Medical record data processing method and device, electronic equipment and readable storage medium

Abstract

The embodiment of the invention provides a medical record data evaluation method and device, electronic equipment and a readable storage medium. The method comprises the following steps: acquiring a preset evaluation rule and a first text and a medical record of a to-be-evaluated medical record; identifying the doctor-seeing record to obtain a second text corresponding to the doctor-seeing record; inputting the first text, the second text and the preset evaluation rule into a text to generate a large model, obtaining a first evaluation result of the first text and a second evaluation result of the second text, and comparing the first evaluation result with the second evaluation result to obtain a comparison result, the comparison result is used for determining whether the first text corresponds to the doctor-seeing record or not, so that the situation that the content of the text corresponding to the doctor-seeing record is inconsistent with the content of the medical record text written by a doctor is accurately recognized, and an evaluation result can be output to assist in targeted modification of the medical record text and the doctor-seeing record manually.

Description

Technical Field

[0001] The present invention belongs to the field of network technology, and in particular relates to a medical record data processing method, device, electronic equipment and readable storage medium. Background Art

[0002] Medical records are important documents in the patient diagnosis and treatment process. In the prior art, in order to improve the work efficiency of doctors, traditional paper medical records have been gradually abandoned and replaced by electronic medical records. Specifically, when or after seeing a patient, the doctor can enter the patient's diagnosis information and medical recordings on the medical record system of an electronic device, and identify the diagnosis information and medical recordings based on the electronic medical record generation model to generate an electronic medical record.

[0003] However, there are two difficulties in the electronic medical record generation model. One is that it is difficult to collect training data and the quality is low. The other is that after the electronic medical record generation model is trained, the actual input recording quality is low, and it is easy for the input data to have low quality and incomplete content, which leads to difficulties in collecting appropriate input data. For example, the audio quality, transcription effect, and content of outpatient recordings will affect the input data quality. At the same time, the medical record text written by the doctor is not based on the consultation recording, but is manually entered into the electronic device based on the actual outpatient results. Therefore, there will be a large number of discrepancies between the transcription results and the medical records written by the doctor, which will affect the input data quality. Summary of the invention

[0004] The present invention provides a medical record data processing method, device, electronic device and readable storage medium to solve the problem of inconsistency between the recording corresponding text and the medical record text input into the electronic medical record generation model due to the complexity of recording and medical record text collection.

[0005] In order to solve the above-mentioned technical problems, the present invention is achieved as follows:

[0006] In a first aspect, the present invention provides a method for processing medical record data, the method comprising:

[0007] Obtaining preset evaluation rules and the first text of the medical records to be evaluated and the consultation recordings;

[0008] Recognize the medical consultation recording to obtain a second text corresponding to the medical consultation recording;

[0009] The first text, the second text and the preset evaluation rules are input into a text generation model to obtain a first evaluation result of the first text and a second evaluation result of the second text, and the first evaluation result is compared with the second evaluation result to obtain a comparison result, which is used to determine whether the first text and the medical recording correspond.

[0010] In a second aspect, the present invention provides a medical record data processing device, the device comprising:

[0011] An acquisition module, used to acquire preset evaluation rules and the first text of the medical record to be evaluated and the consultation recording;

[0012] A recording recognition module, used to recognize the medical consultation recording and obtain a second text corresponding to the medical consultation recording;

[0013] A processing module is used to input the first text, the second text and the preset evaluation rules into a text generation model, obtain a first evaluation result of the first text, a second evaluation result of the second text, and compare the first evaluation result with the second evaluation result to obtain a comparison result, and the comparison result is used to determine whether the first text and the medical recording correspond.

[0014] In a third aspect, the present invention provides an electronic device comprising: a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned medical record data processing method when executing the program.

[0015] In a fourth aspect, the present invention provides a readable storage medium, when the instructions in the storage medium are executed by a processor of an electronic device, the electronic device can execute the above-mentioned medical record data processing method

[0016] In an embodiment of the present invention, a preset evaluation rule and a first text of a medical record to be evaluated and a medical recording are obtained; the medical recording is identified to obtain a second text corresponding to the medical recording; the first text, the second text and the preset evaluation rule are input into a text generation large model to obtain a first evaluation result of the first text and a second evaluation result of the second text, and the first evaluation result is compared with the second evaluation result to obtain a comparison result, which is used to determine whether the first text and the medical recording correspond, thereby generating a large model based on the text, and using the evaluation rules to evaluate and compare the first text of the medical record and the second text corresponding to the medical recording, accurately identifying the inconsistency between the text corresponding to the medical recording and the medical record text written by the doctor, and outputting the evaluation result to make targeted modifications to the medical record text and the medical recording. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 This is one of the step flow charts of a medical record data evaluation method provided by an embodiment of the present invention;

