Auxiliary triage method and device based on illness state description information, equipment and medium

Abstract

The invention provides an auxiliary triage method and device based on illness state description information, equipment and a medium, and the method comprises the steps: determining a reference set from a plurality of preset sets based on a target age, and constructing a target set based on a plurality of habitual descriptors associated with a target part in the reference set; inputting the target set and the illness state description information into an NLP model, and obtaining a target description word determined by the NLP model from the target set based on the illness state description information; and determining preset symptom information associated with the target descriptor as target symptom information, and determining a triage result based on the target symptom information. According to the method, a target set can be dynamically constructed based on habitual descriptors conforming to patient parts and description habits by taking target parts and target ages as information indexes, the misrecognition risk of an NLP model is reduced by eliminating irrelevant classification bases, the habitual descriptors are taken as triage bases, the triage accuracy is improved under the condition that professional nouns are not familiar, and the triage efficiency is improved. The patient experience is improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent triage technology, and in particular to an auxiliary triage method, device, equipment and medium based on condition description information. Background Art

[0002] At present, most hospitals have begun to adopt online registration. After patients register on the hospital's mini program or APP, they can choose the corresponding department and doctor for treatment according to their health status. However, the distribution of departments in different hospitals is different, and each doctor's expertise is also different. Many first-time patients still don't know how to choose the department and doctor when registering, and the patient experience is poor.

[0003] With the development of artificial intelligence technology, natural language processing (NLP) models have begun to be applied to the field of auxiliary triage. In related technologies, NLP models have begun to be deployed in triage equipment. Patients input their condition description information through user terminals or triage equipment. The triage equipment inputs the condition description information into the NLP model and then outputs recommended departments and recommended doctors, which can improve the patient experience to a certain extent. However, related technologies use the condition description information directly as the input of the NLP model. Although the current NLP model has a high recognition accuracy, some patients still have insufficient language expression ability, and the condition description information is usually vague, resulting in the output results of the NLP model still deviating from the patient's needs. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention proposes an auxiliary triage method, device, equipment, and medium based on condition description information, which can optimize the recognition and classification of the NLP model, improve the accuracy of auxiliary triage, and improve the patient experience.

[0005] In a first aspect, an embodiment of the present invention provides an auxiliary triage method based on condition description information, which is applied to a triage device, wherein the triage device is preset with an NLP model, multiple preset symptom information, and multiple preset sets, each of the preset sets includes conventional descriptive words associated with each of the preset symptom information, the conventional descriptive words associated with the same preset symptom information in at least two of the preset sets are different, each of the conventional descriptive words is associated with at least one preset part, and each of the preset sets corresponds to a different age range, and the method includes: Obtain target age, target site and condition description information input by the patient; Determine a reference set from a plurality of preset sets based on the target age, and construct a target set based on a plurality of the conventional descriptors associated with the target part in the reference set; Inputting the target set and the condition description information into the NLP model, and obtaining the target description words determined by the NLP model from the target set based on the condition description information; The preset symptom information associated with the target descriptor is determined as target symptom information, and a triage result is determined based on the target symptom information, wherein the triage result records target doctor information.

[0006] According to some embodiments of the present invention, after determining the triage result based on the target symptom information, the method further includes: Determining the first descriptive words corresponding to the respective target descriptive words in the condition description information, determining the target descriptive words as second descriptive words, and determining the target symptom information as reference symptom information; constructing reference description information based on the first descriptor, the second descriptor and the reference symptom information; The reference description information is recorded, and the reference description information is associated with the target doctor information.

[0007] According to some embodiments of the present invention, the triage device is preset with a plurality of optional doctor information, each of the optional doctor information is pre-associated with at least one of the preset parts and at least one of the preset symptom information; determining the triage result based on the target symptom information includes: Determine the multiple optional doctor information associated with the target part as candidate doctor information, and traverse each of the candidate doctor information based on the condition description information, wherein at least one of the candidate doctor information is pre-associated with at least one of the reference description information; When the first descriptive word associated with the candidate doctor information is recorded in the condition description information, the corresponding candidate doctor information is determined as the target doctor information; When the condition description information fails to successfully match any of the first description words, determining at least one target doctor information from the plurality of candidate doctor information based on the target symptom information; The triage result is generated based on any of the target doctor information.

[0008] According to some embodiments of the present invention, before generating the triage result based on any of the target doctor information, the method further includes: When the target doctor information cannot be determined based on the condition description information and the target symptom information, determining a plurality of associated parts pre-associated with the target part from the plurality of preset parts; determining the plurality of optional doctor information associated with the associated part as associated doctor information, wherein the associated doctor information does not include any of the candidate doctor information; The target doctor information is determined from the plurality of associated doctor information based on the target symptom information and the condition description information.

[0009] According to some embodiments of the present invention, before inputting the target set and the condition description information into the NLP model, the method further includes: Inputting the condition description information and the reference description information associated with the plurality of candidate doctor information into the NLP model, and obtaining target description information determined by the NLP model from the plurality of reference description information according to the condition description information; The candidate doctor information corresponding to the target description information is determined as the target doctor information.

