment and science popularization system based on a language model, comprising:

[0040] The medical condition inquiry module sends a list of medical condition inquiry questions preset by the multidisciplinary diagnosis and treatment mechanism to the user terminal based on the medical condition inquiry request received from the user terminal;

[0041] A medical history generating module generates medical history information based on the medical condition description information received from the user terminal;

[0042] The department acquisition module obtains the department corresponding to the condition consulted by the user terminal based on the medical history information;

[0043] Medical history push module, which sends medical history information to the corresponding diagnosis terminal of the department;

[0044] A diagnosis generation module generates comprehensive diagnosis and treatment information based on the diagnosis and treatment information of a single department sent by the diagnosis terminal;

[0045] Diagnosis push module, sending comprehensive diagnosis and treatment information to user terminals;

[0046] The popular science push module pushes popular science information to user terminals based on medical history information and comprehensive diagnosis and treatment information.

[0047] In another aspect, an embodiment of the present invention further provides a multidisciplinary diagnosis and treatment and popular science device based on a language model, comprising:

[0048] processor;

[0049] A memory storing executable instructions of a processor;

[0050] Among them, the processor is configured to perform the steps of the above-mentioned multidisciplinary diagnosis and treatment and popular science method based on language model by executing executable instructions.

[0051] On the other hand, an embodiment of the present invention further provides a computer program product, including a computer program, which, when executed, implements the steps of the above-mentioned multidisciplinary diagnosis and treatment and popular science method based on the language model.

[0052] The multidisciplinary diagnosis and treatment and popular science method, system, equipment and product based on language model of the present invention provide a multidisciplinary diagnosis and treatment and popular science method based on language model, realize interaction with users and form medical history information through language model, so as to improve the efficiency of communication between doctors and patients; by sending the sorted medical history information to the corresponding department diagnosis terminal respectively, doctors can diagnose and treat the medical history information separately, avoiding the problem that multiple doctors cannot diagnose and treat at the same time, so as to improve the utilization rate of doctors' time; through the language model, comprehensive diagnosis and treatment information is generated based on the single department diagnosis and treatment information of each doctor, so as to improve the communication efficiency between doctors; by using language model, users can obtain timely diagnosis, feedback and popular science push, which can improve patient participation. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The accompanying drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present invention, and together with the specification are used to explain the principles of the present invention. Obviously, the accompanying drawings described below are only some embodiments of the present invention, and for those of ordinary skill in the art, other accompanying drawings can be obtained based on these accompanying drawings without creative work.

[0054] Figure 1 It is a flowchart of the steps of a multidisciplinary diagnosis and treatment and popular science method based on a language model provided by an embodiment of the present invention;

[0055] Figure 2 yes Figure 1 Specific step flow chart of step S120;

[0056] Figure 3 yes Figure 1 Specific step flow chart of step S130;

[0057] Figure 4 yes Figure 1 Specific step flow chart of step S150;

[0058] Figure 5 yes Figure 4 Flow chart of subsequent steps of step S150;

[0059] Figure 6 yes Figure 1 Specific step flow chart of step S170;

[0060] Figure 7 This embodiment of the present invention provides a Figure 1 A flowchart of the subsequent steps of the multidisciplinary diagnosis and treatment and popular science method based on the language model;

[0061] Figure 8 is a module schematic diagram of a multidisciplinary diagnosis and treatment and popular science system based on a language model according to an embodiment of the present invention;

[0062] Fig. 9 is a schematic diagram of communication between a multidisciplinary diagnosis and treatment and popular science system based on a language model and a user terminal and a diagnosis terminal according to an embodiment of the present invention;

[0063] Fig.10 It is a structural schematic diagram of a multidisciplinary diagnosis and treatment and popular science device based on a language model according to an embodiment of the present invention. DETAILED DESCRIPTION

[0064] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that the present invention will be comprehensive and complete and fully convey the concepts of the example embodiments to those skilled in the art. The same reference numerals in the figures represent the same or similar structures, and thus their repeated description will be omitted.

[0065] The words "first", "second" and the like used in the specific description do not indicate any order, quantity or importance, but are only used to distinguish different components. In addition, in the description of the present invention, the orientation or position relationship indicated by the terms "upper" and "lower" are based on the orientation or position relationship shown in the drawings, which are only for the convenience of description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as a limitation of the present invention.

[0066] It should be noted that, in the absence of conflict, the embodiments of the present invention and features in different embodiments may be combined with each other.