[0019] Figure 2 This is the second step flow chart of a medical record data evaluation method provided by an embodiment of the present invention;

[0020] Figure 3 is a structural diagram of a medical record data evaluation device provided by an embodiment of the present invention;

[0021] Figure 4 It is a structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] Figure 1 is a flowchart of a method for evaluating medical record data provided by an embodiment of the present invention. Figure 1 As shown, the method may include:

[0024] Step 101: Obtain preset evaluation rules and the first text of the medical record to be evaluated and the consultation recording.

[0025] Medical records are records of the occurrence, development, and outcome of a patient's disease, as well as the process of medical examinations, diagnosis, and treatment. The first text of a medical record is the text written during the medical treatment process. If the medical staff writes the medical record on paper, the first text of the medical record is the text obtained after the handwritten medical record is recognized. If the medical staff enters the electronic medical record on an electronic device, the first text of the medical record is the text entered into the electronic medical record.

[0026] Medical recording refers to the audio recording of relevant communication content during the patient's medical treatment process. Medical recording can be collected by audio collection equipment during the medical treatment process.

[0027] In the example of the present invention, the preset evaluation rule is used to evaluate the first text of the medical record and the consultation recording. The preset evaluation rule can be preset.

[0028] Step 102: Identify the medical consultation recording to obtain a second text corresponding to the medical consultation recording.

[0029] Since the medical recording is an audio record, the medical recording is recognized by voice to obtain the second text corresponding to the medical recording. Among them, the medical recording can be recognized by automatic speech recognition (Automatic Speech Recognition, ASR) to obtain the second text corresponding to the medical recording. ASR can quickly convert voice audio into text without manual text input. Especially in the process of medical treatment, there is a lot of communication between doctors and patients, and the amount of data in the medical recording is large. ASR is more suitable for processing medical recordings.

[0030] Step 103: Input the first text, the second text, and the preset evaluation rules into a text generation model to obtain a first evaluation result of the first text and a second evaluation result of the second text, and compare the first evaluation result with the second evaluation result to obtain a comparison result, which is used to determine whether the first text and the medical recording correspond.

[0031] The text generation model is an artificial intelligence model based on deep learning technology that can understand and generate natural language text.

[0032] In the example of the present invention, step 103 includes: step 1031 - step 1032.

[0033] Step 1031, generating a large model based on the text, extracting the first text and the second text respectively, and obtaining a target text corresponding to the first text and a target text corresponding to the second text.

[0034] After inputting the first text, the second text and the preset evaluation rules into the text generation model, the text generation model can extract the target text from the first text and the second text, thereby obtaining the target text corresponding to the first text and the target text corresponding to the second text.

[0035] For example, for the second text, since the second text is recognized from a medical consultation recording, and the medical consultation recording may contain content that is unrelated to the condition, such as the casual chat between the doctor and the patient in the medical consultation recording, which is unrelated to the condition, the second text can be extracted to retain the target text and avoid the subsequent use of content that is unrelated to the condition.

[0036] Step 1032, using the preset evaluation rules, respectively evaluate the target text corresponding to the first text and the target text corresponding to the second text to obtain the first text requirement of the target text corresponding to the first text and the second text requirement of the target text corresponding to the first text.

[0037] The preset evaluation rules include at least one text requirement, and the text requirement is a specific rule in the preset evaluation rules for evaluating the number of words in the text and the content of the text.

[0038] For the number of words in the text, the text requirement may be at least one of: the total number of words in the text counted, and the total number of words in the text being within N words (N is a positive integer).

[0039] For text content, the text requirements may be: whether there are typos, the number of typos is within M characters, M is an integer, whether the logic is not smooth, and whether it contains at least one of the description of the disease.

[0040] The text generation model evaluates the target text corresponding to the first text and the target text corresponding to the second text through preset evaluation rules, and obtains the first text requirements that the target text corresponding to the first text in the preset evaluation rules conforms to, and the second text requirements that the target text corresponding to the second text in the preset evaluation rules conforms to.

