Risk assessment list generation method and device, electronic equipment and storage medium

Abstract

The invention provides a risk assessment list generation method and device, electronic equipment and a storage medium. The risk assessment list generation method comprises the following steps: acquiring voice information; obtaining a prompt word according to the voice information and a first prompt word template; a plurality of first assessment results are obtained according to the prompt words and a risk assessment model, and the first assessment results correspond to the risk assessment items; and generating a target risk assessment list according to the plurality of first assessment results and a risk assessment list template, the risk assessment list template including a plurality of risk assessment items. According to the method, the voice information is analyzed through the risk assessment model, the assessment result corresponding to each risk assessment item is determined, and finally the assessment results are filled into the risk assessment list, so that manual intervention is reduced, and the problems that manual assessment is not standard, not rigorous, inaccurate and wrong in filling are avoided.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method and device for generating a risk assessment sheet, an electronic device and a storage medium. Background Art

[0002] Generally, after a patient is admitted to the hospital, the nurse will conduct a risk assessment for the patient to prevent falls, prevent pressure sores, etc. The risk assessment form contains multiple evaluation contents. In related technologies, nurses fill out paper risk assessment forms based on personal experience and judgment by comparing patient information and asking about the patient's condition. After filling out, it will be entered into the nursing system. Some hospitals that have completed information transformation will fill out electronic versions of risk assessment forms on mobile terminals. However, no matter which method is used, the level of intelligence is still low. The quality of the risk assessment form depends entirely on the evaluator, so there are problems such as filling errors, non-standard, imprecise, and inaccurate evaluations. Summary of the invention

[0003] The present invention aims to solve at least one of the technical problems existing in the prior art or related art.

[0004] To this end, a first aspect of the present invention proposes a method for generating a risk assessment sheet.

[0005] A second aspect of the present invention provides a device for generating a risk assessment sheet.

[0006] A third aspect of the present invention provides an electronic device.

[0007] A fourth aspect of the present invention provides a storage medium.

[0008] In view of this, according to the first aspect of the present invention, a method for generating a risk assessment form is proposed, including: obtaining voice information; obtaining a prompt based on the voice information and a first prompt template; obtaining multiple first assessment results based on the prompt and a risk assessment model, wherein the first assessment results correspond to risk assessment items; generating a target risk assessment form based on the multiple first assessment results and a risk assessment form template, wherein the risk assessment form template includes multiple risk assessment items.

[0009] The method for generating a risk assessment form provided by the present invention mainly comprises: first obtaining voice information, wherein the voice information is the voice information of the conversation between the nurse and the patient, and the voice information can be obtained in real time or after the conversation between the nurse and the patient. After obtaining the voice information, the voice information needs to be processed to distinguish the speech content of the nurse and the patient, that is, to distinguish the words spoken by the nurse and the words spoken by the patient respectively. By distinguishing the conversation between the nurse and the patient, it is helpful to make the evaluation result obtained by the subsequent use of the risk assessment model more accurate. After obtaining the voice information, the voice information is added to the pre-set first prompt template to obtain a prompt, which is a bridge for interaction between the user and the model, and the prompt can guide the model to generate the desired output, so adding the voice information to the pre-set first prompt template to obtain the prompt can make the risk assessment model better understand the content of the voice information, and generate an output highly related to the voice information, that is, inputting the voice information into the risk assessment model through the prompt can obtain a more accurate evaluation result. After obtaining the prompt, the prompt is input into the risk assessment model to obtain multiple first assessment results, wherein the first assessment results correspond to the risk assessment items, and the risk assessment model can be a large language model. It can be understood that there will be multiple risk assessment items on the risk assessment form, and the voice information will also contain information related to one or more risk assessment items, so multiple first assessment results will be obtained by inputting the prompt into the risk assessment model, and each first assessment result corresponds to a risk assessment item. Finally, the multiple first assessment results output by the risk assessment model are filled into the risk assessment form template, thereby generating a target risk assessment form, wherein the risk assessment form template includes multiple risk assessment items, and then each first assessment result corresponding to the risk assessment item is filled into the corresponding risk assessment item in the risk assessment form template, thereby completing the filling of the risk assessment form. The present invention first adds the voice information between the nurse and the patient to the corresponding prompt template to obtain the prompt, then inputs the prompt into the risk assessment model, and uses the multimodal recognition, natural language understanding, logical analysis and reasoning of the large language model, and the ability to generate results to analyze the voice information, determine the assessment result corresponding to each risk assessment item, and finally fill the assessment result into the risk assessment form. The present invention completes the electronic version of the risk assessment form through artificial intelligence automatic analysis, extraction, and filling, thereby reducing manual intervention, thereby avoiding the problems of non-standard, imprecise, inaccurate and incorrect filling of manual assessments.

