WO2020034874A1 - 医疗单据审核方法、装置、计算机设备和存储介质 - Google Patents

医疗单据审核方法、装置、计算机设备和存储介质 Download PDF

Info

Publication number
WO2020034874A1
WO2020034874A1 PCT/CN2019/099425 CN2019099425W WO2020034874A1 WO 2020034874 A1 WO2020034874 A1 WO 2020034874A1 CN 2019099425 W CN2019099425 W CN 2019099425W WO 2020034874 A1 WO2020034874 A1 WO 2020034874A1
Authority
WO
WIPO (PCT)
Prior art keywords
medical
current
items
item
medical item
Prior art date
Application number
PCT/CN2019/099425
Other languages
English (en)
French (fr)
Inventor
程吉安
管音
Original Assignee
平安医疗健康管理股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 平安医疗健康管理股份有限公司 filed Critical 平安医疗健康管理股份有限公司
Publication of WO2020034874A1 publication Critical patent/WO2020034874A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
    • G16H10/20—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for electronic clinical trials or questionnaires
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Definitions

  • the present application relates to a method, a device, a computer device, and a storage medium for examining medical documents.
  • the inventors realized that the current medical settlement management and control system relies on threshold control of a certain indicator or risk classification of settlement documents based on a data model. In this way, all medical items are reviewed at the same point in time, which may cause errors in the review.
  • a method for reviewing medical documents includes:
  • the sorted first medical items are reviewed according to the order among the standard medical items in the fixed knowledge.
  • the device includes:
  • An extraction module configured to obtain a medical document to be audited from the server, and extract a diagnosis result set and a current medical item from the medical document to be audited;
  • a clinical path template acquisition module is configured to obtain a current diagnosis result from the diagnosis result set, extract a pre-stored clinical path template set from a server, and compare the current diagnosis result with a standard diagnosis result in the clinical path template set. Performing a comparison to obtain a current clinical path template, where each standard medical item is stored in chronological order;
  • a first sorting module configured to sort the current medical items according to a preset time interval according to the sequence of each standard medical item in the current clinical path template to obtain a first medical item
  • a second sorting module configured to obtain pre-stored fixed knowledge that characterizes a sequence between the standard medical items, and perform a comparison on each preset time interval according to the sequence between the standard medical items in the fixed knowledge; Ranking the first medical item; and
  • An auditing module is configured to audit the sorted first medical items according to an order among standard medical items in the fixed knowledge.
  • a computer device includes a memory and one or more processors.
  • the memory stores computer-readable instructions.
  • the steps of the medical document review method provided in any embodiment of the present application are implemented.
  • One or more non-volatile storage media storing computer-readable instructions, and when the computer-readable instructions are executed by one or more processors, the one or more processors implement the medical treatment provided in any one of the embodiments of the present application. Steps of the document review method.
  • FIG. 1 is a diagram of an application scenario of a method for reviewing medical documents according to one or more embodiments.
  • FIG. 2 is a schematic flowchart of a medical document review method according to one or more embodiments.
  • FIG. 3 is a schematic flowchart of a medical document review method according to another or more embodiments.
  • FIG. 4 is a schematic diagram of an optimal sort sequence insertion process according to one or more embodiments.
  • FIG. 5 is a block diagram of a medical document review device according to one or more embodiments.
  • FIG. 6 is a block diagram of a computer device according to one or more embodiments.
  • the medical document review method provided in this application can be applied to the application environment shown in FIG. 1.
  • the audit terminal 102 communicates with the server 104 through a network.
  • the audit terminal 102 obtains the medical document to be audited from the server 104, then extracts the diagnosis result and the current medical item from the medical document to be audited, sequentially obtains the diagnosis result as the current diagnosis result, and then obtains the current clinical path template corresponding to the current diagnosis result.
  • the current medical items are sorted according to a preset time interval to obtain the first medical item.
  • the first medical item may be sorted by day to obtain the medical items in each day, and then according to the standard medical items in fixed knowledge.
  • the first medical items in each preset time interval, such as each day, are sorted in order, so that the sorted first medical items are audited as the audit result of the pending medical documents, instead of All medical items are audited at the same time, which improves the accuracy of the audit.
  • the audit terminal 102 may be, but is not limited to, various personal computers, notebook computers, smart phones, tablet computers, and portable wearable devices.
  • the server 104 may be implemented by an independent server or a server cluster composed of multiple servers.
  • a method for reviewing medical documents is provided.
  • the method is applied to the review terminal in FIG. 1 as an example, and includes the following steps:
  • S202 Obtain a medical document to be audited from the server, and extract the diagnosis result and the current medical item from the medical document to be audited.
  • the medical documents to be audited are generated by the patient to be audited during a consultation, which may include the diagnosis results and medical items, and the medical items may include the name of the charge item, the number of items, and the amount.
  • the server of the hospital collects the pending medical documents generated in each hospital terminal deployed in the hospital, so that the server stores the pending medical documents corresponding to each patient, so that the review terminal can
  • the server obtains the pending medical documents corresponding to the patients to be reviewed, and extracts the diagnosis result set and the current medical items from the document information.
  • the diagnosis result set and the current medical items are generated by the pending patients during a consultation.
  • S204 Obtain the current diagnosis result from the diagnosis result set, and extract the pre-stored clinical path template set from the server, and compare the current diagnosis result with the standard diagnosis result in the clinical path template set to obtain the current clinical path.