[0041] For example, the target text corresponding to the first text is: "Headache lasts for 1 week, allergic to cyanide, CT scan required, hand fracture in 20XX". Text requirement 1: count the total number of words in the text, text requirement 2: the total number of words in the text is within 15 words, text requirement 3: the number of typos is within 2 words, logical coherence, and contains at least one of the following descriptions of the condition. Then the first text requirement corresponding to the target text is text requirement 1, text requirement 3, and because the total number of words in the text is 17 words, which is greater than 15 words, text requirement 2 is not the text requirement corresponding to the target text.

[0042] In another example of the present invention, step 103 includes: step 1033 - step 1034.

[0043] Step 1033: If the semantics of the target text corresponding to the first text is consistent with the semantics of the target text corresponding to the second text, it is determined that the first text is consistent with the medical recording.

[0044] In the example of the present invention, the target text directly reflects the contents of the first text and the second text. By comparing the semantics of the target text, it can be directly determined whether the semantics of the target text corresponding to the first text is consistent with the semantics of the target text corresponding to the second text. If the semantics of the target text corresponding to the first text is consistent with the semantics of the target text corresponding to the second text, it means that the first text and the medical recording are consistent with the information recorded during the medical consultation.

[0045] When comparing the semantics of the target text corresponding to the first text with the semantics of the target text corresponding to the second text, the feature vector corresponding to the target text can be determined through a bag-of-words model or a word embedding model. When the feature vector corresponding to the target text of the first text and the feature vector corresponding to the target text of the second text are within a threshold range in the vector space, it means that the semantics of the target text corresponding to the first text are consistent with the semantics of the target text corresponding to the second text. Of course, those skilled in the art can also use other methods to determine whether the semantics of the target text corresponding to the first text are consistent with the semantics of the target text corresponding to the second text, and the present invention does not limit this.

[0046] Step 1034, if the semantics of the target text corresponding to the first text are not completely consistent with the semantics of the target text corresponding to the second text, prompts are given for the target text corresponding to the first text, the target text corresponding to the second text, the first text requirement corresponding to the first target text, and the second text requirement corresponding to the second target text.

[0047] When the semantics of the target text corresponding to the first text are not completely consistent with the semantics of the target text corresponding to the second text, it means that the target text corresponding to the first text and the target text corresponding to the second text are inconsistent, and the second text is obtained by recognizing the medical recording. Therefore, the medical recording and the first text are inconsistent.

[0048] In an example of the present invention, when the semantics of the target text corresponding to the first text are not completely consistent with the semantics of the target text corresponding to the second text, a prompt will be given for the incompletely consistent content, so that it can be quickly determined whether the medical recording and the first text correspond according to the prompt and the medical recording and the first text can be modified according to the prompt.

[0049] In the example of the present invention, the evaluation result includes: a target text and a text requirement corresponding to the target text.

[0050] The above scheme obtains the first text of the medical record, the second text corresponding to the consultation recording, and the preset evaluation rules, and inputs the first text, the second text, and the preset evaluation rules into the text generation model. The text generation model evaluates the first text and the second text based on the preset evaluation rules to obtain a first evaluation result of the first text and a second evaluation result of the second text, and compares the first evaluation result with the second evaluation result to obtain a comparison result, so as to conveniently determine whether the first text of the medical record corresponds to the consultation recording based on the comparison result, so as to quickly understand whether there is any content inconsistency between the second text corresponding to the recording and the first text corresponding to the medical record written by the doctor.

[0051] Figure 2 is a flowchart of another method for evaluating medical record data provided by an embodiment of the present invention. Figure 2 As shown, the method may include: steps 201 to 203, wherein steps 201 to 203 are applied after step 1034.

[0052] Step 201: Determine the score value of the first text requirement and the score value of the second text requirement that are not completely semantically consistent according to the preset evaluation rule.

[0053] The preset evaluation rules include not only at least one text requirement but also a scoring standard corresponding to each text requirement. The scoring standard corresponding to each sub-text requirement is used to determine the scoring value of the text requirement corresponding to the target text.

[0054] For example, text requirement 1: the total number of words in the text is within N words. The scoring criteria for this text requirement 1 are: if the total number of words in the text is less than N, then 5 points; if the total number of words in the text is greater than N and less than 2N, then 3 points; if the total number of words in the text is greater than 2N, then 1 point.

[0055] For example, text requirement 2: the number of typos is within M words. The corresponding scoring criteria for text requirement 2 are: if the number of typos is less than M, 5 points; if the number of typos is greater than M and less than 2M, 3 points; if the number of typos is greater than 2M, 1 point.