[0010] The method for generating the risk assessment form according to the present invention may also have the following technical features: In some technical solutions, optionally, before the step of obtaining a prompt based on voice information and a first prompt template, the step includes: collecting multiple risk assessment sheets; splitting each risk assessment sheet to obtain multiple risk assessment items; constructing a second prompt template based on the risk assessment items; and combining multiple second prompt templates to obtain a first prompt template.

[0011] In this technical solution, before the step of obtaining the prompt according to the voice information and the first prompt template, it includes: first collecting a variety of risk assessment forms. It can be understood that there are many types of risk assessment forms, such as: anti-fall risk assessment form, anti-pressure sore risk assessment form, etc. Then each risk assessment form is split into multiple independent risk assessment items, and a second prompt template is constructed for each risk assessment item. Finally, multiple second prompt templates are combined to obtain a first prompt template. By combining multiple second prompt templates corresponding to different risk assessment items to obtain a first prompt template, only one prompt template needs to be used in the future, thereby saving work time.

[0012] In some technical schemes, optionally, the step of obtaining a prompt based on voice information and a first prompt template includes: determining a risk assessment item corresponding to the voice information based on the voice information; inputting the voice information into a second prompt template in the first prompt template based on the risk assessment item to obtain a prompt, wherein the second prompt template corresponds to the risk assessment item.

[0013] In the technical solution, the step of obtaining a prompt according to the voice information and the first prompt template includes: firstly determining the risk assessment item corresponding to the voice information according to the voice information, that is, analyzing the voice information to determine which risk assessment items the voice information is related to, and then adding the voice information to the second prompt template of the first prompt template according to the corresponding risk assessment item, that is, adding the risk assessment item related to the voice information to the corresponding second prompt template to obtain the prompt. By obtaining the prompt by adding the voice information to the corresponding second prompt template, the accuracy of the risk assessment model is improved.

[0014] In some technical solutions, optionally, before the step of obtaining multiple first assessment results according to the prompt and the risk assessment model, it includes: obtaining a training data set; and training a large language model according to the training data set to obtain a risk assessment model.

[0015] In the technical solution, before the step of obtaining multiple first evaluation results according to the prompt and the risk assessment model, it includes: first obtaining a training data set, wherein the training data set includes multiple training data and the evaluation results corresponding to each training data, and then training the large language model according to the multiple training data and the evaluation results corresponding to each training data, so as to obtain the risk assessment model required by the present invention. Since the large language model has a strong language generation ability and in-depth speech understanding ability, the risk assessment model is obtained by training the large language model, so that the evaluation results on the risk assessment sheet generated at the end are more accurate.

[0016] In some technical solutions, optionally, the step of obtaining a training data set includes: collecting multiple text information, wherein the text information includes physical information of the patient; marking the association relationship between each text information and the second evaluation result corresponding to the text information; and generating a training data set based on the marked multiple text information and the multiple second evaluation results corresponding to the text information.

[0017] In this technical solution, the step of obtaining a training data set includes: first, collecting a plurality of text information, wherein the text information refers to text information obtained based on the conversation between the nurse and the patient, so the text information will contain information such as the patient's physical condition. Then, multiple second assessment results corresponding to each text information are determined, and at the same time, the association relationship between each text information and the multiple second assessment results corresponding to the text information is marked. Finally, the multiple marked text information and the multiple second assessment results corresponding to the text information are combined to obtain a training data set. By marking the association relationship between the text information and the second assessment results, and then obtaining a training data set based on the marked text information and the second assessment results corresponding to the text information, the accuracy of the risk assessment model obtained by training is ensured.

[0018] In some technical solutions, optionally, before the step of marking the association between each text information and the second evaluation result corresponding to the text information, it includes: performing a preprocessing operation on the text information, wherein the preprocessing operation includes: one or more combinations of data filtering, data cleaning, data desensitization and paragraph splitting.

[0019] In this technical solution, before the step of marking the association between each text information and the second assessment result corresponding to the text information, it includes: performing a preprocessing operation on the text information, wherein the preprocessing operation may include one or a combination of data filtering, data cleaning, data desensitization, and paragraph splitting. That is to say, before marking the association between the text information and the risk assessment item, it is necessary to perform data filtering, data cleaning, data desensitization, and paragraph splitting on the text information, thereby ensuring the accuracy of the text information.

[0020] In some technical solutions, optionally, the step of obtaining the prompt word according to the voice information and the first prompt word template includes: converting the voice information into text information; and obtaining the prompt word according to the text information and the first prompt word template.

