Methods for generating sample bills for joint cases, methods and equipment for bill recognition
By constructing a shared data object for cases and clinical timeline constraints, a sample of combined invoices that conforms to medical logic is generated, which solves the data inconsistency problem caused by the independent generation of invoice fields in the existing technology and realizes the automatic generation of consistent invoice samples.
Patent Information
- Application Number
- CN202611500998.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-24
- Publication Date
- 2026-11-10
AI Technical Summary
Existing methods for generating invoice samples cannot guarantee consistency across invoice data, especially in the context of medical invoices, where the generated invoice fields are independent and cannot meet semantic consistency requirements.
By constructing a case-shared data object, and combining diagnostic fields, a set of surgical exemptions for internal medicine diseases, and a surgical mapping table, a sample of combined case invoices that conforms to clinical timeline constraints is generated, ensuring data consistency across all invoice fields.
It enables the automatic generation of invoice samples that meet data consistency for synthesized cases, solves the semantic inconsistency problem caused by independent generation across invoice fields, and improves the logical rationality and consistency of generated invoices.
Smart Images

Figure CN123070083A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of bill consolidation processing technology, and in particular to a method for generating consolidated bill samples, a bill recognition method, and equipment. Background Technology
[0002] In scenarios such as intelligent insurance claims review, medical information system testing, and optical character recognition model training, a large number of medical invoices covering the complete diagnosis and treatment process are needed. Since real medical invoices contain sensitive personal information such as patient names, ID numbers, diagnoses, and surgeries, directly using real data for model training and system testing poses compliance risks. Therefore, the industry has begun exploring the creation of high-fidelity medical invoice samples through synthetic methods (i.e., generating entirely by computer programs without using real data).
[0003] Existing methods for generating invoice samples mainly fall into three categories. The first is based on statistical models. This method generates new data with similar statistical characteristics by fitting the statistical distribution of the original data. It can generate variable values sequentially based on conditional distributions, maintaining statistical correlations between variables. However, it primarily targets general statistical table data and does not address semantic consistency constraints across documents in medical invoices. The second method uses generative adversarial networks (GANs) to generate invoice data. It utilizes cyclic conditions to generate medical time-series data. However, by implicitly learning the data distribution through neural networks, it cannot explicitly guarantee precise field-level consistency across invoices (e.g., the same patient's name being identical across all invoices), and the controllability and interpretability of the generated results are weak. The third method is based on template-based random filling for invoice generation. Specifically, it pre-designs blank templates for various medical invoices (e.g., outpatient medical record templates, laboratory report templates, receipt templates, etc.) and uses random functions to independently generate values for each field of each template (e.g., randomly selecting patient names from a name database, randomly generating amounts within a reasonable range, etc.). This approach is simple to implement and can generate various invoice types, but the generation process for each field is independent, failing to guarantee semantic consistency between related invoices. Summary of the Invention
[0004] This invention provides a method for generating sample bills for a combined case, a bill recognition method, and a device to solve the problem that the bill fields generated for a sample bill for a combined case are independent of each other, and the consistency of cross-bill data cannot be guaranteed.
[0005] According to one aspect of the present invention, a method for generating a sample of a joint-stock document is provided, comprising: Construct the case-shared data object for the current case to be synthesized; The case-shared preprocessing data object is determined based on the diagnostic field in the case-shared data object and the set of surgical exemptions for internal medicine diseases; or, the case-shared preprocessing data object is determined based on the diagnostic field in the case-shared data object, the set of surgical exemptions for internal medicine diseases, and the surgical mapping table. Based on the case-shared preprocessing data objects, clinical timeline constraints, and medical invoice templates, a sample of combined invoices is generated.
[0006] According to another aspect of the present invention, a ticket recognition method is provided, comprising: Obtain a sample of the combined invoices generated using the combined invoices sample generation method; The sample of the combined bills is input into a preset machine learning model to train the combined bill recognition model, thus obtaining the combined bill recognition model. Input the relevant documents of the bills to be combined into the bill recognition model to obtain the bill recognition results.
[0007] According to another aspect of the present invention, a combined bill sample generation apparatus is provided, comprising: The shared data object construction module is used to construct the case shared data object for the current case to be synthesized; The preprocessed data object generation module is used to determine the case-shared preprocessed data object based on the diagnostic field in the case-shared data object and the set of surgical exemptions for internal medicine diseases; or, based on the diagnostic field in the case-shared data object, the set of surgical exemptions for internal medicine diseases, and the surgical mapping table. The combined case invoice sample generation module is used to generate combined case invoice samples based on the case shared preprocessing data objects, clinical timeline constraints, and medical invoice templates.