[0010] According to some embodiments of the present invention, each of the preset parts is associated with at least one preset symptom information, and obtaining the target age, target part and condition description information input by the patient includes: Constructing a first visualization interface, displaying part options and age input items on the first visualization interface, wherein the part options include each of the preset parts; Acquire the target age input in the age input item, and determine at least one of the preset parts selected in the part options as the target part; Determine the preset symptom information associated with the target part as candidate symptom information, and determine at least one prompt description information from the plurality of reference description information based on the candidate symptom information, wherein the candidate symptom information matches the reference symptom information recorded in the prompt description information; A second visualization interface is constructed, and the condition input items and each of the prompt description information are displayed on the second visualization interface, and the condition description information input into the condition input items is obtained.

[0011] According to some embodiments of the present invention, the triage device is preset with a first recommendation word set, a second recommendation word set, and a third recommendation word set, the first recommendation word set includes a plurality of preset pain recommendation words, the second recommendation word set includes a plurality of preset symptom information recommendation words, the third recommendation word set includes a plurality of preset condition recommendation words, each of the preset pain recommendation words, the preset symptom information recommendation words, and the preset condition recommendation words is associated with at least one of the age ranges, and after constructing the second visualization interface, the method further includes: When the prompt description information cannot be matched, determining at least one candidate pain recommendation word from the first recommendation word set based on the target age, determining at least one candidate symptom recommendation word from the second recommendation word set, and determining at least one candidate condition recommendation word from the third recommendation word set; Constructing a first input item in the second visualization interface, displaying the candidate pain recommendation words in the first input item, and acquiring a target pain recommendation word selected in the first input item; Constructing a second input item in the second visual interface, displaying the candidate symptom recommendation word in the second input item, and acquiring the target symptom information recommendation word selected in the second input item; Constructing a third input item in the second visual interface, displaying the candidate condition recommendation words in the third input item, and acquiring a target condition recommendation word selected in the third input item; The target pain recommendation word, the target symptom information recommendation word and the target condition recommendation word are combined into the condition description information.

[0012] In a second aspect, an embodiment of the present invention provides an auxiliary triage device based on condition description information, comprising at least one control processor and a memory for communicating with the at least one control processor; the memory stores instructions executable by the at least one control processor, and the instructions are executed by the at least one control processor so that the at least one control processor can execute the auxiliary triage method based on condition description information as described in the first aspect above.

[0013] In a third aspect, an embodiment of the present invention provides an electronic device, comprising the auxiliary triage device based on condition description information as described in the second aspect above.

[0014] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions are used to execute the auxiliary triage method based on condition description information as described in the first aspect above.

[0015] According to the embodiment of the present invention, the auxiliary triage method based on the condition description information has at least the following beneficial effects: obtaining the target age, target part and condition description information input by the patient; determining a reference set from multiple preset sets based on the target age, and constructing a target set based on multiple conventional descriptors associated with the target part in the reference set; inputting the target set and the condition description information into the NLP model, and obtaining the target descriptor determined by the NLP model from the target set based on the condition description information; determining the preset symptom information associated with the target descriptor as the target symptom information, and determining the triage result based on the target symptom information, wherein the triage result records the target doctor information. According to the technical solution of the embodiment of the present invention, the target part and target age can be used as information indexes, and a target set can be dynamically constructed based on conventional descriptors that conform to the patient's part and description habits, and the risk of misidentification of the NLP model can be reduced by excluding irrelevant classification basis, and conventional descriptors are associated with symptoms, so as to improve the accuracy of triage when the patient is not familiar with professional symptom terms and improve the patient experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a schematic diagram of an auxiliary triage method based on condition description information provided by an embodiment of the present invention; Figure 2 is a flow chart of an auxiliary triage method based on condition description information provided by another embodiment of the present invention; Figure 3 is a schematic diagram of an operation interface of a triage device provided by another embodiment of the present invention; Figure 4 is a complete flow chart of an auxiliary triage method based on condition description information provided by another embodiment of the present invention; Figure 5 It is a structural diagram of an auxiliary triage device based on condition description information provided by another embodiment of the present invention. DETAILED DESCRIPTION

[0017] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0018] In the description of the present invention, it should be understood that descriptions involving orientations, such as up, down, front, back, left, right, etc., and orientations or positional relationships indicated are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0019] In the description of the present invention, "several" means one or more, "more" means more than two, "greater than", "less than", "exceed" etc. are understood as not including the number itself, and "above", "below", "within" etc. are understood as including the number itself. If there is a description of "first" or "second", it is only used for the purpose of distinguishing the technical features, and cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0020] In the description of the present invention, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above terms in the present invention based on the specific content of the technical solution.