[0021] In the technical solution, the step of obtaining a prompt according to the voice information and the first prompt template includes: first converting the voice information into text information, and then adding the text information to the first prompt template to obtain the prompt. By converting the voice information into text information, it is convenient to add the voice information to the prompt template to obtain the prompt.

[0022] According to a second aspect of the present invention, a device for generating a risk assessment form is proposed, wherein the device for generating a risk assessment form comprises: a first acquisition module, the first acquisition module is used to acquire voice information; a first processing module, the first processing module is used to obtain a prompt based on the voice information and a first prompt template; a second processing module, the second processing module is used to obtain a plurality of first assessment results based on the prompt and a risk assessment model, wherein the first assessment results correspond to risk assessment items; a generation module, the generation module is used to generate a target risk assessment form based on the plurality of first assessment results and a risk assessment form template, wherein the risk assessment form template includes a plurality of risk assessment items.

[0023] The device for generating a risk assessment form provided by the present invention mainly comprises: a first acquisition module, a first processing module, a second processing module and a generation module. Among them, the first acquisition module needs to acquire voice information, wherein the voice information is the voice information of the conversation between the nurse and the patient, and the voice information can be acquired in real time or acquired after the conversation between the nurse and the patient. After acquiring the voice information, the voice information needs to be processed to distinguish the speech content of the nurse and the patient, that is, to distinguish the words spoken by the nurse and the words spoken by the patient respectively. By distinguishing the conversation between the nurse and the patient, it is helpful to make the evaluation result obtained by the subsequent use of the risk assessment model more accurate. After obtaining the voice information, the first processing module adds the voice information to the pre-set first prompt template to obtain a prompt, which is a bridge for interaction between the user and the model, and the prompt can guide the model to generate the desired output, so adding the voice information to the pre-set first prompt template to obtain the prompt can make the risk assessment model better understand the content of the voice information and generate an output highly related to the voice information, that is, inputting the voice information into the risk assessment model through the prompt can obtain a more accurate evaluation result. After obtaining the prompt, the second processing module inputs the prompt into the risk assessment model to obtain multiple first assessment results, wherein the first assessment results correspond to the risk assessment items, and the risk assessment model can be a large language model. It can be understood that there will be multiple risk assessment items on the risk assessment form, and the voice information will also contain information related to one or more risk assessment items, so inputting the prompt into the risk assessment model will obtain multiple first assessment results, and each first assessment result corresponds to a risk assessment item. Finally, the generation module fills in the multiple first assessment results output by the risk assessment model into the risk assessment form template, thereby generating a target risk assessment form, wherein the risk assessment form template includes multiple risk assessment items, and then each first assessment result corresponding to the risk assessment item is filled in the corresponding risk assessment item in the risk assessment form template, thereby completing the filling of the risk assessment form. The present invention first adds the voice information between the nurse and the patient to the corresponding prompt template to obtain the prompt, then inputs the prompt into the risk assessment model, and uses the multimodal recognition, natural language understanding, logical analysis and reasoning of the large language model, and the ability to generate results to analyze the voice information, determine the assessment result corresponding to each risk assessment item, and finally fill the assessment result into the risk assessment form. The present invention completes the electronic version of the risk assessment form through artificial intelligence automatic analysis, extraction, and filling, thereby reducing manual intervention, thereby avoiding the problems of non-standard, imprecise, inaccurate and incorrect filling of manual assessments.

[0024] According to a third aspect of the present invention, an electronic device is proposed, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of a method for generating a risk assessment sheet as described above are implemented.

[0025] The electronic device provided by the present invention implements the steps of the method for generating the risk assessment form when the processor executes the computer program, which can achieve the technical effects of any of the above-mentioned technical solutions and will not be repeated here.

[0026] According to a fourth aspect of the present invention, a storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method for generating a risk assessment sheet as described above are implemented.

[0027] The storage medium provided by the present invention implements the steps of the above-mentioned method for generating the risk assessment form when the computer program is executed by the processor, and can achieve the technical effects of any of the above-mentioned technical solutions, which will not be repeated here.