[0008] According to another aspect of the present invention, a ticket recognition device is provided, comprising: The module for obtaining sample of consolidated invoices is used to obtain sample of consolidated invoices generated by the method for generating sample of consolidated invoices. The model training module is used to input the sample of the combined bills into the preset machine learning model to train the combined bills recognition model and obtain the combined bills recognition model. The combined invoice recognition module is used to input the associated files of the invoices to be combined into the combined invoice recognition model to obtain the invoice recognition results.
[0009] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and a memory communicatively connected to said at least one processor; The memory stores a computer program that can be executed by the at least one processor. The computer program is executed by the at least one processor to enable the at least one processor to execute the method for generating a combined invoice sample according to any embodiment of the present invention, or to execute the invoice recognition method according to any embodiment of the present invention.
[0010] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions, the computer instructions being configured to cause a processor to execute, implement the method for generating a sample of a combined invoice as described in any embodiment of the present invention, or to execute the method for recognizing an invoice as described in any embodiment of the present invention.
[0011] According to another aspect of the present invention, a computer program product is provided, the computer program product comprising a computer program that, when executed by a processor, implements the method for generating a sample of a combined invoice as described in any embodiment of the present invention, or performs the method for recognizing an invoice as described in any embodiment of the present invention.
[0012] The technical solution of this invention constructs a case-shared data object for the current case to be synthesized, thereby obtaining a case-shared preprocessing data object based on the diagnostic fields and the set of surgical exemptions for internal medicine diseases in the case-shared data object; or, it determines the case-shared preprocessing data object based on the diagnostic fields, the set of surgical exemptions for internal medicine diseases, and the surgical mapping table in the case-shared data object, and then generates a combined case invoice sample based on the case-shared preprocessing data object, clinical timeline constraints, and medical invoice templates. In this solution, the invoice fields in the combined case invoice sample are not generated independently, but rather the case-shared preprocessing data object, which stores shared field values and conforms to clinical medical logic, is used as the data source referenced by different medical invoice templates. Furthermore, the rationality of the virtual creation time of the combined case invoice sample is maintained through clinical timeline constraints. This solves the problem that the generation of invoice fields in the invoice sample generated for a single synthesized case is independent and cannot guarantee cross-invoice data consistency, enabling the automatic generation of invoice samples that meet data consistency requirements for a single synthesized case.
[0013] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description
[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0015] Figure 1 This is a flowchart of a method for generating a sample of a combined invoice according to Embodiment 1 of the present invention; Figure 2 This is a flowchart of a method for generating a sample of a joint-purpose invoice according to Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of a mapping mechanism between a primary diagnostic name and surgical clinical association provided in Embodiment 2 of the present invention; Figure 4 This is a schematic diagram of the overall process of a method for generating a sample of a joint-purpose invoice provided in Embodiment 2 of the present invention; Figure 5 This is a schematic diagram of a sample quality assessment process provided in Embodiment 2 of the present invention; Figure 6 This is a flowchart of a ticket recognition method provided in Embodiment 3 of the present invention; Figure 7 This is a schematic diagram of a combined bill sample generation device provided in Embodiment 4 of the present invention; Figure 8 This is a schematic diagram of the structure of a ticket recognition device provided in Embodiment 5 of the present invention; Figure 9 A schematic diagram of an electronic device that can be used to implement embodiments of the present invention is shown. Detailed Implementation
[0016] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.
[0017] It should be noted that the terms "currently," "at present," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0018] Example 1 Figure 1 This is a flowchart of a method for generating a sample of consolidated invoices according to Embodiment 1 of the present invention. This embodiment is applicable to the automatic generation of invoice samples that meet data consistency requirements. This method can be executed by a device for generating a sample of consolidated invoices, which can be implemented in hardware and / or software and can be configured in an electronic device. This electronic device may include, but is not limited to, a computer or a server. Figure 1 As shown, the method includes: Step 110: Construct the case-shared data object for the current case to be synthesized.