[0021] The embodiment of the present invention provides an auxiliary triage method, device, equipment, and medium based on condition description information, wherein the auxiliary triage method based on condition description information includes: obtaining the target age, target part, and condition description information input by the patient; determining a reference set from multiple preset sets based on the target age, and constructing a target set based on multiple conventional descriptors associated with the target part in the reference set; inputting the target set and the condition description information into the NLP model, and obtaining the target descriptor determined by the NLP model from the target set based on the condition description information; determining the preset symptom information associated with the target descriptor as the target symptom information, and determining the triage result based on the target symptom information, wherein the triage result records the target doctor information. According to the technical solution of the embodiment of the present invention, the target part and target age can be used as information indexes, and a target set can be dynamically constructed based on conventional descriptors that conform to the patient's part and description habits, and the risk of misidentification of the NLP model can be reduced by excluding irrelevant classification basis, and conventional descriptors are associated with symptoms, so as to improve the accuracy of triage when the patient is not familiar with professional symptom terms, and improve the patient experience.

[0022] The following is based on the Figure 1 The schematic diagram of the auxiliary triage method based on condition description information shown is used to further illustrate the technical solution of the embodiment of the present invention.

[0023] The embodiment of the present invention provides an auxiliary triage method based on condition description information, which is applied to a triage device. The triage device is preset with an NLP model, multiple preset symptom information and multiple preset sets. Each preset set includes a conventional descriptive word associated with each preset symptom information. The conventional descriptive words associated with the same preset symptom information in at least two preset sets are different. Each conventional descriptive word is associated with at least one preset part. Each preset set corresponds to a different age range. Figure 2 , Figure 2 A flowchart of an auxiliary triage method based on condition description information provided by an embodiment of the present invention includes but is not limited to the following steps: S10, obtaining target age, target part and condition description information input by the patient; S20, determining a reference set from a plurality of preset sets based on the target age, and constructing a target set based on a plurality of conventional descriptors associated with the target part in the reference set; S30, inputting the target set and the condition description information into the NLP model, and obtaining the target description words determined by the NLP model from the target set based on the condition description information; S40, determining the preset symptom information associated with the target descriptor as the target symptom information, and determining the triage result based on the target symptom information, wherein the triage result records the target doctor information.

[0024] It should be noted that the triage device 10 can be a human-computer interaction device installed in the outpatient hall. When the patient is not clear about the department he needs to register, he can enter his age, disease site and condition description information in the triage device after arriving at the outpatient hall. The triage device 10 generates triage results based on the technical solution of this embodiment and assists the patient in registration, thereby improving the patient experience.

[0025] It should be noted that the target age and target part can be input from the visual interface of the triage device, and the specific input method is not limited here. Patients of similar age have a certain degree of similarity in the way they describe their condition. For example, older patients are more vague in describing their condition, and children may have exaggerated elements in describing their condition. Young patients usually describe their condition more accurately. If all commonly used descriptors are used as the classification basis of the NLP model, it will cause misidentification because different age groups use the same commonly used descriptors for different symptom information. For example, for the symptom of "labor pain", the commonly used descriptor for young patients is "labor pain", the commonly used descriptor for older patients is "very painful", and the commonly used descriptor for children is "always painful". For the needle-like "severe pain", the commonly used descriptor for children is "very painful", which is similar to the commonly used descriptors for "labor pain" of older patients, resulting in the risk of misidentification when the NLP model recognizes the description information of the condition. Based on this, since each preset set in this embodiment corresponds to a different age range, different preset sets are used to distinguish the description habits of different age groups. After determining the target age, a preset set is uniquely matched according to the target age as a reference set, reducing the risk of similar idiomatic descriptions for different preset symptom information in the reference set, and providing an information basis for improving the recognition accuracy of the NLP model. At the same time, compared with inputting all idiomatic descriptions for training, it can effectively reduce the number of classification categories, simplify the calculation process, and improve recognition efficiency.

[0026] It should be noted that the conventional descriptive words in this embodiment are not professional medical terms, but colloquial descriptive words. When patients enter the description information of their condition, they may not necessarily use professional medical terms accurately, but will most likely use colloquial descriptions. Since people of the same age group have similar description habits for similar symptoms, at least one conventional descriptive word can be set for each preset symptom information in the same age range according to actual needs. The target descriptive word can be identified through the colloquial conventional descriptive words and the colloquial condition description information, thereby determining the appropriate target doctor information and completing the triage.

[0027] It is worth noting that the triage device of the present embodiment is an auxiliary device used to assist patients who are unable to accurately describe their condition and are not familiar with how to perform registration and triage. Patients who are proficient in describing symptoms, such as patients with long-term follow-up of chronic diseases, can directly complete the registration through the conventional online registration process, which is not within the scope of use of the triage device of the present embodiment. Therefore, the triage device of the present embodiment does not use professional medical terms at all in its conventional descriptive words. The NLP model is not used for accurate symptom recognition, but to recognize the words used in colloquial descriptions, which will not be repeated later.

[0028] It should be noted that the reference set is determined based on the target age, and the reference set records descriptive words for different symptoms, such as pain, pain level, duration of illness, etc. In order to improve the accuracy of triage, this embodiment further obtains the target part from the triage device. The target part is a part of the body. Each preset set corresponds to a preset part. Therefore, according to the target part, a target set can be uniquely determined from multiple reference sets, so that the commonly used descriptive words recorded in the target set are used to describe the symptoms of the target part, avoiding NLP model recognition errors due to the same symptoms in different parts.