[0056] For another example, text requirement 3: The target text is extracted correctly. The corresponding scoring criteria for this text requirement 3 are: 5 points if it is extracted correctly; 3 points if part of it is extracted correctly and the other part is not extracted correctly; 1 point if it is not extracted correctly.

[0057] Therefore, when the target text corresponding to the first text, the target text corresponding to the second text, the first text requirement corresponding to the first target text, and the second text requirement corresponding to the second target text are obtained, if the semantics of the target text corresponding to the first text are not completely consistent with the semantics of the target text corresponding to the second text, then the scoring value of the first text requirement and the scoring value of the second text requirement are determined according to the preset evaluation rules.

[0058] In the example of the present invention, the first target text can correspond to at least one text requirement, and the second target text can correspond to at least one text requirement. Therefore, based on the scoring criteria corresponding to each text requirement, the scoring value of each first text requirement corresponding to the first target text and the scoring value of each second text requirement corresponding to the second target text can be determined, and the sum of the scoring values ​​of each first text requirement is used as the scoring value of the first text, and the sum of the scoring values ​​of each second text requirement is used as the scoring value of the second text requirement.

[0059] Step 202: compare the score of the first text requirement and the score of the second text requirement, and use the text requirement with the lower score between the first text requirement and the second text requirement as the third text requirement.

[0060] Step 203: determine the first text or the medical consultation recording corresponding to the third text requirement, and modify the first text or the medical consultation recording corresponding to the third text requirement.

[0061] The third text requirement is obtained by determining the text requirement with a lower score value between the first text requirement and the second text requirement, and then determining the target text corresponding to the third text requirement. Based on the third text requirement corresponding to the target text, it is determined which of the first text of the medical record to be evaluated and the medical recording has lower data quality, so that the first text of the medical record and the medical recording with lower data quality are modified, and the modified first text, medical recording, and preset evaluation rules are input into the text generation big model to repeat step 103 until the comparison result output by the text generation big model indicates that the modified first text of the medical record and the medical recording correspond.

[0062] In the example of the present invention, the first text and the second text are evaluated by scoring values, thereby evaluating the quality of the first text and the medical recording, and assisting in modification based on the evaluation values.

[0063] Specifically, when the semantics of the target text corresponding to the first text and the target text corresponding to the second text are not completely consistent, prompts are given to the target text corresponding to the first text, the target text corresponding to the second text, the first text requirement corresponding to the first text, and the second text requirement corresponding to the second text, and the target text corresponding to the first text and the target text corresponding to the second text that are not completely consistent in semantics are modified. Since prompts have been given to the texts that are not completely consistent, the prompts can facilitate manual modification of the prompt content.

[0064] In another example of the present invention, when a modified first text and a medical recording are obtained, and the modified first text and the medical recording correspond to the second text, it means that the modified first text and the modified medical recording are good training samples. The modified first text and the modified medical recording can be used as training samples for the corresponding task of manual annotation in the electronic medical record generation model, and the electronic medical record generation model can be trained to obtain a trained electronic medical record generation model.

[0065] In another example of the present invention, the target text includes: at least one target sub-text; the target sub-text type includes: any one of the chief complaint text type, the present medical history text type, the past medical history text type, and the allergy history text type.

[0066] The target subtext corresponding to the chief complaint text type is the chief complaint text, which is a description of the symptoms and symptom duration of the patient's visit in the medical record.

[0067] The target subtext corresponding to the current medical history text type is the current medical history text, which is a description of the development and evolution of the patient's condition from the onset of the disease to the current visit.

[0068] The target subtext corresponding to the past history text type is the past history text, which is a description of the diseases the patient has suffered from and the surgical records the patient has undergone.

[0069] The target subtext corresponding to the allergy history text type is the allergy history text, which is a description of the patient's allergies to various drugs.

[0070] Since the target text includes: at least one target sub-text, when extracting the first text and the second text, the extraction can be performed according to the target sub-text type, so that each target sub-text corresponds to a target sub-text type.

[0071] For example, when extracting the first text, you can extract the text related to the chief complaint in the first text to obtain the target subtext of the chief complaint text type, extract the text related to the current medical history to obtain the target subtext of the current medical history text type, extract the text related to the past medical history to obtain the target subtext of the past medical history text type, extract the text related to the allergy history to obtain the target subtext of the allergy history text type. Similarly, the second text can also be extracted based on the target subtext type, which will not be elaborated here.

[0072] In the example of the present invention, the preset evaluation rules include subtext requirements corresponding to each target subtext type.