[0028] Additional aspects and advantages of the present invention will become apparent in the following description or may be learned by practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which: Figure 1 A schematic diagram showing a flow chart of a method for generating a risk assessment form according to an embodiment of the present invention; Figure 2 A schematic diagram of a process flow before the step of obtaining a prompt message according to voice information and a first prompt message template in a method for generating a risk assessment form according to an embodiment of the present invention is shown; Figure 3 One of the flowchart diagrams shows the step of obtaining a prompt message according to voice information and a first prompt message template in a method for generating a risk assessment form according to an embodiment of the present invention; Figure 4 A schematic diagram of a process flow before the step of obtaining a plurality of first assessment results according to prompts and a risk assessment model in a method for generating a risk assessment form according to an embodiment of the present invention is shown; Figure 5 A schematic flow chart showing the steps of obtaining a training data set in a method for generating a risk assessment form according to an embodiment of the present invention; Figure 6 A second flow chart showing the step of obtaining a prompt message according to voice information and a first prompt message template in a method for generating a risk assessment form according to an embodiment of the present invention; Figure 7A schematic diagram showing the principle of a method for generating a risk assessment sheet according to an embodiment of the present invention; Figure 8 A schematic block diagram showing a device for generating a risk assessment form according to an embodiment of the present invention; Fig. 9 A schematic block diagram of an electronic device according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0030] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0031] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0032] Figure 1 A schematic diagram of a process for generating a risk assessment form according to an embodiment of the present invention is shown. The method comprises: Step 102: Acquire voice information; Step 104: Obtaining a prompt message according to the voice information and the first prompt message template; Step 106: obtaining a plurality of first assessment results according to the prompt and the risk assessment model, wherein the first assessment results correspond to the risk assessment items; Step 108: Generate a target risk assessment form according to the multiple first assessment results and the risk assessment form template, wherein the risk assessment form template includes multiple risk assessment items.

[0033] The method for generating a risk assessment form provided by the present invention mainly comprises: first obtaining voice information, wherein the voice information is the voice information of the conversation between the nurse and the patient, and the voice information can be obtained in real time or after the conversation between the nurse and the patient. After obtaining the voice information, the voice information needs to be processed to distinguish the speech content of the nurse and the patient, that is, to distinguish the words spoken by the nurse and the words spoken by the patient respectively. By distinguishing the conversation between the nurse and the patient, it is helpful to make the evaluation result obtained by the subsequent use of the risk assessment model more accurate. After obtaining the voice information, the voice information is added to the pre-set first prompt template to obtain a prompt, which is a bridge for interaction between the user and the model, and the prompt can guide the model to generate the desired output, so adding the voice information to the pre-set first prompt template to obtain the prompt can make the risk assessment model better understand the content of the voice information, and generate an output highly related to the voice information, that is, inputting the voice information into the risk assessment model through the prompt can obtain a more accurate evaluation result. After obtaining the prompt, the prompt is input into the risk assessment model to obtain multiple first assessment results, wherein the first assessment results correspond to the risk assessment items, and the risk assessment model can be a large language model. It can be understood that there will be multiple risk assessment items on the risk assessment form, and the voice information will also contain information related to one or more risk assessment items, so multiple first assessment results will be obtained by inputting the prompt into the risk assessment model, and each first assessment result corresponds to a risk assessment item. Finally, the multiple first assessment results output by the risk assessment model are filled into the risk assessment form template, thereby generating a target risk assessment form, wherein the risk assessment form template includes multiple risk assessment items, and then each first assessment result corresponding to the risk assessment item is filled into the corresponding risk assessment item in the risk assessment form template, thereby completing the filling of the risk assessment form. The present invention first adds the voice information between the nurse and the patient to the corresponding prompt template to obtain the prompt, then inputs the prompt into the risk assessment model, and uses the multimodal recognition, natural language understanding, logical analysis and reasoning of the large language model, and the ability to generate results to analyze the voice information, determine the assessment result corresponding to each risk assessment item, and finally fill the assessment result into the risk assessment form. The present invention completes the electronic version of the risk assessment form through artificial intelligence automatic analysis, extraction, and filling, thereby reducing manual intervention, thereby avoiding the problems of non-standard, imprecise, inaccurate and incorrect filling of manual assessments.

[0034] Figure 2 A flow chart of a method for generating a risk assessment form according to an embodiment of the present invention is shown before the step of obtaining a prompt message according to voice information and a first prompt message template; wherein, before the step of obtaining a prompt message according to voice information and a first prompt message template, the method includes: Step 202: Collect multiple risk assessment sheets; Step 204: Split each risk assessment sheet into multiple risk assessment items; Step 206: construct a second prompt template according to the risk assessment item; Step 208: Combine multiple second prompt templates to obtain a first prompt template.

[0035] In this embodiment, before the step of obtaining the prompt according to the voice information and the first prompt template, it includes: firstly collecting a plurality of risk assessment forms. It is understandable that there are many types of risk assessment forms, such as: anti-fall risk assessment form, anti-pressure sore risk assessment form, etc. Then each risk assessment form is split into multiple independent risk assessment items, and a second prompt template is constructed for each risk assessment item, and finally the multiple second prompt templates are combined to obtain the first prompt template. By combining multiple second prompt templates corresponding to different risk assessment items to obtain the first prompt template, only one prompt template needs to be used in the subsequent process, thereby saving working time.