[0029] It should be noted that this embodiment does not use the NLP model to diagnose the disease and determine the doctor who needs triage, but excludes inapplicable common descriptors through the two dimensions of target site and target age, so that the target set only records the common descriptors of various symptoms based on the target site of the same age group. Using the common descriptors of the target set as the classification basis of the NLP model can effectively reduce the number of classification categories of the NLP model, eliminate the classification categories that are inevitably wrong, simplify the model training process, and improve the calculation efficiency and accuracy. The specific principles and algorithms of the NLP model are well known to those skilled in the art, and this embodiment does not involve improvements to the NLP algorithm.

[0030] It should be noted that after the target set and the condition description information are input into the NLP model, the NLP model uses the condition description information as the training input and the commonly used descriptive words of the target set as the classification categories, identifies the commonly used descriptive words corresponding to the condition description information as the target descriptive words, and then determines the preset symptom information corresponding to the target descriptive words as the target symptom information. Since the target site is also determined, a suitable doctor can be determined if the symptom information and site are available, and the matched target doctor information is used as the triage result. This embodiment uses the target descriptive words as the basis and the target symptom information as the transit information to determine the target doctor information required for the final triage.

[0031] For example, Figure 1As shown, the patient inputs the target age as age 1 and the target part as part A in the triage device 10. The triage device matches multiple reference sets corresponding to the age range [age 1-age 2], and then determines the target set according to part A. The target set includes idiomatic term 1-1 of symptom information 1, and idiomatic term 2-1 and idiomatic term 2-2 of symptom information 2. Taking symptom information 1 as labor pains and symptom information 2 as severe pain as an example, idiomatic term 1-1 is "painful sometimes and not painful sometimes", idiomatic term 2-1 is "very painful", idiomatic term 2-2 is "extremely painful", and the condition description information is "sometimes painful and sometimes not", so the NLP model will output idiomatic term 1-1 as the target description word, thereby determining symptom information 1 as the target symptom information, and determining doctor information 1 that can process symptom information 1 as the target doctor information, and then generating a triage result to guide the patient to perform the registration operation according to the target doctor information.

[0032] In addition, in one embodiment, referring to Figure 4 After executing step S40, the following steps are also included but not limited to: S51, determining the first descriptive words corresponding to the respective target descriptive words in the condition description information, determining the target descriptive words as the second descriptive words, and determining the target symptom information as the reference symptom information; S52, constructing reference description information based on the first description word, the second description word and the reference symptom information; S53, record the reference description information, and associate the reference description information with the target doctor information.

[0033] It should be noted that the condition description information may contain more descriptive information, and therefore may match multiple target descriptive words. In this embodiment, the word segmentation in the condition description information that matches each target descriptive word is determined as the first descriptive word, the matched target descriptive word is determined as the second descriptive word, and the target symptom information corresponding to the target descriptive word is determined as the reference symptom information. The first descriptive word may be determined by obtaining the recognition log of the NLP model, and the character string used to identify each target descriptive word may be obtained from the recognition log.

[0034] Exemplarily, the disease description information is "the pain after meals every day stops after a while", and the target description word is identified as "after meals" based on "the pain stops after a while", and the target description word is identified as "pains for a while and then stops after a while". The identified target symptom information is "labor pain", and the first description words include "after meals" and "pains for a while and then stops after a while", and the second description words include "after meals" and "pains for a while and then stops after a while".

[0035] It should be noted that this embodiment constructs reference description information based on the first description word, the second description word and the reference symptom information, and associates it with the target doctor information. The target doctor information can be associated with an information set constructed based on the patient's condition description information. When new condition description information is obtained next time, the model recognition can be skipped directly and matched directly with the reference description information. If the content of the text record is the same, it can be determined that the new patient has the same symptoms as the previous patient, and can be directly matched with the same target doctor information, thereby improving information matching efficiency. It can also realize the adaptive collection of matching information, and the corpus associated with the target doctor information is continuously increased with the continuous use of the triage equipment, thereby improving matching accuracy.

[0036] In addition, in one embodiment, referring to Figure 4 The triage device is preset with a plurality of optional doctor information, each optional doctor information is pre-associated with at least one preset part and at least one preset symptom information; in step S40, the triage result is determined based on the target symptom information, specifically including but not limited to the following steps: S41, determining multiple optional doctor information associated with the target part as candidate doctor information, and traversing each candidate doctor information based on the condition description information, wherein at least one candidate doctor information is pre-associated with at least one reference description information; S42, when the first descriptive word associated with the candidate doctor information is recorded in the condition description information, the corresponding candidate doctor information is determined as the target doctor information; S43, when the condition description information fails to successfully match any first description word, determining at least one target doctor information from multiple candidate doctor information based on the target symptom information; S44, generating a triage result based on any target doctor information.