[0067] The inventor of this case has provided a solution to the problems existing in the prior art through careful and in-depth research. The present invention provides a multidisciplinary diagnosis and treatment and popular science method, system, device and product based on a language model, wherein the multidisciplinary diagnosis and treatment and popular science method based on a language model includes: based on receiving a condition inquiry request from a user terminal, sending a list of condition inquiry questions preset by a multidisciplinary diagnosis and treatment mechanism to the user terminal; based on the condition description information received from the user terminal, generating medical history information; obtaining the department corresponding to the condition consulted by the user terminal according to the medical history information; sending the medical history information to the diagnosis terminal corresponding to the department; based on receiving the single department diagnosis and treatment information sent by the diagnosis terminal, generating comprehensive diagnosis and treatment information; sending the comprehensive diagnosis and treatment information to the user terminal. The multidisciplinary diagnosis and treatment and popular science method, system, equipment and product based on language model of the present invention provide a multidisciplinary diagnosis and treatment and popular science method based on language model, realize interaction with users and form medical history information through language model, so as to improve the efficiency of communication between doctors and patients; by sending the sorted medical history information to the corresponding department diagnosis terminal respectively, doctors can diagnose and treat the medical history information separately, avoiding the problem that multiple doctors cannot diagnose and treat at the same time, so as to improve the utilization rate of doctors' time; through the language model, comprehensive diagnosis and treatment information is generated based on the single department diagnosis and treatment information of each doctor, so as to improve the communication efficiency between doctors; by using language model, users can obtain timely diagnosis, feedback and popular science push, which can improve patient participation.

[0068] The above is the core idea of ​​the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention.

[0069] On one hand, Figure 1 As shown, an embodiment of the present invention provides a multidisciplinary diagnosis and treatment and popular science method based on a language model, comprising:

[0070] S110, based on receiving a condition inquiry request from a user terminal, sending a condition inquiry question list preset by a multidisciplinary diagnosis and treatment mechanism to the user terminal;

[0071] S120, generating medical history information based on the condition description information received from the user terminal;

[0072] S130, obtaining the department corresponding to the condition consulted by the user terminal according to the medical history information;

[0073] S140, sending the medical history information to the corresponding diagnosis terminal of the department;

[0074] S150, generating comprehensive diagnosis and treatment information based on the single department diagnosis and treatment information received from the diagnosis terminal;

[0075] S160, sending comprehensive diagnosis and treatment information to the user terminal;

[0076] S170. Push popular science information to the user terminal based on the medical history information and comprehensive diagnosis and treatment information.

[0077] It is worth noting that the above S110 to S170 are merely step numbers, which are used to facilitate reference and avoid text duplication. Unless otherwise specified, the above and subsequent step numbers will not limit the order of implementation of the various steps of the method. In other embodiments, the above steps of the method can also be written and implemented in an interchangeable order, and are not limited to this. For example, S160 and S170 only need to be after S150, and there is no order of precedence.

[0078] Among them, for the above-mentioned step S110, it is intended to inquire about the patient's condition with the user terminal, that is, the patient, and guide the patient to input the condition. Specifically, the user first initiates an inquiry to the system via the user terminal. Among them, the user can be the patient who needs medical treatment, or the patient's caregiver when the patient is unable to seek medical treatment independently, to initiate a condition inquiry request to the system on behalf of the patient. After receiving the user's condition inquiry request, the system guides the user to describe the patient's condition according to a preset list of condition inquiry questions. The list of condition inquiry questions is formulated by doctors in the hospital, and is intended to use the shortest time to inquire about the patient's basic information and condition. Compared with the existing multidisciplinary diagnosis and treatment process, multiple doctors and patients communicate separately, resulting in patients repeatedly describing their conditions. This application improves the efficiency of doctor-patient communication and can also avoid omissions when patients repeatedly describe their conditions.

[0079] Furthermore, step S110 is further described below through two examples.

[0080]

Example 1

[0081] “The system asks: Can you tell me why you feel unwell and want to see a doctor (chief complaint)?

[0082] The patient replied: My right hand is a little numb.

[0083] System asks: How long has this symptom lasted?

[0084] The patient replied: Half a year.

[0085] System question: Does the attack occur all the time or in waves?

[0086] The patient replied: In waves.

[0087] System questions: Do you have a history of hypertension, diabetes, coronary heart disease, and hyperlipidemia?

[0088] The patient replied: I have a history of coronary heart disease and diabetes.

[0089] System question: Are you taking any medications?

[0090] Patient's answer: Taking metformin and statins.

[0091] System question: Have you had any imaging examinations?

[0092] The patient replied: I had an MRI.

[0093] The system asks: Can you take a picture for me? Just click the upper right corner of the screen, then take a picture of the film and upload it.

[0094] Patient’s answer: It has been uploaded.