[0073] In the example of the present invention, step 1032 includes:

[0074] Step 10321, using the subtext requirements corresponding to each target subtext type, respectively evaluate the target subtext in the first text and the target subtext in the second text to obtain a first subtext requirement and a second subtext requirement.

[0075] Since different target subtext types have different functions in corresponding target subtexts, the content and word count of target subtexts corresponding to different target subtext types are different, and corresponding subtext requirements can be set for target subtexts of different target subtext types. Thus, the target subtexts can be evaluated by using the subtext requirements corresponding to each target subtext type, so as to make a more detailed evaluation of different parts of the target text.

[0076] In the example of the present invention, step 1034 includes:

[0077] Step 10341, for each target sub-text of the same target sub-text type, if the semantics of the target sub-text corresponding to the first text is inconsistent with the semantics of the target sub-text corresponding to the second text, prompts are given for the target sub-text corresponding to the first text, the target sub-text corresponding to the second text, the first sub-text requirements, and the second sub-text requirements.

[0078] By comparing the semantics of the target sub-texts of the same target sub-text type in the first text and the second text, it can be determined whether the semantics of the target sub-text corresponding to the first text is consistent with the semantics of the target sub-text corresponding to the second text. The specific process of comparing the semantics of the target sub-text can be referred to step 1033 and will not be repeated here.

[0079] If the semantics of the target subtexts of the same target subtext type in the first text and the second text are the same, it means that the contents of the target subtexts of the target subtext type in the first text and the medical recording are consistent.

[0080] If the semantics of the target sub-text of the same target sub-text type in the first text and the second text do not match, the target sub-text corresponding to the first text under the target sub-text type, the target sub-text corresponding to the second text, the first sub-text requirements, the second sub-text requirements can be output by a large text generation model, and prompts can be given.

[0081] Specifically, the target subtext and subtext requirements can be highlighted during the prompt, or the non-conforming target subtext and subtext requirements can be displayed in the form of a comparison table. The comparison table includes: target subtext type, target subtext, and the position of the target subtext in the first text or the second text. The specific situation of the non-conforming target subtext can be intuitively seen through the comparison table. The comparison table is shown in Table 1.

[0082] Table 1:

[0083]

[0084] In the example of the present invention, step 201 includes:

[0085] Step 2011: Determine the scoring value of the first subtext requirement and the scoring value of the second subtext requirement for each target subtext type.

[0086] The preset evaluation rules include not only the sub-text requirements but also the scoring criteria corresponding to the sub-text requirements.

[0087] Subtext requirements are word count requirements and content requirements for subtexts. For example, the subtext requirements set for the chief complaint text type may be word count requirements and content requirements for the chief complaint text. For example, the subtext requirements set for the chief complaint text type may be whether the chief complaint text contains the name of the disease, whether the chief complaint text contains the duration of the disease, and whether the number of words in the chief complaint text is X, where X is a positive integer.

[0088] The scoring criteria corresponding to each sub-text requirement are used to determine the scoring value of the sub-text requirement that meets the target sub-text. For example, the scoring criteria corresponding to the sub-text requirement of the chief complaint text type setting are: if the chief complaint text contains the name of the disease, it will score 5 points, if the chief complaint text does not contain the name of the disease, it will score 0 points; if the chief complaint text contains the duration of the disease, it will score 3 points, if the chief complaint text does not contain the duration of the disease, it will score 0 points.

[0089] In the example of the present invention, the target subtext of the first text may correspond to at least one subtext requirement, and the target subtext of the second text may correspond to at least one subtext requirement. Therefore, the text generation model may determine the score value of each text subrequirement corresponding to the target subtext of the first text based on the scoring criteria corresponding to each subtext requirement, and use the sum of the score values ​​of each text subrequirement corresponding to the target subtext of the first text as the score value of the first subtext requirement. Similarly, the score value of each subtext requirement corresponding to the target subtext of the second text may be determined, and the sum of the score values ​​of each text subrequirement corresponding to the target subtext of the second text may be used as the score value of the second subtext requirement.

[0090] In the example of the present invention, step 202 includes:

[0091] Step 2021, for the first sub-text requirement and the second sub-text requirement corresponding to the same target sub-text type, compare the score value of the first sub-text requirement and the score value of the second sub-text requirement, and use the text requirement with the lower score between the first sub-text requirement and the second sub-text requirement as the third sub-text requirement.

[0092] For step 2021, please refer to the description in step 202, which will not be repeated here.

[0093] In the example of the present invention, step 203 includes:

[0094] Step 2031: determine that the third subtext is required to be at a target position of the first text or the target position of the medical recording, and modify the data at the target position.