[0019] The current case to be synthesized can be an event that generates a virtual medical invoice for a specific disease for the current virtual patient. The case-shared data object can be the case data objects that the current case to be synthesized needs to share. The case-shared data object can include, but is not limited to, patient identity field groups, hospital information field groups, diagnosis field groups, and surgical field groups. Optionally, the patient identity field group can include identity-related fields such as name, gender, ID number, date of birth, telephone number, and address. The hospital information field group includes hospital-related fields such as hospital name and department. The diagnosis field group includes diagnostic fields such as the primary diagnosis name (i.e., the diagnosis name of the chief complaint) and a list of secondary diagnoses, with the field names conforming to medical disease naming conventions. The surgical field group includes surgical-related fields such as surgery name and surgery date.
[0020] In this embodiment of the invention, a case-shared data object can be predefined for the current case to be synthesized, so as to centrally store the partial field values that need to be shared by all the bills in the current case to be synthesized.
[0021] For example, in the case-shared data object, except for the field values of the surgical field group that need further determination, the other field values are randomly generated by the computer according to the corresponding field attributes. The field values of the patient identity field group conform to the corresponding field attributes, such as gender can only be male or female, and the ID number must be 18 digits, etc.
[0022] Step 120: Determine the case-shared preprocessing data object based on the diagnosis field in the case-shared data object and the set of surgical exemptions for internal medicine diseases; or, determine the case-shared preprocessing data object based on the diagnosis field in the case-shared data object, the set of surgical exemptions for internal medicine diseases, and the surgical mapping table.
[0023] The medical surgery exemption set can be any diagnosis name that falls under the category of medical conditions and does not require surgery. For example, the medical surgery exemption set includes, but is not limited to, type 2 diabetes, grade 3 hypertension, pneumonia, cirrhosis, and chronic obstructive pulmonary disease. The case-sharing preprocessing data object can be used to store all field values that need to be shared across all tickets in the current case to be synthesized, excluding date-related fields. The surgical procedure mapping table can be a key-value pair mapping table from surgical diagnosis names to a list of surgical procedure names. For example, in the surgical procedure mapping table, acute appendicitis corresponds to laparoscopic appendectomy, fracture corresponds to internal fixation of fracture, coronary artery disease corresponds to coronary stenting / coronary artery bypass grafting, and gastric cancer corresponds to radical gastrectomy.
[0024] In this embodiment of the invention, the primary diagnosis name of a virtual patient can be determined based on the diagnosis field in the case-shared data object. If the primary diagnosis name falls into the medical surgery exemption set, the surgery field in the case-shared data object is deleted or set to null, resulting in a case-shared preprocessed data object. If the primary diagnosis name does not fall into the medical surgery exemption set, the field value of the surgery field is generated based on the chief complaint diagnosis name and the surgical mapping table. The relevant field values of the surgery field are then filled into the corresponding positions of the surgery field groups in the case-shared data object, resulting in a case-shared preprocessed data object.
[0025] Step 130: Generate a sample of combined invoices based on the case-shared preprocessing data object, clinical timeline constraints, and medical invoice template.
[0026] The clinical timeline constraint can be a pre-set rule for the sequence of medical events. This constraint ensures that the virtual creation date corresponding to the generated invoice is logically consistent with clinical practice, preventing inconsistencies such as tests being performed after discharge. The medical invoice template can be a medical invoice template with a reserved field for input. The combined invoice sample can be a sample of invoices from the current case to be combined. The combined invoice sample can contain various invoice types.
[0027] Optionally, clinical timeline constraints may include, but are not limited to: outpatient visit time less than or equal to admission time in outpatient medical records; admission time less than or equal to laboratory test submission time in inpatient medical records, and laboratory test submission time less than discharge time in inpatient medical records; admission date less than or equal to surgery date in inpatient medical records, and surgery date less than or equal to discharge date in inpatient medical records; surgery date less than or equal to pathology report date in inpatient medical records, and pathology report date less than or equal to discharge date in inpatient medical records; and payment settlement date on receipts equal to discharge date. Clinical timeline constraints can be understood as clinical timeline data objects, which may include date fields for 12 medical events in a single inpatient case, such as outpatient visit time, admission date, laboratory test submission date, imaging examination date, imaging report date, surgery date, pathology test submission date, pathology report date, prescription date, discharge date, and payment settlement date.
[0028] The medical invoice templates can include various types such as outpatient medical records, inpatient medical records, laboratory reports, imaging reports, surgical records, pathology reports, prescriptions, receipts, settlement lists, and expense details.