[0036] Figure 3 One of the flow charts of the step of obtaining a prompt message according to voice information and a first prompt message template in a method for generating a risk assessment form according to an embodiment of the present invention is shown; wherein the step of obtaining a prompt message according to voice information and a first prompt message template includes: Step 302: determining a risk assessment item corresponding to the voice information according to the voice information; Step 304: inputting the voice information into the second prompt template in the first prompt template according to the risk assessment item to obtain a prompt, wherein the second prompt template corresponds to the risk assessment item.

[0037] In this embodiment, the step of obtaining a prompt according to the voice information and the first prompt template includes: firstly determining the risk assessment item corresponding to the voice information according to the voice information, that is, analyzing the voice information to determine which risk assessment items the voice information is related to, and then adding the voice information to the second prompt template of the first prompt template according to the corresponding risk assessment item, that is, adding the risk assessment item related to the voice information to the corresponding second prompt template to obtain the prompt. The prompt is obtained by adding the voice information to the corresponding second prompt template, thereby improving the accuracy of the risk assessment model.

[0038] Figure 4A flow chart of a method for generating a risk assessment form according to an embodiment of the present invention is shown before the step of obtaining a plurality of first assessment results according to prompts and a risk assessment model; wherein, before the step of obtaining a plurality of first assessment results according to prompts and a risk assessment model, the method includes: Step 402: Obtain a training data set; Step 404: Train the large language model according to the training data set to obtain a risk assessment model.

[0039] In this embodiment, before the step of obtaining multiple evaluation results according to the prompt and the risk assessment model, it includes: first obtaining a training data set, wherein the training data set includes multiple training data and the evaluation results corresponding to each training data, and then training the large language model according to the multiple training data and the evaluation results corresponding to each training data, so as to obtain the risk assessment model required by the present invention. Since the large language model has a strong language generation ability and in-depth speech understanding ability, the risk assessment model is obtained by training the large language model, so that the evaluation results on the risk assessment sheet generated at the end are more accurate.

[0040] Figure 5 A flow chart showing the step of obtaining a training data set in a method for generating a risk assessment form according to an embodiment of the present invention is provided; wherein the step of obtaining a training data set includes: Step 502: Collect multiple text information, wherein the text information includes the patient's physical information; Step 504: marking the association relationship between each text information and the second evaluation result corresponding to the text information; Step 506: Generate a training data set according to the annotated multiple text information and multiple second evaluation results corresponding to the text information.

[0041] In this embodiment, the step of obtaining a training data set includes: the step of obtaining a training data set includes: first, collecting a plurality of text information, wherein the text information refers to text information obtained based on the conversation between the nurse and the patient, so the text information will contain information such as the patient's physical condition. Then, multiple second assessment results corresponding to each text information are determined, and at the same time, the association relationship between each text information and the multiple second assessment results corresponding to the text information is marked. Finally, the multiple marked text information and the multiple second assessment results corresponding to the text information are combined to obtain a training data set. By marking the association relationship between the text information and the second assessment results, and then obtaining a training data set based on the marked text information and the second assessment results corresponding to the text information, the accuracy of the risk assessment model obtained by training is ensured.

[0042] In some embodiments, optionally, before the step of marking the association between each text information and the second evaluation result corresponding to the text information, it includes: performing a preprocessing operation on the text information, wherein the preprocessing operation includes: one or more combinations of data filtering, data cleaning, data desensitization, and paragraph splitting.

[0043] In this embodiment, before the step of marking the association between each text information and the second assessment result corresponding to the text information, it includes: performing a preprocessing operation on the text information, wherein the preprocessing operation may include one or a combination of data filtering, data cleaning, data desensitization, and paragraph splitting. That is to say, before marking the association between the text information and the risk assessment item, it is necessary to perform data filtering, data cleaning, data desensitization, and paragraph splitting on the text information, thereby ensuring the accuracy of the text information.

[0044] Figure 6 A second flow chart of the step of obtaining a prompt message according to voice information and a first prompt message template in a method for generating a risk assessment form according to an embodiment of the present invention is shown; wherein the step of obtaining a prompt message according to voice information and a first prompt message template further includes: Step 602: converting the voice information into text information; Step 604: Obtain a prompt message according to the text information and the first prompt message template.

[0045] In this embodiment, the step of obtaining the prompt according to the voice information and the first prompt template includes: first converting the voice information into text information, and then adding the text information to the first prompt template to obtain the prompt. By converting the voice information into text information, it is convenient to add the voice information to the prompt template to obtain the prompt.