[0095] When determining the first text or the medical recording corresponding to the third text requirement, the target position where the third text requirement appears in the first text or the medical recording can be determined, and the data at the target position in the first text or the medical recording can be modified.

[0096] For step 2031, please refer to the description in step 203, which will not be repeated here.

[0097] In an example of the present invention, in the process of obtaining the second text based on the medical recording, the timestamp of each sentence of the second text appearing in the medical recording can be identified at the same time. Therefore, when a third text requirement corresponds to a target sub-text in the second text, the timestamp of the medical recording corresponding to the target sub-text can be determined based on the target sub-text corresponding to the third text requirement. Thus, the target position of the third text requirement in the medical recording can be determined based on the timestamp.

[0098] In an example of the present invention, the scoring values ​​of the first sub-text requirement and the second sub-text requirement output by the text generation model are determined, and the text requirement with a lower scoring value is selected from the first sub-text requirement and the second sub-text requirement as the third sub-text requirement. The data corresponding to the third sub-text requirement in the first text or the medical recording is modified so that the modified first text or the medical recording corresponds, thereby improving the efficiency of manual data modification and saving labor costs.

[0099] In another example of the present invention, when providing prompts through a comparison table, the sub-text requirements corresponding to each target sub-text and the evaluation values ​​corresponding to each sub-text evaluation requirement can also be displayed in the comparison table, as shown in Table 2, so that the quality of the target sub-text can be clearly understood through the comparison table.

[0100] Table 2

[0101]

[0102] Figure 3 is a structural diagram of a medical record data evaluation device 40 provided in an embodiment of the present invention. The device 40 may include:

[0103] The acquisition module 401 is used to acquire the preset evaluation rules and the first text of the medical record to be evaluated and the consultation recording.

[0104] The recording recognition module 402 is used to recognize the medical consultation recording to obtain a second text corresponding to the medical consultation recording.

[0105] Processing module 403 is used to input the first text, the second text and the preset evaluation rules into a text generation model, obtain a first evaluation result of the first text, a second evaluation result of the second text, and compare the first evaluation result with the second evaluation result to obtain a comparison result, and the comparison result is used to determine whether the first text and the medical recording correspond.

[0106] Optionally, the preset evaluation rules include: text requirements. Processing module 403 includes:

[0107] The first processing module 4031 is used to generate a large model based on the text, extract the first text and the second text respectively, and obtain a target text corresponding to the first text and a target text corresponding to the second text.

[0108] The second processing module 4032 is used to evaluate the target text corresponding to the first text and the target text corresponding to the second text respectively by using the preset evaluation rules to obtain the first text requirement of the target text corresponding to the first text and the second text requirement of the target text corresponding to the first text.

[0109] Optionally, the processing module 403 includes:

[0110] The third processing module 4033 is used to determine that the first text is consistent with the medical consultation recording if the semantics of the target text corresponding to the first text is consistent with the semantics of the target text corresponding to the second text.

[0111] The fourth processing module 4033 is used to provide prompts for the target text corresponding to the first text, the target text corresponding to the second text, the first text requirement corresponding to the first target text, and the second text requirement corresponding to the second target text if the semantics of the target text corresponding to the first text are not completely consistent with the semantics of the target text corresponding to the second text.

[0112] Optionally, the preset evaluation rules further include: scoring rules corresponding to the text requirements. The device further includes:

[0113] The scoring value determination module is used to determine the scoring value of the first text requirement and the scoring value of the second text requirement that are not completely consistent with the semantics according to the preset evaluation rule.

[0114] The selection module is used to compare the score value of the first text requirement and the score value of the second text requirement, and use the text requirement with the lower score between the first text requirement and the second text requirement as the third text requirement.

[0115] A modification module is used to determine the first text or the medical recording corresponding to the third text requirement, and to modify the first text or the medical recording corresponding to the third text requirement.

[0116] Optionally, the target text includes: at least one target sub-text; the target sub-text type includes: any one of the chief complaint text type, the current medical history text type, the past medical history text type, and the allergy history text type; the second processing module 4032 includes:

[0117] The first processing submodule 40321 is used to evaluate the target subtext in the first text and the target subtext in the second text respectively by using the evaluation rules corresponding to each target subtext, so as to obtain the first subtext requirement and the second subtext requirement.