[0029] In this embodiment of the invention, a medical invoice template associated with a diagnosis field can be selected. That is, the invoice template corresponding to the actual medical invoice that may be generated is selected according to the disease corresponding to the diagnosis field. Field value filling is performed based on the case shared preprocessing data object. Each selected medical invoice template has an independent filling function, but all filling functions reference field values from the case shared preprocessing data object. In accordance with the clinical timeline constraints, the generation time is created for the medical invoice corresponding to the selected medical invoice template to obtain a combined invoice sample.
[0030] For example, for the selected medical bill template associated with the diagnosis field, the corresponding field values of each field group in the shared preprocessed data object of the same case are referenced according to their respective fill functions to fill the fields, so that the patient identity and the main diagnosis name are consistent in the filled combined bill samples.
[0031] Optionally, the medical invoice template also includes a cost field group, which includes cost fields such as total payment amount, medical insurance payment amount, and personal payment amount. For different primary diagnosis names, a corresponding total amount range is set for the invoice type corresponding to the medical invoice template, and a value is randomly selected from this range as the corresponding total payment amount for the invoice. It is important to note that the total payment amount = medical insurance payment amount + personal payment amount; that is, the corresponding field values for the medical insurance payment amount and personal payment amount are determined according to the constraint that the sum of the medical insurance payment amount and the personal payment amount is the total payment amount.
[0032] The technical solution of this invention constructs a case-shared data object for the current case to be synthesized, thereby obtaining a case-shared preprocessing data object based on the diagnostic fields and the set of surgical exemptions for internal medicine diseases in the case-shared data object; or, it determines the case-shared preprocessing data object based on the diagnostic fields, the set of surgical exemptions for internal medicine diseases, and the surgical mapping table in the case-shared data object, and then generates a combined case invoice sample based on the case-shared preprocessing data object, clinical timeline constraints, and medical invoice templates. In this solution, the invoice fields in the combined case invoice sample are not generated independently, but rather the case-shared preprocessing data object, which stores shared field values and conforms to clinical medical logic, is used as the data source referenced by different medical invoice templates. Furthermore, the rationality of the virtual creation time of the combined case invoice sample is maintained through clinical timeline constraints. This solves the problem that the generation of invoice fields in the invoice sample generated for a single synthesized case is independent and cannot guarantee cross-invoice data consistency, enabling the automatic generation of invoice samples that meet data consistency requirements for a single synthesized case.
[0033] Example 2 Figure 2 This is a flowchart of a method for generating a combined case invoice sample according to Embodiment 2 of the present invention. This embodiment is based on the above embodiment and provides specific optional implementation methods for generating a combined case invoice sample according to the case shared preprocessing data object, clinical timeline constraints, and medical invoice template. Figure 2 As shown, the method includes: Step 210: Construct the case-shared data object for the current case to be synthesized.
[0034] Step 220: Determine the case-shared preprocessing data object based on the diagnosis field in the case-shared data object and the set of surgical exemptions for internal medicine diseases; or, determine the case-shared preprocessing data object based on the diagnosis field in the case-shared data object, the set of surgical exemptions for internal medicine diseases, and the surgical mapping table.
[0035] In an optional embodiment of the present invention, determining the case-shared preprocessing data object based on the diagnosis field in the case-shared data object and the medical surgery exemption set may include: parsing the diagnosis field in the case-shared data object to obtain the current primary diagnosis name; if the current primary diagnosis name exists in the medical surgery exemption set, setting the surgery field in the case-shared data object to null to obtain the case-shared preprocessing data object.
[0036] The current primary diagnosis name can be the diagnosis name of the current virtual patient's chief complaint.
[0037] In this embodiment of the invention, the diagnosis field in the case shared data object can be parsed to obtain the current primary diagnosis name, and based on the current primary diagnosis name, the internal medicine surgery exemption set can be queried. If it can be successfully matched with the internal medicine surgery exemption set, it indicates that the disease corresponding to the primary diagnosis name of the virtual patient does not require surgery. Then, the surgery field in the case shared data object can be deleted or the surgery field can be set to null to obtain the case shared preprocessed data object.
[0038] In an optional embodiment of the present invention, determining the case-shared preprocessing data object based on the diagnostic field, the set of medical disease surgical exemptions, and the surgical mapping table in the case-shared data object may include: when the current primary diagnosis name does not exist in the set of medical disease surgical exemptions, determining the surgical name corresponding to the current primary diagnosis name based on the current primary diagnosis name and the surgical mapping table; and determining the case-shared preprocessing data object based on the surgical allocation probability assigned to the current primary diagnosis name and the surgical name corresponding to the current primary diagnosis name.