[0046] Figure 7 FIG. 1 is a schematic diagram showing a method for generating a risk assessment form according to an embodiment of the present invention; Figure 7As shown, in the present invention, the real-time conversation voice information between the doctor and the patient is first obtained, and then the conversation voice information is converted into text information through ASR (Automatic Speech Recognition) technology, and then the text information is input into the prompt engineering, and the text information and the prompt template of the risk assessment form are used together to generate a prompt, and then the prompt is input into the risk assessment model, and after the risk assessment model understands the conversation content, risk assessment items and assessment criteria, the assessment results of the risk assessment items are given, and finally the assessment results are filled with data, and the assessment results are filled into the risk assessment form, and finally a risk assessment report, i.e., a target risk assessment form, is generated. Among them, the method for generating the risk assessment model includes: first, for each type of analysis and evaluation form, collect the text of the doctor-patient dialogue, and then preprocess the text. The preprocessing methods include: filtering, cleaning, desensitization, and paragraph splitting, etc. At the same time, determine the risk assessment results corresponding to the corresponding text, and mark the correlation between the risk assessment results and the text, that is, the dialogue content, and then generate fine-tuning corpus according to the dialogue content and the corresponding risk assessment results. Finally, fine-tune the fine-tuning corpus and the open source large language model Llama (Large Language Model Family of AI, large language model) to generate a risk assessment model, wherein the LoRA (Low-Rank Adaptation) fine-tuning technology can be used to fine-tune the Llama3-7B model to obtain the final risk assessment model. Furthermore, the process of constructing the prompt template includes: first, collect various types of risk assessment forms, then split the risk assessment forms into independent assessment items, then construct the prompt templates for each assessment item, and finally combine the prompt templates of multiple assessment items to obtain the final prompt template. The method for generating a risk assessment form provided by the present invention, on the one hand, greatly reduces the quality requirements for risk assessment personnel, and on the other hand, reduces the workload of manual entry by risk assessors. At the same time, since the risk assessment model also has the ability to learn and improve, as well as the ability to adapt to the same assessment form of different hospitals, it can also allow more hospitals to have unified assessment capabilities in various risk assessment forms and meet medical standards.

[0047] Figure 8 A schematic block diagram of a device for generating a risk assessment form according to an embodiment of the present invention is shown; wherein the device for generating a risk assessment form 80 comprises: A first acquisition module 802, used to acquire voice information; A first processing module 804, configured to obtain a prompt according to the voice information and a first prompt template; A second processing module 806 is used to obtain a plurality of first assessment results according to the prompt and the risk assessment model, wherein the first assessment results correspond to the risk assessment items; The generating module 808 is used to generate a target risk assessment form according to the multiple first assessment results and the risk assessment form template, wherein the risk assessment form template includes multiple risk assessment items.

[0048] The device 80 for generating a risk assessment form provided by the present invention mainly includes: a first acquisition module 802, a first processing module 804, a second processing module 806 and a generation module 808. Among them, the first acquisition module 802 needs to obtain voice information, wherein the voice information is the voice information of the conversation between the nurse and the patient, and the voice information can be obtained in real time or after the conversation between the nurse and the patient. After obtaining the voice information, it is also necessary to process the voice information to distinguish the speech content of the nurse and the patient, that is, to distinguish the words spoken by the nurse and the words spoken by the patient respectively. By distinguishing the conversation between the nurse and the patient, it is helpful to make the evaluation results obtained by the subsequent use of the risk assessment model more accurate. After obtaining the voice information, the first processing module 804 adds the voice information to the pre-set first prompt template to obtain the prompt. The prompt is a bridge for interaction between the user and the model. The prompt can guide the model to produce the desired output. Therefore, adding the voice information to the pre-set first prompt template to obtain the prompt can make the risk assessment model better understand the content of the voice information and generate an output highly related to the voice information, that is, inputting the voice information into the risk assessment model through the prompt can obtain a more accurate assessment result. After obtaining the prompt, the second processing module 806 inputs the prompt into the risk assessment model to obtain multiple first assessment results, wherein the first assessment result corresponds to the risk assessment item, and the risk assessment model can be a large language model. It can be understood that there will be multiple risk assessment items on the risk assessment sheet, and the voice information will also contain information related to one or more risk assessment items, so inputting the prompt into the risk assessment model will obtain multiple first assessment results, and each first assessment result corresponds to a risk assessment item. Finally, the generation module 808 fills in the multiple first assessment results output by the risk assessment model into the risk assessment single template, thereby generating a target risk assessment single, wherein the risk assessment single template includes multiple risk assessment items, and then fills in each first assessment result corresponding to the risk assessment item into the corresponding risk assessment item in the risk assessment single template, thereby completing the filling of the risk assessment single. The present invention first adds the voice information between the nurse and the patient to the corresponding prompt template to obtain the prompt, and then inputs the prompt into the risk assessment model, and uses the multimodal recognition, natural language understanding, logical analysis and reasoning of the large language model, and the ability to generate results to analyze the voice information, determine the assessment result corresponding to each risk assessment item, and finally fill the assessment result into the risk assessment single. The present invention completes the electronic version of the risk assessment single by means of artificial intelligence automatic analysis, extraction, and filling, thereby reducing manual intervention, thereby avoiding the problems of non-standard, imprecise, inaccurate and incorrect filling of manual assessments.