[0037] It should be noted that the triage equipment is equipped with multiple optional doctor information, and each optional doctor information can be dynamically activated according to the doctor's schedule to avoid the inability to complete registration after the triage result is determined. Unless otherwise specified, the subsequent optional doctor information is the doctor information that can perform the registration operation. Doctor information that is not scheduled or has been fully booked will not be determined as optional doctor information.

[0038] It should be noted that, in this embodiment, the candidate doctor information is first filtered out from the optional doctor information through the target part, ensuring that the patient part associated with the candidate doctor information includes the target part. For example, when the target part is the abdomen, the candidate doctor information can be determined by departments such as internal medicine, urology, and endocrinology, where the abdomen is an optional patient part. The subsequent matching range is narrowed through the target part, thereby improving the matching accuracy of the doctor information.

[0039] It should be noted that after determining the candidate doctor information, according to the description of the above embodiment, when each optional doctor information is determined as the target doctor information, the reference description information generated based on the triage result will be associated with the target doctor information. Therefore, with the use of the triage device, each optional doctor information can be associated with multiple reference description information. The first description word carried in the reference description information can be used as a matching basis for the patient's colloquial description. When two patients have the same description of their condition, assigning them the same target doctor information can ensure that the triage result is correct. It is possible to skip complex model calculations and use historical information to improve triage efficiency.

[0040] It should be noted that, in this embodiment, whenever the target age, target part and condition description information input by a patient are obtained, the candidate doctor information is first determined according to the target part, and then the condition description information is used to match the first description word recorded in the candidate doctor information. At this time, there is no need to distinguish different target ages. According to the description of the above embodiment, the condition description information can have multiple participles. If each participle can match the first description word of the same reference description information, then the current patient and the patient corresponding to the successfully matched reference description information have the same description habit. At this time, there is no need to consider the age difference. Instead, the result of the condition description information matching the first description word is used as the basis for the successful description habit matching, and the corresponding candidate doctor information is directly determined as the target doctor information.

[0041] For example, Figure 1 As shown, after the condition description information is obtained, multiple candidate doctor information are screened out according to part A, namely doctor information 1, doctor information 2 and doctor information 3. When the condition description information matches the description word 1 of doctor information 1, doctor information 1 is determined as the target doctor information, and the triage result is determined.

[0042] It should be noted that when the first description word is not successfully matched, that is, at least one word in the condition description information is not recorded in the reference description information of the same candidate doctor information, it cannot be considered that different patients have the same description of the same symptom. At this time, it is necessary to determine the target symptom information according to step 30 and step S40, and match the target doctor information according to the preset symptom information associated with each doctor information, for example Figure 1 As shown, when the target symptom information determined based on the NLP model is symptom information 1, doctor information 1, doctor information 2 and doctor information 3 can be determined as target doctor information. When there are multiple target doctor information, you can select any one to generate a triage result. I will not go into details here.

[0043] In addition, in one embodiment, referring to Figure 4 Before executing step S44, the following steps are also included but not limited to: S431, when the target doctor information cannot be determined based on the condition description information and the target symptom information, a plurality of associated parts pre-associated with the target part are determined from a plurality of preset parts; S432, determining a plurality of optional doctor information associated with the associated part as associated doctor information, wherein the associated doctor information does not include any candidate doctor information; S433, determining target doctor information from multiple associated doctor information based on the target symptom information and the condition description information.

[0044] It should be noted that when the target doctor information is determined based on the condition description information and the target symptom information, it can be determined that the current patient has incorrectly described the condition when using the triage equipment. For some symptoms, pain radiation is likely to occur due to the association between parts. For example, toothache can cause headaches due to nerve association. This embodiment further sets multiple associated parts in each preset part. After determining the target part, the associated parts are used to filter out the associated doctor information that has been successfully matched. This can serve as a compensation mechanism for the use of incorrect descriptions in the condition description information, thereby improving the matching success rate of the triage results.

[0045] It is worth noting that since the candidate doctor information has failed to match, the associated doctor information does not include the previously determined candidate doctor information. The candidate doctor information can be marked first. After matching multiple associated doctor information according to the associated parts, the marked associated doctor information can be removed to ensure that no reference doctor information is included.

[0046] It should be noted that, for determining the target doctor information from the associated doctor information based on the target symptom information and the condition description information, reference can be made to the description of steps S42 and S43 above, for example, by matching the reference description information of the associated doctor information according to the condition description information, or by determining whether the target symptom information belongs to the preset symptom information of the associated doctor information. Details will not be repeated here.

[0047] In addition, in one embodiment, referring to Figure 4 Before executing step S30, the following steps are also included but not limited to: S61, inputting the condition description information and the reference description information associated with the multiple candidate doctor information into the NLP model, and obtaining the target description information determined by the NLP model from the multiple reference description information according to the condition description information; S62, determining the candidate doctor information corresponding to the target description information as the target doctor information.