[0039] The probability of surgical allocation can be a pre-set probability of surgical execution.
[0040] In this embodiment of the invention, if the current primary diagnosis name is not found in the set of surgical exemptions for internal medicine diseases, it indicates that the disease corresponding to the current primary diagnosis name of the virtual patient requires surgery. Therefore, based on the current primary diagnosis name, the surgical mapping table is queried to determine the surgical name corresponding to the current primary diagnosis name. Based on the surgical allocation probability assigned to the current primary diagnosis name, when determining the surgical allocation for the current patient, the surgical name corresponding to the current primary diagnosis name is used as the field value of the surgical field group in the case shared data object to obtain the case shared preprocessing data object.
[0041] For example, assuming a 50% probability of surgery assignment, after determining the surgery name corresponding to the current primary diagnosis, one can choose to leave the surgery field empty, or choose to use the surgery name corresponding to the current primary diagnosis as the field value of the surgery field group in the case shared data object. The response probability for both options is 50%. The primary diagnosis name and surgical clinical association mapping mechanism is as follows: Figure 3 As shown.
[0042] Step 230: Based on the case-shared preprocessing data object and clinical timeline constraints, fill in the fields of the medical invoice template to obtain a sample of the combined invoices to be evaluated.
[0043] Among them, the sample of the combined bills to be evaluated may be a sample of the combined bills that has not been labeled.
[0044] In this embodiment of the invention, the original medical invoice template, after being filled with fields based on the case-shared preprocessing data object and clinical timeline constraints, can be used as a sample of combined invoices to be evaluated.
[0045] Step 240: Determine the sample of consolidated invoices based on the sample of consolidated invoices that have passed the sample quality assessment.
[0046] The sample quality assessment may include at least one of the following: patient identity consistency assessment, clinical rationality assessment, temporal consistency assessment, and financial consistency assessment. Patient identity consistency assessment determines whether the virtual patient's name, ID number, and other identity information across different invoices are completely consistent. Clinical rationality assessment determines the matching degree between the diagnosis and the department / test item. Temporal consistency assessment determines whether the dates on the invoice meet clinical timeline constraints. Financial consistency assessment determines whether the total amount equation is true.
[0047] In this embodiment of the invention, a sample quality assessment can be performed on the sample of the joint invoices to be evaluated, and the sample of the joint invoices to be evaluated that has passed the sample quality assessment can be labeled with a sample tag to obtain a sample of the joint invoices.
[0048] In an optional embodiment of the present invention, determining the sample of combined invoices based on the sample of combined invoices to be evaluated that has passed the sample quality assessment may include: converting the sample of combined invoices to be evaluated that has passed the sample quality assessment into a sample of combined invoices to be evaluated in image format; and performing field priority labeling and invoice association field truth value labeling on the sample of combined invoices to be evaluated in image format to obtain the sample of combined invoices.
[0049] The field priority label can be annotated with the priority of field names in the invoice. For example, the primary diagnosis name, name, medical insurance payment amount, and personal payment amount have the highest priority. The invoice-related field truth value label can be a pre-set sample label for the expected invoice type, field name, and field value.
[0050] In this embodiment of the invention, the sample of the joint invoice to be evaluated that has passed the sample quality assessment can be converted into an image format to simulate the form of a real invoice. This allows for the labeling of field priority tags and the labeling of truth values for invoice-related fields on the image-format sample of the joint invoice to be evaluated. The invoice sample with completed labeling is then used as the sample of the joint invoice.
[0051] For example, the sample of the bills to be evaluated can be rendered as a bill image based on HTML (HyperText Markup Language).
[0052] In a specific example, the overall process of generating a sample of a joint-case invoice can be found in [link to documentation]. Figure 4 For the sample quality assessment process, please refer to [link / reference]. Figure 5 . Figure 5 The "passed verification items" refer to the dimensions that passed the evaluation, namely, identity, clinical rationality, temporal consistency, and financial consistency. The score is a set value for the evaluation result. For example, the average of the consistency score and the rationality score can be used as the final score. The warning list is used to display the failed verification items. Improvement suggestions are improvements to the generated invoice samples and can be generated by a large model. For example, the average of the identity consistency score, temporal consistency score, and financial consistency score can be used as the consistency score. The calculation rules for the identity consistency score, temporal consistency score, financial consistency score, and rationality score can be preset.