[0095] System query: OK, please pay through the Internet hospital and wait patiently, 4 experts will reply to you within 12 hours. "

[0096]

Example 2

[0097] “The system asks: Patient’s age and gender (all questions are in Tibetan)?

[0098] Family member’s answer: 53 years old, my husband (hinting for a middle-aged male).

[0099] The system asks: Did you fall down first and then become unable to speak, or did you become unable to speak first and then fall?

[0100] The family member replied: He became unable to speak first and then fell down (excluding bleeding after trauma, suggesting that the fall was caused by acute cerebral stroke).

[0101] System question: Have you ever had difficulty speaking clearly or felt weak in your right hands and feet?

[0102] The family member replied: It happened twice, and my husband didn’t take it seriously, thinking it was caused by overwork (the possibility of cerebral infarction is greater than cerebral hemorrhage).

[0103] System asks: Call the patient and see if he responds?

[0104] The family member replied: He has reactions but cannot speak (to judge his consciousness and condition).

[0105] System summary: OK, don't worry. The patient may have acute cerebral infarction. The fact that he is still conscious indicates that the area of ​​infarction is not large. Once cerebral infarction is confirmed, thrombolytic drugs must be used as soon as possible or interventional thrombectomy must be considered to restore speech and limb movement. From now on, please follow my instructions and act according to the steps (from inquiry to guidance stage).

[0106] System guidance: Please adjust the patient to lie on his side to avoid aspiration. Please reply to me when you are done (pre-hospital emergency guidance).

[0107] The family member replied: It’s done.

[0108] System guidance: When you arrive at the hospital and meet the nurse at the triage desk, tell her immediately that the patient may have had an acute stroke and needs to activate the stroke channel. In this way, the hospital will immediately follow the emergency process for stroke and provide the patient with diagnosis and treatment guidance as quickly as possible (embedded diagnosis and treatment process guidance).”

[0109] It can be seen that, in addition to guiding the patient to input the condition, the above step S110 can also be used to provide first aid guidance to the patient to protect the patient's life safety.

[0110] The above example is merely a specific explanatory description of one situation of step S110 of the present application, and is not a limitation of the present application. The list of medical condition inquiry questions of the present application may also include other questions, which are not listed one by one here.

[0111] like Figure 2 As shown, in some embodiments, the above step S120 may include:

[0112] S121, receiving condition description information sent from a user terminal; wherein the condition description information includes text, voice, image and video, and the language used in the condition description information includes Chinese, Tibetan and English;

[0113] S122, based on the language model program, converting the disease description information in the form of voice, image and video into text form, and converting the disease description information in a language different from the preset language into the preset language; wherein the language model program has the ability to process text, voice, image and video;

[0114] S123. Based on the language model program, the converted condition description information is modularized according to at least the following categories: the patient's basic information, chief complaint information, past medical history, and medication information, and the medical history information is obtained.

[0115] Among them, for the above-mentioned step S121, the form in which the user inputs the condition, that is, the form of the condition description information can be text, voice, picture or video. The condition input in the form of pictures can be photos of the patient's physical signs or current situation, photos of medical records, photos or screenshots of examination results, etc. The condition input in the form of videos can be recordings of the patient's physical signs or current situation, etc. The condition input in the form of text or voice can be a text or voice description of the patient's physical signs or current situation, medical records or examination results, etc., and the present application does not limit this. Further, for voice information, a voice activity detection algorithm can be used to distinguish between voice and non-voice parts in a complex environment, and a sliding window mechanism can be used to pre-process the voice signal to be recognized. The Mel Frequency Cepstral Coefficient (MFCC) method is further used to extract voice features in the Mel frequency spectrum, and a deep neural network based on the Transformer architecture is designed to convert audio to text, and the present application does not limit this.

[0116] The above step S122 is intended to process the condition description information and convert other forms of condition description information into text form. Information in other languages ​​is converted into a preset language. For example, the language model program can translate Tibetan and English into preset Chinese based on the built-in machine translation program, so as to improve the versatility of medical treatment for users with different language backgrounds. Furthermore, expert label data can be used to fine-tune large language models such as Wenxin Yiyan and Llama3 in vertical applications to achieve matching between patients' multimodal input information and diseases, and automatically generate text medical records.

[0117] For the above-mentioned step S123, it is intended to extract data for the condition description information. The condition description information is extracted and classified according to the above-mentioned modules so that it can be processed by the diagnosis and treatment terminal. When the diagnosis and treatment terminal is a doctor, the efficiency of doctor-patient communication can be improved, or when the diagnosis and treatment terminal is a diagnosis and treatment neural network model, the processing efficiency of the diagnosis and treatment terminal can be improved. Specifically, taking the patient's condition in the above-mentioned [Example 1] as an example, the converted condition description information can be: the patient is an elderly male (the patient's basic information), complaining of paroxysmal right upper limb numbness for half a year (chief complaint information), previous history of coronary heart disease and diabetes (past medical history), taking metformin and statins, and the control situation is unknown (medication information). In addition, according to the magnetic resonance imaging examination photos uploaded by the patient, it can also be obtained: magnetic resonance imaging suggests stenosis of the left middle cerebral artery.