[0118] The fourth processing module 4033 includes:

[0119] The second processing submodule 40331 is used to provide prompts for the target subtext corresponding to the first text, the target subtext corresponding to the second text, the first subtext requirements, and the second subtext requirements for the target subtext of the same target subtext type if the semantics of the target subtext corresponding to the first text are inconsistent with the semantics of the target subtext corresponding to the second text.

[0120] Optionally, a scoring value determination module includes:

[0121] The scoring value determination submodule determines the scoring value of the first subtext requirement and the scoring value of the second subtext requirement for each target subtext type.

[0122] A selection submodule is used to compare the score value of the first sub-text requirement and the score value of the second sub-text requirement corresponding to the same target sub-text type, and use the text requirement with the lower score value between the first sub-text requirement and the second sub-text requirement as the third sub-text requirement.

[0123] The modification submodule is used to determine the target position of the first text or the target position of the medical recording, and to modify the data at the target position.

[0124] To summarize, the medical record data processing device provided by the embodiment of the present invention obtains a first text of the medical record, a second text corresponding to the medical consultation recording, and preset evaluation rules, and inputs the first text, the second text, and the preset evaluation rules into a text generation model. The text generation model evaluates the first text and the second text based on the preset evaluation rules to obtain a first evaluation result of the first text and a second evaluation result of the second text, and compares the first evaluation result with the second evaluation result to obtain a comparison result, thereby conveniently determining whether the first text of the medical record corresponds to the medical consultation recording based on the comparison result, so as to quickly understand whether there is any content inconsistency between the second text corresponding to the recording and the first text corresponding to the medical record written by the doctor.

[0125] The present invention also provides an electronic device, see Figure 4 , including: a processor 501, a memory 502, and a computer program 5021 stored in the memory and executable on the processor, wherein when the processor executes the program, the medical record data evaluation method of the aforementioned embodiment is implemented.

[0126] The present invention also provides a readable storage medium, when the instructions in the storage medium are executed by a processor of an electronic device, the electronic device can execute the medical record data evaluation method of the aforementioned embodiment.

[0127] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0128] The algorithm and display provided herein are not inherently related to any particular computer, virtual system or other device. Various general purpose systems can also be used together with the teachings based on this. According to the above description, it is obvious that the structure required for constructing such systems. In addition, the present invention is not directed to any specific programming language either. It should be understood that various programming languages ​​can be utilized to realize the content of the present invention described herein, and the description of the above specific languages ​​is for disclosing the best mode of the present invention.

[0129] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0130] Similarly, it should be understood that in order to streamline the present invention and aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of the present invention, the various features of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, this disclosed method should not be interpreted as reflecting the following intention: that the claimed invention requires more features than the features explicitly recited in each claim. More specifically, as reflected in the claims below, inventive aspects lie in less than all the features of the individual embodiments disclosed above. Therefore, the claims that follow the specific embodiment are hereby expressly incorporated into the specific embodiment, with each claim itself serving as a separate embodiment of the present invention.

[0131] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and in addition they may be divided into a plurality of submodules or subunits or subcomponents. Except that at least some of such features and / or processes or units are mutually exclusive, all features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device disclosed in this manner may be combined in any combination. Unless otherwise expressly stated, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0132] The various component embodiments of the present invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. It should be understood by those skilled in the art that a microprocessor or digital signal processor (DSP) may be used in practice to implement some or all of the functions of some or all of the components in the sorting device according to the present invention. The present invention may also be implemented as a device or apparatus program for executing part or all of the methods described herein. Such a program for implementing the present invention may be stored on a computer-readable medium, or may be in the form of one or more signals. Such a signal may be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0133] It should be noted that the above embodiments illustrate the present invention rather than limit it, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference symbol between brackets shall not be construed as a limitation on the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "one" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising a number of different elements and by means of a suitably programmed computer. In a unit claim enumerating a number of devices, several of these devices may be embodied by the same hardware item. The use of the words first, second, and third, etc., does not indicate any order. These words may be interpreted as names.

[0134] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0135] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0136] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A medical record data processing method, characterized in that: The method comprises: Obtaining preset evaluation rules and the first text of the medical records to be evaluated and the consultation recordings; Recognize the medical consultation recording to obtain a second text corresponding to the medical consultation recording; The first text, the second text and the preset evaluation rules are input into a text generation model to obtain a first evaluation result of the first text and a second evaluation result of the second text, and the first evaluation result is compared with the second evaluation result to obtain a comparison result, which is used to determine whether the first text and the medical recording correspond.