[0048] It should be noted that, according to the description of steps S20 and S30, the common descriptive words recorded in the target set can be used as the classification basis to realize the recognition of the condition description information. When the number of common descriptive words recorded in the target set is large, it is very likely that misrecognition will occur due to the close semantics of multiple common descriptive words. In this embodiment, the NLP model is first configured with the second descriptive word of the reference description information. Since the second descriptive word is the target descriptive word determined in advance, if the NLP model can identify the second descriptive word based on the condition description information at this time, the reference description information of the identified second descriptive word is determined as the target description information, and the corresponding candidate doctor information is determined as the target doctor information.

[0049] It is worth noting that step S30 of this embodiment and the above-mentioned embodiment both utilizes the NLP model to identify descriptive words. Since the classification categories of this embodiment are further reduced compared to the target set in step S30, this embodiment can set a first similarity threshold for the NLP model and set a second similarity threshold in step S30. The first similarity threshold is greater than the second similarity threshold, so that this embodiment can perform information recognition through a higher similarity threshold based on fewer classification categories. If the matching of the target description information can still be completed at this time, it can be determined that the second descriptive word in the target description information is relatively similar to the condition description information, and can be determined as the target description information and the target doctor information can be determined.

[0050] In addition, in one embodiment, referring to Figure 4 Each preset part is associated with at least one preset symptom information, and step S10 specifically includes but is not limited to the following steps: S11, constructing a first visualization interface, displaying part options and age input items on the first visualization interface, wherein the part options include various preset parts; S12, obtaining a target age input in the age input item, and determining at least one preset part selected in the part options as the target part; S13, determining the preset symptom information associated with the target part as candidate symptom information, and determining at least one prompt description information from multiple reference description information based on the candidate symptom information, wherein the candidate symptom information matches the reference symptom information recorded in the prompt description information; S14, constructing a second visualization interface, displaying the condition input items and various prompt description information on the second visualization interface, and acquiring the condition description information input into the condition input items.

[0051] It should be noted that each preset part can be associated with any number of preset symptom information, which can be pre-configured according to actual needs. This embodiment uses the target part to filter out candidate symptom information, excludes preset symptom information that the target part does not have, filters out wrong selections from the information selection level, and improves the accuracy of information matching.

[0052] It should be noted that if Figure 3 As shown, a first visualization interface can be generated in the triage device 10, in which a part selection 11 and an age input item 12 are displayed. The target selection can be in the form of a drop-down menu, and the preset parts are the various display items of the drop-down menu. Multiple target parts can be selected through a multiple selection operation, which is not limited here.

[0053] It should be noted that in step S13, this embodiment uses the preset symptom information associated with the target part as candidate symptom information. According to the description of the above embodiment, each reference description information includes the reference symptom information determined at that time. This embodiment uses the candidate symptom information as an index and determines the matched multiple reference description information as prompt description information.

[0054] It should be noted that, in this embodiment, the condition input items and prompt description information are displayed in the second visual interface, such as Figure 3 As shown in the example interface 1, each candidate symptom information and the corresponding prompt description information are displayed in the example interface 1, and the patient is prompted to describe his or her own condition through visual text, thereby improving the accuracy of the condition description information. For example, when the patient sees the prompt description 1, the patient subjectively judges that it is similar to his or her own symptoms, and the prompt description 1 can be directly selected as the condition description information. At this time, the candidate symptom information 1 can be directly determined as the target symptom information, and the subsequent target doctor information can be determined.

[0055] It is worth noting that when recording reference description information, the target age can also be saved as the reference age of the reference description information, so that after entering the new target age, the prompt description information can be filtered out in combination with the target age to ensure that the description prompts for the patient can be in line with the patient's age group, lower the patient's understanding threshold, and improve the input accuracy of the condition description information.

[0056] In addition, in one embodiment, referring to Figure 4 The triage device is preset with a first recommendation word set, a second recommendation word set and a third recommendation word set, the first recommendation word set includes a plurality of preset pain recommendation words, the second recommendation word set includes a plurality of preset symptom information recommendation words, the third recommendation word set includes a plurality of preset condition recommendation words, each preset pain recommendation word, preset symptom information recommendation word and preset condition recommendation word is associated with at least one age range, in step S14, after constructing the second visualization interface, further including but not limited to the following steps: S141, when the prompt description information is not matched, determining at least one candidate pain recommendation word from the first recommendation word set, determining at least one candidate symptom recommendation word from the second recommendation word set, and determining at least one candidate condition recommendation word from the third recommendation word set based on the target age; S142, constructing a first input item on the second visualization interface, displaying candidate pain recommendation words in the first input item, and obtaining a target pain recommendation word selected in the first input item; S143, constructing a second input item on the second visual interface, displaying candidate symptom recommendation words in the second input item, and obtaining a target symptom information recommendation word selected in the second input item; S144, constructing a third input item in the second visualization interface, displaying candidate condition recommendation words in the third input item, and obtaining a target condition recommendation word selected in the third input item; S145, combining the target pain recommendation words, the target symptom information recommendation words and the target condition recommendation words into condition description information.