[0118] like Figure 3 As shown, in some embodiments, step S130 may include:

[0119] S131, performing keyword matching on text-based medical history information based on a preset disease condition keyword list;

[0120] S132. Based on the successful matching of the condition keyword, the department corresponding to the successfully matched condition keyword is obtained.

[0121] The above steps S131 and S132 are intended to determine the department required for the diagnosis and treatment of the user's disease based on the medical history information obtained after the system has sorted it out. Specifically, a keyword matching technology can be used, and each department has a disease keyword list. When the disease keyword in the medical history information matches the disease keyword in the department's keyword list, the department is selected as the department required for the diagnosis and treatment of the user's disease.

[0122] Specifically, taking the patient's condition in the above-mentioned [Example 1] as an example, the departments corresponding to the successfully matched condition keywords may be neurology, neurosurgery, imaging, and rehabilitation.

[0123] In some embodiments, for the above step S140, the medical history information is sent to the diagnosis terminal corresponding to the department. Among them, the diagnosis terminal can be a doctor or a diagnosis and treatment neural network model, and the present application does not limit this. The diagnosis and treatment neural network model is a trained neural network model, which can obtain diagnostic suggestions for specific diseases based on the patient's medical history information. Furthermore, the subscription / publishing (Pub / Sub) mode can be used to send it to the diagnosis terminal corresponding to the relevant department. In addition, the reply time of the diagnosis terminal can also be limited, for example, it can be replied within 12 hours. When the diagnosis terminal does not reply within 12 hours, a prompt message can be sent to the diagnosis terminal to remind the diagnosis terminal to reply as soon as possible. The above setting enables doctors to diagnose and treat the medical history information separately by sending the sorted medical history information to the corresponding department diagnosis terminal, avoiding the problem that multiple doctors cannot diagnose and treat at the same time, so as to improve the utilization of doctors' time.

[0124] like Figure 4 As shown, in some embodiments, step S150 may include:

[0125] S151, receiving single-department diagnosis and treatment information sent from a diagnosis terminal; wherein the single-department diagnosis and treatment information includes text and voice;

[0126] S152, converting the diagnosis and treatment information of a single department in voice form into text form;

[0127] S153, semantically segmenting each department's diagnosis and treatment information in the processed text form into a preliminary diagnosis module, a further examination module, and a treatment suggestion module;

[0128] S154, continue to perform semantic segmentation on each preliminary diagnosis module, further examination module, and treatment suggestion module, and divide them into multiple sub-items;

[0129] S155. Perform semantic vector conversion on each sub-item of the preliminary diagnosis module, further examination module and treatment suggestion module respectively, merge two sub-items whose semantic vector difference is within a preset range, and use them as sub-items of the preliminary diagnosis module, further examination module and treatment suggestion module of the comprehensive diagnosis and treatment information respectively.

[0130] Regarding the above step S152, when the single department diagnosis and treatment information is in the form of voice, the above method can be used to convert the voice into text, which will not be repeated here.

[0131] For the above steps S153 and S154, taking the patient's condition in the above [Example 1] as an example, the semantic segmentation results of the diagnosis information of the neurology department's diagnosis terminal A, the neurosurgery department's diagnosis terminal B, the imaging department's diagnosis terminal C, and the rehabilitation department's diagnosis terminal D can be as follows:

[0132]

[0133] Among them, the preliminary diagnosis module of the single department diagnosis and treatment information of the diagnosis terminal A of the Department of Neurology has cerebral infarction, diabetes and hypertension as its three sub-items respectively. The treatment recommendation module of the single department diagnosis and treatment information of the diagnosis terminal D of the Department of Rehabilitation has hyperbaric oxygen and acupuncture as its two sub-items respectively. The sub-items of other modules of other diagnosis terminals are similar and will not be repeated here.

[0134] For the above-mentioned step S155, it is intended to summarize the single-department diagnosis and treatment information of all diagnostic terminals, and then form comprehensive diagnosis and treatment information. Specifically, multiple single-department diagnosis and treatment information can be merged and obtained by eliminating duplicate sub-items. Due to the existence of synonymous expressions, sub-items expressed in different words may express the same meaning. At this time, sub-items expressing the same meaning should be eliminated. Specifically, each sub-item can be converted into a semantic vector. When the difference between the semantic vectors is within a preset range, it is considered that the two sub-items express the same meaning, and then only one of them is retained as the sub-item of the final comprehensive diagnosis and treatment information. The above setting generates comprehensive diagnosis and treatment information based on the single-department diagnosis and treatment information of each doctor through a language model to improve the efficiency of communication between doctors.