[0053] The technical solution of this invention constructs a case-shared data object for the current case to be synthesized, and then determines the case-shared preprocessing data object based on the diagnostic fields and the set of surgical exemptions for internal medicine diseases in the case-shared data object; or, it determines the case-shared preprocessing data object based on the diagnostic fields, the set of surgical exemptions for internal medicine diseases, and the surgical mapping table in the case-shared data object. Then, based on the case-shared preprocessing data object and clinical timeline constraints, it fills in the fields of the medical invoice template to obtain a sample of the combined invoice to be evaluated. Further, based on the sample of the combined invoice to be evaluated that has passed the sample quality assessment, it determines the combined invoice sample. In this solution, the fields of each invoice in the combined invoice sample are not generated independently. Instead, the case-shared preprocessing data object, which stores shared field values and conforms to clinical medical logic, is used as the data source referenced by different medical invoice templates. Furthermore, the rationality of the virtual creation time of the combined invoice sample is maintained through clinical timeline constraints. This solves the problem that the generation of invoice fields in the invoice sample generated for a single synthesized case is independent and cannot guarantee cross-invoice data consistency. It can automatically generate invoice samples that meet data consistency for a single synthesized case.
[0054] Example 3 Embodiment 3 of the present invention provides a ticket recognition method, such as... Figure 6 As shown, Step 310: Obtain the sample of the combined invoice generated by the combined invoice sample generation method.
[0055] Step 320: Input the sample of the combined bills into the preset machine learning model to train the combined bill recognition model and obtain the combined bill recognition model.
[0056] The preset machine learning model can be any type of machine learning model, as long as it can be trained for combined medical invoice recognition. This embodiment of the invention does not limit the specific type of the preset machine learning model. The combined medical invoice recognition model can be used to identify the content of genuine medical invoices.
[0057] In this embodiment of the invention, a sample of combined invoices can be input into a preset machine learning model, so that the combined invoices recognition model can be trained based on the sample of combined invoices through the preset machine learning model to obtain a combined invoices recognition model.
[0058] Step 330: Input the associated file of the invoice to be identified into the combined invoice recognition model to obtain the invoice recognition result.
[0059] The associated file for the invoice to be identified can be a medical treatment-related file uploaded by the current user. The invoice recognition result can be the recognition result of the medical invoice in the associated file for the invoice to be identified. The invoice recognition result can include the invoice type, fields and field values in the invoice, and field priority, etc.
[0060] In this embodiment of the invention, after inputting the associated file of the invoice to be identified into the combined invoice identification model, the combined invoice identification model can be used to summarize the invoices of the same patient and the same primary diagnosis name in the associated file of the invoice to be identified, and obtain the invoice identification result.
[0061] In an optional embodiment of the present invention, after obtaining the invoice recognition result, the method may further include: determining the matching claim rule based on the invoice recognition result and the insurance claim rule base; and calculating the target claim amount based on the matching claim rule.
[0062] The insurance claims rule base can be a database storing insurance claims rules. The matching claims rule can be a claims rule adapted to the ticket recognition result. The target claim amount can be the claim amount calculated based on the matching claims rule.
[0063] In this embodiment of the invention, the invoice recognition result can be matched with the rules in the insurance claim rule base to obtain the claim rule that matches the invoice recognition result, i.e. the hit claim rule. Then, based on the hit claim rule, the corresponding target claim amount can be calculated for the invoice recognition result.
[0064] It should be noted that all information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for display, data used for analysis, etc.) involved in this disclosure are information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data comply with the relevant laws, regulations and standards of the relevant regions.
[0065] Example 4 Figure 7 This is a schematic diagram of a combined invoice sample generation device provided in Embodiment 4 of the present invention. Figure 7 As shown, the device includes: The shared data object construction module 410 is used to construct the case shared data object of the current case to be synthesized; The preprocessing data object generation module 420 is used to determine the case-shared preprocessing data object based on the diagnostic field in the case-shared data object and the set of surgical exemptions for internal medicine diseases; or, based on the diagnostic field in the case-shared data object, the set of surgical exemptions for internal medicine diseases, and the surgical mapping table. The combined case invoice sample generation module 430 is used to generate combined case invoice samples based on the case shared preprocessing data object, clinical timeline constraints, and medical invoice templates.