[0057] It should be noted that when the prompt description information cannot be matched, it can be determined that there is no previous patient describing the same condition. On this basis, this embodiment further guides the patient to enter accurate condition description information through the first input item, the second input item and the third input item.

[0058] It should be noted that the triage device may be preset with a first set of recommended words, a second set of recommended words, and a third set of recommended words, which correspond to pain, symptoms, and conditions, respectively. The preset pain recommendation words are used as a reference for pain description, for example Figure 3 As shown, the first recommendation word set includes pain recommendation word 1, pain recommendation word 2, pain recommendation word 3, pain recommendation word 4 and pain recommendation word 5. For example, pain recommendation word 1 is "not much pain", pain recommendation word 2 is "general pain, occasional pain", pain recommendation word 3 is "general pain, always painful", pain recommendation word 4 is "very painful, it hurts after eating", and pain recommendation word 5 is "very painful, it hurts all the time". This embodiment uses common descriptive words to identify and classify the patient's colloquial description information. Therefore, the preset pain recommendation words, preset symptom recommendation words and preset condition recommendation words should also be colloquial terms to ensure that they can meet the usage habits of patients who are not familiar with professional medical terms. The preset symptom recommendation words can be visible symptom descriptions such as "itchy skin" and "acne on the skin". The preset condition recommendation words are used to describe the duration of the condition and the timing of onset, such as "lasting for a month" and "starting to hurt after eating xx food", etc., which are not repeated here.

[0059] It should be noted that, according to the description of step S20, this embodiment is based on the different description habits of different age groups, so as to eliminate incorrect classification categories of the NLP model. Therefore, after determining the target age, this embodiment matches the corresponding content from the first description word set, the second description word set and the third description word set. For example, each preset pain description word is associated with an age range, and the candidate pain description words are matched by matching the age range, and the rest are similar.

[0060] It should be noted that this embodiment displays the first input item, the second input item and the third input item in sequence instead of displaying them simultaneously. The strategy of displaying them in sequence can gradually guide the patient to accurately describe his condition from different dimensions, avoiding affecting the accuracy of the description when considering multiple dimensions at the same time.

[0061] For example, Figure 3 As shown in example interface 2, after inputting the target part and target age, the first input item is first displayed on the second visualization interface, and each candidate pain description word is displayed through the first input item. After obtaining the target pain recommendation word, the second input item is displayed in the second visualization interface, and each candidate symptom descriptor word is displayed through the second input item. After obtaining the target symptom description word, the third input item is displayed in the second visualization interface, and each candidate condition recommendation word is displayed through the third input item. Finally, the condition description information is combined according to the patient's selection in each input item to improve the accuracy of subsequent recognition.

[0062] like Figure 5 As shown, Figure 5 : is a structural diagram of an auxiliary triage device based on condition description information provided by an embodiment of the present invention. The present invention also provides an auxiliary triage device based on condition description information, including: The processor 401 may be implemented by a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present application; The memory 402 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 402 can store an operating system and other application programs. When the technical solution provided in the embodiment of this specification is implemented by software or firmware, the relevant program code is stored in the memory 402, and the processor 401 calls and executes the auxiliary triage method based on the condition description information of the embodiment of this application; Input / output interface 403, used to implement information input and output; Communication interface 404, used to realize communication interaction between the device and other devices, which can be realized through wired mode (such as USB, network cable, etc.) or wireless mode (such as mobile network, WIFI, Bluetooth, etc.); A bus 405 that transmits information between various components of the device (e.g., processor 401, memory 402, input / output interface 403, and communication interface 404); The processor 401 , the memory 402 , the input / output interface 403 and the communication interface 404 are connected to each other in communication within the device via the bus 405 .

[0063] An embodiment of the present application also provides an electronic device, including the auxiliary triage device based on condition description information as described above.

[0064] An embodiment of the present application also provides a storage medium, which is a computer-readable storage medium. The storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned auxiliary triage method based on condition description information is implemented.

[0065] As a non-transient computer-readable storage medium, the memory can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage devices. In some embodiments, the memory may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof. The device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and are implemented to be located in one place, or may also be distributed to multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment.

[0066] It will be appreciated by those skilled in the art that all or some of the steps and systems in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or transient medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically include computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0067] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the above-mentioned implementation mode. Technical personnel familiar with the field can also make various equivalent deformations or substitutions under the shared conditions without violating the spirit of the present invention. These equivalent deformations or substitutions are all included in the scope defined by the claims of the present invention.

Claims

1. An auxiliary triage method based on condition description information, characterized in that: The method is applied to a triage device, wherein the triage device is preset with an NLP model, a plurality of preset symptom information, and a plurality of preset sets, each of the preset sets includes a conventional descriptor associated with each of the preset symptom information, the conventional descriptors associated with the same preset symptom information in at least two of the preset sets are different, each of the conventional descriptors is associated with at least one preset part, and each of the preset sets corresponds to a different age range, and the method comprises: Obtain target age, target site and condition description information input by the patient; Determine a reference set from a plurality of preset sets based on the target age, and construct a target set based on a plurality of the conventional descriptors associated with the target part in the reference set; Inputting the target set and the condition description information into the NLP model, and obtaining the target description words determined by the NLP model from the target set based on the condition description information; The preset symptom information associated with the target descriptor is determined as target symptom information, and a triage result is determined based on the target symptom information, wherein the triage result records target doctor information.