[0135] Furthermore, taking the above-mentioned single department diagnosis and treatment information as an example, the comprehensive diagnosis and treatment information obtained after the sub-items are finally merged is as follows:

[0136]

[0137]

[0138] like Figure 5 As shown, in some embodiments, after step S155, the following may also be included:

[0139] S156, sending the preliminary diagnosis module, the further examination module and the treatment suggestion module of the comprehensive diagnosis and treatment information, and the popular rewriting instruction to the language model program;

[0140] S157. Obtain the colloquial rewritten text data generated based on the colloquial rewriting instructions returned by the language model program as comprehensive diagnosis and treatment information; wherein the text data returned by the language model program includes a preliminary diagnosis module, a further examination module and a treatment suggestion module.

[0141] The above steps S156 and S157 are intended to convert professional and difficult medical terms into popular language that ordinary users can understand, so as to facilitate user understanding. Specifically, an application program interface can be developed to call the fine-tuned large language model program, and the large language model program can be instructed by popular rewriting instructions to automatically generate medical reports that are easy for patients to understand, thereby improving the efficiency of doctor-patient communication and increasing the participation of patients and their families.

[0142] like Figure 6 As shown, in some embodiments, step S170 may include:

[0143] S171. Based on a preset disease condition keyword list, keyword matching is performed on the medical history information and the comprehensive diagnosis and treatment information; wherein the medical history information and the comprehensive diagnosis and treatment information are in text form;

[0144] S172, based on the successful matching of the condition keyword, obtaining the matched condition keyword;

[0145] S173. Based on the successfully matched disease keywords, popular science articles and videos are retrieved through the language model;

[0146] S174. Based on a preset screening mechanism, obtain screened popular science articles and videos;

[0147] S175. Send the screened popular science articles and videos to the user terminal.

[0148] Specifically, for the above step S171, the medical history information and the comprehensive diagnosis and treatment information have been converted from possible other forms of original information (voice, picture or video, etc.) into text form through the above steps to facilitate the subsequent keyword matching operation. The list of disease keywords can refer to step S130, which will not be repeated here.

[0149] For the above step S173, the large language model can be connected to the Internet to search from popular science articles and videos on the Internet. Popular science articles can be directly matched with keywords in full text, and popular science videos can be matched by text tags in the videos, so as to retrieve popular science articles and videos that match the above disease keywords.

[0150] For the above step S174, since there are a large number of popular science articles and videos, and the quality varies, it is necessary to screen the popular science articles and videos, and a preset screening mechanism is used to achieve the above screening. The preset screening mechanism may include that the number of fans of the blogger of the popular science articles and videos meets a certain value, the number of video playbacks meets a certain value, and the blogger's unit is located in a tertiary hospital or a specific university, etc., and this application does not limit this.

[0151] The above step S170 is intended to enable users to better understand their own diseases after the diagnosis is completed, improve the efficiency of doctor-patient communication, and increase the participation of patients and their families.

[0152] like Figure 7 As shown, in some embodiments, after step S160, the following steps may also be included:

[0153] S180, in response to receiving the inquiry information sent by the user terminal, performing keyword matching on the inquiry information, obtaining the department corresponding to the keyword that successfully matches, and sending the inquiry information to the diagnosis terminal corresponding to the department; wherein the inquiry information includes the inspection result information corresponding to the further inspection module;

[0154] S190 . In response to receiving reply information from the diagnosis terminal, send the reply information to the user terminal.

[0155] Since steps S160 and S170 are not restricted in sequence, steps S180 and S190 only need to be placed after S160. Figure 6 The placement of steps S180 and S190 in step S170 is merely an example and is not a limitation on the order of implementing steps S170 , S180 and S190 .

[0156] The above steps S180 and S190 are intended to allow users to ask further questions after receiving the comprehensive diagnosis and treatment information, or to allow users to execute them when they have a follow-up visit after further examination. For example, any questions that the user has about the comprehensive diagnosis and treatment information, or questions that are not mentioned in the comprehensive diagnosis and treatment information, can be fed back to the diagnosis terminal in the form of follow-up information. Alternatively, after the patient is examined according to the examination items of the further examination module of the comprehensive diagnosis and treatment information, the examination results are provided to the diagnosis terminal again as follow-up information. In these cases, the system will provide the follow-up information to the diagnosis terminals of the corresponding departments respectively, and after receiving the reply from the diagnosis terminal, it will be conveyed to the user to improve the efficiency of doctor-patient communication.