[0049] In some embodiments, optionally, the risk assessment form generating device 80 also includes: a first collection module, used to collect multiple risk assessment forms; a splitting module, used to split each risk assessment form to obtain multiple risk assessment items; a construction module, used to construct a second prompt template according to the risk assessment items; a first combination module, used to combine multiple second prompt templates to obtain a first prompt template.

[0050] In this embodiment, the risk assessment form generation device 80 also includes: a first collection module, a splitting module, a construction module and a first combination module. Among them, the first collection module can collect multiple risk assessment forms. It can be understood that there are many types of risk assessment forms, such as: anti-fall risk assessment form, anti-pressure sore risk assessment form, etc. The splitting module then splits each risk assessment form into multiple independent risk assessment items. At the same time, the construction module constructs a second prompt template for each risk assessment item. Finally, the first combination module combines multiple second prompt templates to obtain a first prompt template. By combining multiple second prompt templates corresponding to different risk assessment items, a first prompt template is obtained, so that only one prompt template needs to be used in the future, thereby saving work time.

[0051] In some embodiments, optionally, the first processing module 804 is specifically used to determine a risk assessment item corresponding to the voice information based on the voice information; input the voice information into a second prompt template in the first prompt template according to the risk assessment item to obtain a prompt, wherein the second prompt template corresponds to the risk assessment item.

[0052] In this embodiment, the first processing module 804 is specifically used to determine the risk assessment items corresponding to the voice information according to the voice information, that is, by analyzing the voice information, determine which risk assessment items the voice information is related to, and then add the voice information to the second prompt template of the first prompt template according to the corresponding risk assessment items, that is, add the risk assessment items related to the voice information to the corresponding second prompt template to obtain the prompt. By adding the voice information to the corresponding second prompt template to obtain the prompt, the accuracy of the risk assessment model is improved.

[0053] In some embodiments, optionally, the risk assessment form generating device 80 further includes: a second collection module for acquiring a training data set; and a training module for training a large language model according to the training data set to obtain a risk assessment model.

[0054] In this embodiment, the risk assessment sheet generation device 80 also includes: a second collection module and a training module. The second collection module can obtain a training data set, wherein the training data set includes multiple training data and an evaluation result corresponding to each training data, and then the training module trains the large language model according to the multiple training data and the evaluation result corresponding to each training data, thereby obtaining the risk assessment model required by the present invention. Since the large language model has a powerful language generation ability and in-depth speech understanding ability, the risk assessment model is obtained by training the large language model, so that the evaluation result on the risk assessment sheet generated at the end is more accurate.

[0055] In some embodiments, optionally, the second collection module is specifically used to collect multiple text information, wherein the text information includes the patient's physical information; annotate the association between each text information and the second evaluation result corresponding to the text information; and generate a training data set based on the annotated multiple text information and the multiple second evaluation results corresponding to the text information.

[0056] In this embodiment, the second collection module is specifically used to collect multiple text information, wherein the text information refers to the text information obtained based on the conversation between the nurse and the patient, so the text information will contain information such as the patient's physical condition. Then, multiple second assessment results corresponding to each text information are determined, and the association relationship between each text information and the multiple second assessment results corresponding to the text information is marked. Finally, the multiple annotated text information and the multiple second assessment results corresponding to the text information are combined to obtain a training data set. By marking the association relationship between the text information and the second assessment results, and then obtaining a training data set based on the annotated text information and the second assessment results corresponding to the text information, the accuracy of the risk assessment model obtained by training is ensured.

[0057] In some embodiments, optionally, the second collection module is also specifically used to perform preprocessing operations on the text information, wherein the preprocessing operations include: one or a combination of data filtering, data cleaning, data desensitization, and paragraph splitting.

[0058] In this embodiment, the second collection module is also specifically used to perform preprocessing operations on the text information, wherein the preprocessing operations may include one or a combination of data filtering, data cleaning, data desensitization, and paragraph splitting. That is to say, before marking the association between the text information and the risk assessment items, it is necessary to perform data filtering, data cleaning, data desensitization, and paragraph splitting on the text information, thereby ensuring the accuracy of the text information.