[0066] Optionally, the preprocessed data object generation module 420 includes a first preprocessed data object generation unit and a second preprocessed data object generation unit. The first preprocessed data object generation unit is used to parse the diagnosis field in the case shared data object to obtain the current primary diagnosis name; when the current primary diagnosis name exists in the internal medicine disease surgical exemption set, the surgical field in the case shared data object is set to empty to obtain the case shared preprocessed data object.
[0067] The technical solution of this invention constructs a case-shared data object for the current case to be synthesized, thereby obtaining a case-shared preprocessing data object based on the diagnostic fields and the set of surgical exemptions for internal medicine diseases in the case-shared data object; or, it determines the case-shared preprocessing data object based on the diagnostic fields, the set of surgical exemptions for internal medicine diseases, and the surgical mapping table in the case-shared data object, and then generates a combined case invoice sample based on the case-shared preprocessing data object, clinical timeline constraints, and medical invoice templates. In this solution, the invoice fields in the combined case invoice sample are not generated independently, but rather the case-shared preprocessing data object, which stores shared field values and conforms to clinical medical logic, is used as the data source referenced by different medical invoice templates. Furthermore, the rationality of the virtual creation time of the combined case invoice sample is maintained through clinical timeline constraints. This solves the problem that the generation of invoice fields in the invoice sample generated for a single synthesized case is independent and cannot guarantee cross-invoice data consistency, enabling the automatic generation of invoice samples that meet data consistency requirements for a single synthesized case.
[0068] Optionally, the second preprocessing data object generation unit is used to determine the surgical name corresponding to the current primary diagnosis name based on the current primary diagnosis name and the surgical mapping table when the current primary diagnosis name does not exist in the set of surgical exemptions for internal medicine diseases; and to determine the case-shared preprocessing data object based on the surgical allocation probability assigned to the current primary diagnosis name and the surgical name corresponding to the current primary diagnosis name.
[0069] Optionally, the case-sharing invoice sample generation module 430 is used to fill in the fields of the medical invoice template according to the case-shared preprocessing data object and the clinical timeline constraints to obtain a case-sharing invoice sample to be evaluated; and to determine the case-sharing invoice sample based on the case-sharing invoice sample to be evaluated that has passed the sample quality assessment; wherein, the sample quality assessment includes at least one of patient identity consistency assessment, clinical rationality assessment, temporal consistency assessment, and financial consistency assessment.
[0070] Optionally, the combined invoice sample generation module 430 is used to convert the combined invoice sample to be evaluated that has passed the sample quality assessment into an image format combined invoice sample to be evaluated; and to perform field priority labeling and invoice association field truth value labeling on the image format combined invoice sample to be evaluated to obtain the combined invoice sample.
[0071] The combined bill sample device provided in the embodiments of the present invention can execute the combined bill sample method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the method execution.
[0072] Example 5 Figure 8 This is a schematic diagram of the structure of a ticket recognition device provided in Embodiment 5 of the present invention. Figure 8 As shown, the device includes: The module 510 for obtaining a sample of a combined invoice is used to obtain a sample of a combined invoice generated by the method for generating a sample of a combined invoice. The model training module 520 is used to input the sample of the combined bills into the preset machine learning model to train the combined bills recognition model and obtain the combined bills recognition model. The invoice recognition module 530 is used to input the associated files of the invoices to be merged into the merging invoice recognition model to obtain the invoice recognition results.
[0073] Optionally, the invoice recognition device further includes a claim amount calculation module, used to determine the hit claim rule based on the invoice recognition result and the insurance claim rule library; and to calculate the target claim amount based on the hit claim rule.
[0074] Example 6 Figure 9A schematic diagram of an electronic device that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.
[0075] like Figure 9 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as ROM 12, RAM 13, etc., communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded into the RAM 13 from the storage unit 18. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An I / O interface 15 is also connected to the bus 14. The ROM 12 is a read-only memory, the RAM 13 is a random access memory, and the I / O interface 15 is an input / output interface.
[0076] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.
[0077] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as a method for generating sample bills or a bill recognition method.
[0078] In some embodiments, the method for generating a sample of combined invoices, or the method for recognizing invoices, may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded into and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the method for generating a sample of combined invoices, or the method for recognizing invoices, described above, may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the method for generating a sample of combined invoices, or the method for recognizing invoices, by any other suitable means (e.g., by means of firmware).
[0079] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.
[0080] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.