[0157] Specifically, taking the patient's condition in the above [Example 1] as an example, the above steps S170, S180 and S190 can be embodied in the following forms:

[0158] “User question: The patient is feeling much better after taking the medicine, is there anything else I need to do?

[0159] System reply: We have selected 3 popular science articles and videos on acute stroke produced by doctors from well-known domestic hospitals, focusing on the causes, treatment and rehabilitation of stroke. You can read them while accompanying the patient (popular science rating and targeted push). You don’t have to worry about not being able to understand Chinese characters, we will read them aloud for you in Tibetan (voice and semantic interaction).”

[0160] In addition, during this process, the system can also convert text information into voice information and specific language to facilitate user understanding.

[0161] In some embodiments, the above-mentioned condition description information, medical history information, single department diagnosis and treatment information, comprehensive diagnosis and treatment information, follow-up information and reply information can all be integrated into the multidisciplinary diagnosis and treatment and popular science system based on the language model through blockchain technology to protect user privacy.

[0162] The multidisciplinary diagnosis and treatment and popular science method based on a language model of the present invention provides a multidisciplinary diagnosis and treatment and popular science method based on a language model, realizes interaction with users and forms medical history information through a language model, so as to improve the efficiency of communication between doctors and patients; by sending the sorted medical history information to the corresponding department diagnosis terminals respectively, doctors can diagnose and treat the medical history information separately, avoiding the problem that multiple doctors cannot diagnose and treat at the same time, so as to improve the utilization rate of doctors' time; through the language model, comprehensive diagnosis and treatment information is generated based on the single department diagnosis and treatment information of each doctor, so as to improve the efficiency of communication between doctors; through the use of language models, users can obtain timely diagnosis, feedback and popular science push, which can improve patient participation.

[0163] On the other hand, Figure 8 and Fig. 9 As shown, the embodiment of the present invention further provides a multidisciplinary diagnosis and treatment and popular science system 50 based on a language model, comprising:

[0164] The condition inquiry module 51 sends a condition inquiry question list preset by the multidisciplinary diagnosis and treatment mechanism to the user terminal 10 based on the condition inquiry request received from the user terminal 10;

[0165] The medical history generating module 52 generates medical history information based on the condition description information received from the user terminal 10;

[0166] The department acquisition module 53 obtains the department corresponding to the condition consulted by the user terminal 10 according to the medical history information;

[0167] The medical history push module 54 sends the medical history information to the diagnosis terminal 30 corresponding to the department;

[0168] The diagnosis generation module 55 generates comprehensive diagnosis and treatment information based on the single department diagnosis and treatment information received from the diagnosis terminal 30;

[0169] The diagnosis push module 56 sends comprehensive diagnosis and treatment information to the user terminal 10;

[0170] The science popularization push module 57 pushes science popularization information to the user terminal 10 based on the medical history information and comprehensive diagnosis and treatment information.

[0171] In addition to the aforementioned modules, the multidisciplinary diagnosis and treatment and popular science system 50 based on the language model may also include other modules, but since they are irrelevant to the content of the embodiment of the present invention, the specific description of these modules is omitted here. In addition, since the functions of the multidisciplinary diagnosis and treatment and popular science system 50 based on the language model are similar to the details of the multidisciplinary diagnosis and treatment and popular science method based on the language model, the repeated description of some of the same contents is omitted here.

[0172] The technical effects of the multidisciplinary diagnosis and treatment and popular science system based on a language model of the present invention can be referred to the embodiments of the multidisciplinary diagnosis and treatment and popular science method based on a language model, and will not be repeated here.

[0173] On the other hand, Fig.10 As shown, an embodiment of the present invention further provides a multidisciplinary diagnosis and treatment and popular science device based on a language model, comprising: a processor; a memory, wherein executable instructions of the processor are stored. The processor is configured to execute the steps of the multidisciplinary diagnosis and treatment and popular science method based on a language model by executing the executable instructions.

[0174] It will be appreciated by those skilled in the art that various aspects of the present invention may be implemented as systems, methods or program products. Therefore, various aspects of the present invention may be specifically implemented in the following forms, namely: complete hardware implementation, complete software implementation (including firmware, microcode, etc.), or a combination of hardware and software implementations, which may be collectively referred to herein as "circuits", "modules" or "platforms".

[0175] Refer to the following Fig.10 The electronic device 600 according to this embodiment of the present invention is described. Fig.10 The electronic device 600 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0176] like Fig.10 As shown, the electronic device 600 is in the form of a general computing device. The components of the electronic device 600 may include, but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different platform components (including the storage unit 620 and the processing unit 610), a display unit 640, etc.