2. The method according to claim 1, characterized in that: The preset evaluation rules include: text requirements; The step of inputting the first text, the second text, and the preset evaluation rule into a text generation model to obtain a first evaluation result of the first text and a first evaluation result of the second text includes: Generate a large model based on the text, extract the first text and the second text respectively, and obtain a target text corresponding to the first text and a target text corresponding to the second text; The preset evaluation rules are used to evaluate the target text corresponding to the first text and the target text corresponding to the second text, respectively, to obtain the first text requirement of the target text corresponding to the first text and the second text requirement of the target text corresponding to the first text.

3. The method according to claim 2, characterized in that The comparing the first evaluation result with the second evaluation result to obtain a comparison result includes: If the semantics of the target text corresponding to the first text is consistent with the semantics of the target text corresponding to the second text, it is determined that the first text is consistent with the medical recording; If the semantics of the target text corresponding to the first text are not completely consistent with the semantics of the target text corresponding to the second text, prompts are given to the target text corresponding to the first text, the target text corresponding to the second text, the first text requirement corresponding to the first text, and the second text requirement corresponding to the second text.

4. The method according to claim 3, characterized in that The preset evaluation rules also include: scoring rules corresponding to the text requirements; The method further comprises: Determining, according to the preset evaluation rule, a score value of the first text requirement and a score value of the second text requirement that are not completely semantically consistent; Comparing the score of the first text requirement and the score of the second text requirement, and using the text requirement with the lower score between the first text requirement and the second text requirement as the third text requirement; Determine the first text or the medical recording corresponding to the third text requirement, and modify the first text or the medical recording corresponding to the third text requirement.

5. The method according to claim 3, characterized in that: The target text includes: at least one target sub-text; the target sub-text type includes: any one of the chief complaint text type, the current medical history text type, the past medical history text type, and the allergy history text type; The method of using the preset evaluation rule to evaluate the target text corresponding to the first text and the target text corresponding to the second text respectively to obtain a first text requirement of the target text corresponding to the first text and a second text requirement of the target text corresponding to the first text includes: The target subtext in the first text and the target subtext in the second text are evaluated respectively by using the subtext requirements corresponding to each target subtext type to obtain a first subtext requirement and a second subtext requirement.

6. The method according to claim 5, characterized in that If the semantics of the target text corresponding to the first text are not completely consistent with the semantics of the target text corresponding to the second text, prompting the target text corresponding to the first text, the target text corresponding to the second text, the first text requirement corresponding to the first target text, and the second text requirement corresponding to the second target text includes: For the target sub-texts of the same target sub-text type, if the semantics of the target sub-text corresponding to the first text are inconsistent with the semantics of the target sub-text corresponding to the second text, prompts are given for the target sub-text corresponding to the first text, the target sub-text corresponding to the second text, the first sub-text requirements, and the second sub-text requirements.

7. The method according to claim 5, characterized in that The step of determining the score value of the first text requirement and the score value of the second text requirement according to the preset evaluation rule includes: Determine, for each target subtext type, the first subtext requirement and the second subtext requirement corresponding to each target subtext type, a scoring value of the first subtext requirement and a scoring value of the second subtext requirement; The step of comparing the score of the first text requirement and the score of the second text requirement, and using the text requirement with the lower score between the first text requirement and the second text requirement as the third text requirement includes: For the first sub-text requirement and the second sub-text requirement corresponding to the same target sub-text type, compare the score value of the first sub-text requirement and the score value of the second sub-text requirement, and use the text requirement with the lower score value between the first sub-text requirement and the second sub-text requirement as the third sub-text requirement; Determining the first text or the medical consultation recording corresponding to the third text requirement, and modifying the first text or the medical consultation recording corresponding to the third text requirement, including: Determine that the third subtext is required to be at a target position of the first text or a target position of the medical recording, and modify the data at the target position.

8. A medical record data processing device, characterized in that: The device comprises: An acquisition module, used to acquire preset evaluation rules and the first text of the medical record to be evaluated and the consultation recording; A recording recognition module, used to recognize the medical consultation recording and obtain a second text corresponding to the medical consultation recording; A processing module is used to input the first text, the second text and the preset evaluation rules into a text generation model, obtain a first evaluation result of the first text, a second evaluation result of the second text, and compare the first evaluation result with the second evaluation result to obtain a comparison result, and the comparison result is used to determine whether the first text and the medical recording correspond.

9. An electronic device, characterized in that: include: A processor, a memory, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the medical record data processing method according to any one of claims 1 to 7 is implemented.

10. A readable storage medium, characterized in that: When the instructions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the medical record data processing method described in any one of claims 1-7.

Images