2. The auxiliary triage method based on condition description information according to claim 1, characterized in that: After determining the triage result based on the target symptom information, the method further includes: Determining the first descriptive words corresponding to the respective target descriptive words in the condition description information, determining the target descriptive words as second descriptive words, and determining the target symptom information as reference symptom information; constructing reference description information based on the first descriptor, the second descriptor and the reference symptom information; The reference description information is recorded, and the reference description information is associated with the target doctor information.

3. The auxiliary triage method based on condition description information according to claim 2, characterized in that: The triage device is preset with a plurality of optional doctor information, each of the optional doctor information is pre-associated with at least one of the preset parts and at least one of the preset symptom information; Determining a triage result based on the target symptom information includes: Determine the multiple optional doctor information associated with the target part as candidate doctor information, and traverse each of the candidate doctor information based on the condition description information, wherein at least one of the candidate doctor information is pre-associated with at least one of the reference description information; When the first descriptive word associated with the candidate doctor information is recorded in the condition description information, the corresponding candidate doctor information is determined as the target doctor information; When the condition description information fails to successfully match any of the first description words, determining at least one target doctor information from the plurality of candidate doctor information based on the target symptom information; The triage result is generated based on any of the target doctor information.

4. The auxiliary triage method based on condition description information according to claim 3, characterized in that: Before generating the triage result based on any of the target doctor information, the method further includes: When the target doctor information cannot be determined based on the condition description information and the target symptom information, determining a plurality of associated parts pre-associated with the target part from the plurality of preset parts; determining the plurality of optional doctor information associated with the associated part as associated doctor information, wherein the associated doctor information does not include any of the candidate doctor information; The target doctor information is determined from the plurality of associated doctor information based on the target symptom information and the condition description information.

5. The auxiliary triage method based on condition description information according to claim 3, characterized in that: Before inputting the target set and the condition description information into the NLP model, the method further includes: Inputting the condition description information and the reference description information associated with the plurality of candidate doctor information into the NLP model, and obtaining target description information determined by the NLP model from the plurality of reference description information according to the condition description information; The candidate doctor information corresponding to the target description information is determined as the target doctor information.

6. The auxiliary triage method based on condition description information according to claim 2, characterized in that: Each of the preset parts is associated with at least one preset symptom information, and the target age, target part and condition description information input by the patient are obtained, including: Constructing a first visualization interface, displaying part options and age input items on the first visualization interface, wherein the part options include each of the preset parts; acquiring the target age input in the age input item, and determining at least one of the preset parts selected in the part options as the target part; Determine the preset symptom information associated with the target part as candidate symptom information, and determine at least one prompt description information from the plurality of reference description information based on the candidate symptom information, wherein the candidate symptom information matches the reference symptom information recorded in the prompt description information; A second visualization interface is constructed, and the condition input items and each of the prompt description information are displayed on the second visualization interface, and the condition description information input into the condition input items is obtained.

7. The auxiliary triage method based on condition description information according to claim 6, characterized in that: The triage device is preset with a first recommendation word set, a second recommendation word set and a third recommendation word set, the first recommendation word set includes a plurality of preset pain recommendation words, the second recommendation word set includes a plurality of preset symptom information recommendation words, the third recommendation word set includes a plurality of preset condition recommendation words, each of the preset pain recommendation words, the preset symptom information recommendation words and the preset condition recommendation words is associated with at least one of the age ranges, and after constructing the second visualization interface, the method further includes: When the prompt description information cannot be matched, determining at least one candidate pain recommendation word from the first recommendation word set based on the target age, determining at least one candidate symptom recommendation word from the second recommendation word set, and determining at least one candidate condition recommendation word from the third recommendation word set; Constructing a first input item in the second visualization interface, displaying the candidate pain recommendation words in the first input item, and acquiring a target pain recommendation word selected in the first input item; Constructing a second input item in the second visual interface, displaying the candidate symptom recommendation word in the second input item, and acquiring the target symptom information recommendation word selected in the second input item; Constructing a third input item in the second visual interface, displaying the candidate condition recommendation words in the third input item, and acquiring a target condition recommendation word selected in the third input item; The target pain recommendation word, the target symptom information recommendation word and the target condition recommendation word are combined into the condition description information.

8. An auxiliary triage device based on condition description information, characterized in that: It includes at least one control processor and a memory for communicating with the at least one control processor; the memory stores instructions that can be executed by the at least one control processor, and the instructions are executed by the at least one control processor so that the at least one control processor can execute the auxiliary triage method based on condition description information as described in any one of claims 1 to 7.

9. An electronic device, characterized in that: Including the auxiliary triage device based on condition description information as described in claim 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the auxiliary triage method based on condition description information as described in any one of claims 1 to 7.

Images