[0177] The storage unit stores program codes, which can be executed by the processing unit 610, so that the processing unit 610 performs the steps according to various exemplary embodiments of the present invention described in the above method section of this specification. For example, the processing unit 610 can perform the following steps: Figure 1 Follow the steps shown in .

[0178] The storage unit 620 may include a readable medium in the form of a volatile storage unit, such as a random access storage unit (RAM) 6201 and / or a cache storage unit 6202 , and may further include a read-only storage unit (ROM) 6203 .

[0179] The storage unit 620 may also include a program / utility 6204 having a set (at least one) of program modules 6205, such program modules 6205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0180] Bus 630 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0181] The electronic device 600 may also communicate with one or more external devices 700 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 600, and / or communicate with any device that enables the electronic device 600 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 650. Furthermore, the electronic device 600 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 660. The network adapter 660 may communicate with other modules of the electronic device 600 via a bus 630. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms, etc.

[0182] The technical effects of the multidisciplinary diagnosis and treatment and popular science equipment based on language models of the present invention can be referred to the embodiments of the multidisciplinary diagnosis and treatment and popular science methods based on language models mentioned above, and will not be repeated here.

[0183] In another aspect, an embodiment of the present invention further provides a computer program product, which includes a computer program stored in a computer-readable storage medium. A processor of a computing device can read the computer program from the computer-readable storage medium, and the processor executes the computer program, so that the computing device executes the multidisciplinary diagnosis and treatment and popular science methods based on language models described in the above embodiments.

[0184] The computer program included in the computer program product of an embodiment of the present invention can be written in any combination of one or more programming languages. Programming languages ​​include object-oriented programming languages, such as Java, C++, etc., and also include conventional procedural programming languages, such as "C" language or similar programming languages. The computer program can be executed entirely on the user computing device, partially on the user device, as an independent software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, using an Internet service provider to connect through the Internet).

[0185] The technical effects of the computer program product of the present invention can be referred to the aforementioned embodiments of the multidisciplinary diagnosis and treatment and popular science method based on the language model, and will not be repeated here.

[0186] In summary, the multidisciplinary diagnosis and treatment and popular science method, system, storage medium and program product based on language model of the present invention, by providing a multidisciplinary diagnosis and treatment and popular science method based on language model, realizes interaction with users and forms medical history information through language model, so as to improve the efficiency of communication between doctors and patients; by sending the sorted medical history information to the corresponding department diagnosis terminal respectively, doctors can diagnose and treat the medical history information separately, avoiding the problem that multiple doctors cannot diagnose and treat at the same time, so as to improve the utilization rate of doctors' time; through the language model, comprehensive diagnosis and treatment information is generated based on the single department diagnosis and treatment information of each doctor, so as to improve the efficiency of communication between doctors; by using language model, users can obtain timely diagnosis, feedback and popular science push, which can improve patient participation.

[0187] The above contents are further detailed descriptions of the present invention in combination with specific optional implementation methods, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the scope of protection of the present invention.

Claims

1. A multidisciplinary diagnosis and treatment and popular science method based on a language model, characterized in that: include: Based on receiving a condition inquiry request from a user terminal, sending a condition inquiry question list preset by a multidisciplinary diagnosis and treatment mechanism to the user terminal; Generate medical history information based on the condition description information received from the user terminal; Obtaining the department corresponding to the condition consulted by the user terminal according to the medical history information; Sending the medical history information to the diagnosis terminal corresponding to the department; Generate comprehensive diagnosis and treatment information based on the single department diagnosis and treatment information received from the diagnosis terminal; Sending the comprehensive diagnosis and treatment information to the user terminal; Based on the medical history information and the comprehensive diagnosis and treatment information, popular science information is pushed to the user terminal.

2. The multidisciplinary diagnosis and treatment and popular science method based on language model according to claim 1, characterized in that: The pushing of popular science information to the user terminal based on the medical history information and the comprehensive diagnosis and treatment information includes: Based on a preset disease condition keyword list, keyword matching is performed on the medical history information and the comprehensive diagnosis and treatment information; wherein the medical history information and the comprehensive diagnosis and treatment information are in text form; Based on the successful matching of the condition keyword, obtaining the successfully matched condition keyword; Retrieve popular science articles and videos based on the successfully matched disease keywords; Based on the preset screening mechanism, obtain the filtered popular science articles and videos; The screened popular science articles and videos are sent to the user terminal.