[0059] In some embodiments, optionally, the first processing module 804 is further specifically configured to convert the voice information into text information; and obtain the prompt message according to the text information and the first prompt message template.

[0060] In this embodiment, the first processing module 804 is also specifically used to convert the voice information into text information, and then add the text information to the first prompt template to obtain the prompt. By converting the voice information into text information, it is convenient to add the voice information to the prompt template to obtain the prompt.

[0061] Fig. 9 A schematic block diagram of an electronic device of an embodiment of the present invention is shown; wherein the electronic device 90 includes a memory 902, a processor 904, and a computer program stored in the memory 902 and executable on the processor 904, and when the processor 904 executes the computer program, the steps of the method for generating a risk assessment form as described above are implemented.

[0062] The electronic device 90 provided by the present invention implements the steps of the method for generating the risk assessment form when the processor 904 executes the computer program, which can achieve the technical effects of any of the above embodiments and will not be repeated here.

[0063] An embodiment of the present invention provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for generating a risk assessment form as described above are implemented.

[0064] The storage medium provided by the present invention implements the steps of the above-mentioned method for generating a risk assessment form when the computer program is executed by a processor, and can achieve the technical effects of any of the above-mentioned embodiments, which will not be repeated here.

[0065] In the description of this specification, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance unless otherwise clearly specified and limited; the terms "connection", "installation", "fixation" and the like should be understood in a broad sense, for example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be directly connected or indirectly connected through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0066] In the description of this specification, the description of the terms "one embodiment", "some embodiments", "specific embodiments", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0067] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for generating a risk assessment form, characterized in that: include: Get voice information; Obtaining a prompt according to the voice information and a first prompt template; Obtaining a plurality of first assessment results according to the prompt and the risk assessment model, wherein the first assessment results correspond to the risk assessment items; A target risk assessment form is generated according to the plurality of the first assessment results and a risk assessment form template, wherein the risk assessment form template includes a plurality of the risk assessment items.

2. The method for generating a risk assessment form according to claim 1, characterized in that: Before the step of obtaining the prompt word according to the voice information and the first prompt word template, the method for generating the risk assessment form includes: Collect various risk assessment sheets; Splitting each risk assessment sheet to obtain a plurality of risk assessment items; Constructing a second prompt template according to the risk assessment item; The first prompt template is obtained by combining a plurality of the second prompt templates.

3. The method for generating a risk assessment form according to claim 2, characterized in that: The step of obtaining the prompt word according to the voice information and the first prompt word template comprises: Determining the risk assessment item corresponding to the voice information according to the voice information; The voice information is input into the second prompt template in the first prompt template according to the risk assessment item to obtain the prompt, wherein the second prompt template corresponds to the risk assessment item.

4. The method for generating a risk assessment form according to claim 1, characterized in that: Before the step of obtaining a plurality of first assessment results according to the prompt and the risk assessment model, the method for generating the risk assessment sheet includes: Get the training data set; The risk assessment model is obtained by training a large language model according to the training data set.

5. The method for generating a risk assessment form according to claim 4, characterized in that: The step of obtaining a training data set includes: Collecting a plurality of text information, wherein the text information includes physical information of the patient; Annotating the association relationship between each piece of text information and the second evaluation result corresponding to the text information; The training data set is generated according to the labeled multiple text information and the multiple second evaluation results corresponding to the text information.

6. The method for generating a risk assessment form according to claim 5, characterized in that: Before the step of marking the association relationship between each piece of text information and the second assessment result corresponding to the text information, the method for generating the risk assessment sheet includes: The text information is preprocessed, wherein the preprocessing operation includes: one or a combination of data filtering, data cleaning, data desensitization, and paragraph splitting.

7. The method for generating a risk assessment form according to any one of claims 1 to 6, characterized in that: The step of obtaining the prompt word according to the voice information and the first prompt word template comprises: Converting the voice information into text information; The prompt is obtained according to the text information and the first prompt template.

8. A device for generating a risk assessment form, characterized in that: include: A first acquisition module, wherein the first acquisition module is used to acquire voice information; A first processing module, the first processing module is used to obtain a prompt according to the voice information and a first prompt template; A second processing module, the second processing module is used to obtain a plurality of first assessment results according to the prompt and the risk assessment model, wherein the first assessment results correspond to the risk assessment items; A generation module is used to generate a target risk assessment form according to the plurality of the first assessment results and a risk assessment form template, wherein the risk assessment form template includes a plurality of the risk assessment items.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method for generating a risk assessment sheet according to any one of claims 1 to 7 are implemented.

10. A 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 for generating a risk assessment sheet according to any one of claims 1 to 7 are implemented.

Images