[0081] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, RAM, ROM, erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.
[0082] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).
[0083] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.
[0084] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS servers, such as high management difficulty and weak business scalability.
[0085] This application also discloses a computer program product, which includes a computer program that, when executed by a processor, implements the method for generating a sample of combined invoices or the method for recognizing invoices provided in any embodiment of this application. This program product shares the same inventive concept as the method for generating a sample of combined invoices or the method for recognizing invoices disclosed in the embodiments of this application, and therefore will not be described in detail here.
[0086] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.
[0087] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.
Claims
1. A method for generating a sample of combined invoices, characterized in that, include: Construct the case-shared data object for the current case to be synthesized; The case-shared preprocessing data object is determined based on the diagnostic field in the case-shared data object and the set of surgical exemptions for internal medicine diseases; or, the case-shared preprocessing data object is determined based on the diagnostic field in the case-shared data object, the set of surgical exemptions for internal medicine diseases, and the surgical mapping table. Based on the shared preprocessed data objects, clinical timeline constraints, and medical invoice templates of the cases, a sample of combined invoices is generated.
2. The method according to claim 1, characterized in that, Based on the diagnostic field in the case-shared data object and the set of surgical exemptions for internal medicine diseases, the case-shared preprocessing data object is determined, including: Parse the diagnosis field in the case shared data object to obtain the current primary diagnosis name; When the current primary diagnosis name exists in the set of surgical exemptions for internal medicine diseases, the surgical field in the case-shared data object is set to empty, and the case-shared preprocessed data object is obtained.
3. The method according to claim 2, characterized in that, Based on the diagnostic field, the set of surgical exemptions for internal medicine diseases, and the surgical mapping table in the case-shared data object, the case-shared preprocessing data object is determined, including: If the current primary diagnosis name is not found in the set of surgical exemptions for internal medicine diseases, the surgical name corresponding to the current primary diagnosis name is determined according to the current primary diagnosis name and the surgical mapping table. Based on the surgical assignment probability assigned to the current primary diagnosis name and the surgical name corresponding to the current primary diagnosis name, the case shared preprocessing data object is determined.
4. The method according to claim 1, characterized in that, Based on the shared preprocessed data objects, clinical timeline constraints, and medical invoice templates, a sample of combined invoices is generated, including: Based on the case-shared preprocessing data object and the clinical timeline constraints, the fields of the medical invoice template are filled to obtain a sample of combined invoices to be evaluated. The sample of consolidated invoices to be evaluated is determined based on the sample of consolidated invoices that has passed the sample quality assessment; The sample quality assessment includes at least one of patient identity consistency assessment, clinical rationality assessment, temporal consistency assessment, and financial consistency assessment.
5. The method according to claim 4, characterized in that, Based on the sample of consolidated invoices to be evaluated that has passed the sample quality assessment, the sample of consolidated invoices is determined, including: The sample of the consolidated invoices to be evaluated, which has passed the sample quality assessment, will be converted into an image format sample of the consolidated invoices to be evaluated. For the sample of the consolidated invoices to be evaluated in image format, field priority labels and truth value labels for invoice-related fields are added to obtain the sample of the consolidated invoices.
6. A method for recognizing invoices, characterized in that, include: Obtain a sample of a combined bill of exchange generated by the method for generating a combined bill of exchange sample according to any one of claims 1-5; The sample of the combined bills is input into a preset machine learning model for training the combined bill recognition model to obtain the combined bill recognition model. The associated file of the invoice to be identified is input into the combined invoice recognition model to obtain the invoice recognition result.
7. The method according to claim 6, characterized in that, After obtaining the ticket recognition result, the following is also included: Based on the ticket recognition results and the insurance claims rule base, the matching claims rule is determined; Calculate the target claim amount based on the stated hit claim rules.
8. An electronic device, characterized in that, The electronic device includes: At least one processor, and a memory communicatively connected to said at least one processor; The memory stores a computer program that can be executed by the at least one processor, which is executed by the at least one processor to enable the at least one processor to perform the method for generating a combined invoice sample according to any one of claims 1-5, or to perform the invoice recognition method according to any one of claims 6-7.
9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the method for generating a sample of a combined invoice as described in any one of claims 1-5, or to execute the invoice recognition method as described in any one of claims 6-7.
10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the method for generating a sample of a combined invoice according to any one of claims 1-5, or performs the invoice recognition method according to any one of claims 6-7.