3. The multidisciplinary diagnosis and treatment and popular science method based on language model according to claim 1, characterized in that: The generating of medical history information based on the received medical condition description information sent by the user terminal includes: Receiving the condition description information sent from the user terminal; wherein the condition description information includes text, voice, image and video, and the language used by the condition description information includes Chinese, Tibetan and English; Based on the language model program, the disease description information in the form of voice, image and video is converted into text form, and the disease description information in a language different from the preset language is converted into the preset language; wherein the language model program has the ability to process text, voice, image and video; The converted condition description information is modularized based on a language model program according to at least the following categories: basic information of the patient, chief complaint information, past medical history and medication information, and the medical history information is obtained.

4. The multidisciplinary diagnosis and treatment and popular science method based on language model according to claim 3 is characterized in that: The obtaining, according to the medical history information, of the department corresponding to the condition consulted by the user terminal comprises: Based on a preset disease condition keyword list, keyword matching is performed on the medical history information in text form; Based on the successful matching of the condition keyword, the department corresponding to the condition keyword that has been successfully matched is obtained.

5. The multidisciplinary diagnosis and treatment and popular science method based on language model according to claim 1, characterized in that: The generating of comprehensive diagnosis and treatment information based on the single department diagnosis and treatment information received from the diagnosis terminal includes: Receiving the single-department diagnosis and treatment information sent from the diagnosis terminal; wherein the single-department diagnosis and treatment information includes text and voice; Converting the single-department diagnosis and treatment information in voice form into text form; Semantically segmenting each of the processed text-based diagnosis and treatment information of the single department into a preliminary diagnosis module, a further examination module, and a treatment recommendation module; Semantic segmentation is continued for each preliminary diagnosis module, further examination module, and treatment suggestion module, and each module is divided into multiple sub-items; Semantic vector conversion is performed on each sub-item of the preliminary diagnosis module, further examination module and treatment suggestion module respectively, and two sub-items with a difference in the semantic vector within a preset range are merged as sub-items of the preliminary diagnosis module, further examination module and treatment suggestion module of the comprehensive diagnosis and treatment information respectively.

6. The multidisciplinary diagnosis and treatment and popular science method based on language model according to claim 5, characterized in that: The semantic vector conversion is performed on each sub-item of the preliminary diagnosis module, the further examination module and the treatment suggestion module respectively, and the difference of the semantic vectors is within a preset range. The two sub-items are merged to serve as the sub-items of the preliminary diagnosis module, the further examination module and the treatment suggestion module of the comprehensive diagnosis and treatment information respectively, and then the following is also included: sending the preliminary diagnosis module, the further examination module and the treatment suggestion module of the comprehensive diagnosis and treatment information, as well as the colloquial rewriting instructions to the language model program; The popular rewritten text data generated based on the popular rewriting instruction and returned by the language model program is obtained as the comprehensive diagnosis and treatment information; wherein the text data returned by the language model program includes a preliminary diagnosis module, a further examination module and a treatment suggestion module.

7. The multidisciplinary diagnosis and treatment and popular science method based on language model according to claim 1, characterized in that: After sending the comprehensive diagnosis and treatment information to the user terminal, the method further includes: In response to receiving the inquiry information sent by the user terminal, performing keyword matching on the inquiry information, obtaining the department corresponding to the keyword that successfully matches, and sending the inquiry information to the diagnosis terminal corresponding to the department; wherein the inquiry information includes the inspection result information corresponding to the further inspection module; In response to receiving the reply information from the diagnosis terminal, the reply information is sent to the user terminal.

8. A multidisciplinary diagnosis and treatment and popular science system based on a language model, characterized in that: include: A condition inquiry module, based on receiving a condition inquiry request from a user terminal, sends a condition inquiry question list preset by a multidisciplinary diagnosis and treatment mechanism to the user terminal; A medical history generating module, generating medical history information based on the condition description information received from the user terminal; A department acquisition module, which obtains the department corresponding to the condition consulted by the user terminal according to the medical history information; A medical history push module sends the medical history information to the diagnosis terminal corresponding to the department; A diagnosis generation module, which generates comprehensive diagnosis and treatment information based on the single department diagnosis and treatment information received from the diagnosis terminal; A diagnosis push module, sending the comprehensive diagnosis and treatment information to the user terminal; The popular science push module pushes popular science information to the user terminal based on the medical history information and the comprehensive diagnosis and treatment information.

9. A multidisciplinary diagnosis and treatment and popular science device based on a language model, characterized in that: include: processor; A memory storing executable instructions of the processor; Wherein, the processor is configured to execute the steps of the multidisciplinary diagnosis and treatment and popular science method based on the language model as described in any one of claims 1 to 7 by executing the executable instructions.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed, the steps of the multidisciplinary diagnosis and treatment and popular science method based on a language model described in any one of claims 1 to 7 are implemented.

Images