  • Template stores each standard medical item in chronological order.
  • Clinical Pathway refers to the establishment of a set of standardized treatment models and procedures for a certain disease. It is a comprehensive model of clinical treatment. It is guided by evidence-based medical evidence and guidelines to promote treatment organization and disease management. The method ultimately plays a role in regulating medical behavior, reducing variation, reducing costs, and improving quality. Each medical item is stored in the clinical path template corresponding to each disease according to time.
  • the diagnosis result in the diagnosis result set refers to the disease suffered by the patient, for example, it can be the name of the disease or the disease code, and the clinical path template corresponding to the diagnosis result can be matched according to the diagnosis result. And optionally, if there are multiple diagnosis results, one of the diagnosis results may be selected as the current diagnosis result first, and then the corresponding current clinical path template is matched by the current diagnosis result.
  • S206 According to the order of each standard medical item in the current clinical path template, sort the current medical items according to a preset time interval to obtain a first medical item.
  • each medical item can be sorted according to the time divided in the clinical path template, for example, sorted by day to obtain the first medical item in each day, such as the current There are 10 medical items. After the current clinical path is sorted, 3 of the 10 current medical items are sorted into the first day. Then these 3 current medical items are the first medical items on the first day. If 4 are sorted into the second day, these 4 current medical items will be regarded as the first medical item on the second day, and the remaining 3 items cannot be sorted according to the current clinical path.
  • NLP Neuro-Linguistic Programming
  • the division time in is divided by day to get the first medical item in each day.
  • S208 Obtain pre-stored fixed knowledge that characterizes the sequence between the standard medical items, and sort the first medical item within each preset time interval according to the order between the standard medical items in the fixed knowledge. Fixed order between standard medical items in knowledge.
  • fixed knowledge refers to the knowledge of the sequence between medical items. For example, it can indicate that fasting examinations are required for blood glucose, liver function, and the like, and that the medications are evenly distributed according to time periods.
  • the review terminal may first obtain the first medical item in each day, and then sort the first medical item according to the order between the standard medical items in the fixed knowledge. Specifically, the ranking method is also to sort the first medical item. The name of the item is matched with the name of the medical item in the fixed knowledge. After the matching is successful, the order of the medical items corresponding to the corresponding fixed knowledge is obtained, so that the first medical item is sorted according to the order.
  • the diagnosis results and current medical items are first extracted from the medical documents, and then the clinical path template is obtained based on the diagnosis results. Therefore, the current medical items are sorted by day, and in order to further improve the accuracy of sorting, they are also sorted by fixed knowledge, so that the medical items corresponding to the medical documents are sorted by time point, and then each sorted medical item is sorted. Auditing, so that medical items are audited in order according to the order of time nodes, which can improve the accuracy of the audit.
  • the above medical document review method does not directly review the document information after obtaining the document information, but extracts the diagnosis results and current medical items from the document information, and obtains the current clinical path template corresponding to the diagnosis results.
  • the path template sorts the current medical items by day, and sorts the first medical item sorted by day according to the order between standard medical items in fixed knowledge, so that the first medical item sorted can be reviewed. It is not that all medical items are audited at the same time, which improves the accuracy of the audit.
  • the method may further include: determining whether There are pending medical items that cannot be sorted according to the order between the standard medical items in the fixed knowledge; when the medical items are to be processed, the medical items to be processed are based on the probability model and the fixed knowledge
  • the standard medical items in the sequence are sorted and the first medical item is sorted to get the best sorted sequence.
  • the probability model uses a single data training performed by the nursing doctor with different timestamps carried by different diseases to generate and characterize each standard medical item. The probability value of the sort order. Thereby, the first medical item after being sorted is reviewed, including: the best sorting sequence is checked.
  • the probabilistic model is generated by performing single-data training using nursing doctors' orders with different timestamps for different diseases, which gives each diagnosis result, that is, the sort order of each medical item corresponding to the disease, and inputs different medical items.
  • the sort order of the different medical items and the probability corresponding to the different sort order can be obtained.
  • the medical items to be processed may include two types. One is that when the first medical item is sorted according to the order between standard medical items in the fixed knowledge, the existing ones are successfully matched with the fixed knowledge, but cannot be based on the fixed knowledge.
  • the second medical item in the order of standard medical items, and the other is a third medical item that failed to match the fixed knowledge.
  • the second and third are only for the purpose of distinguishing. There is no specific limitation here.
  • the audit terminal matches the first medical item with the fixed knowledge, the second medical item and the medical item in the fixed knowledge are successfully matched.
  • the order between standard medical items in fixed knowledge cannot determine the order with other first medical items, that is, the second medical item may have multiple different ranking positions, or it may not be successfully matched with fixed knowledge Third medical project.
  • the review terminal needs to determine whether there are different ranking positions after sorting according to the order between the standard medical items in the fixed knowledge.
  • the review terminal may first match the name of the first medical item with the name of the medical item in the fixed knowledge to obtain a first item that can be sorted according to the order between the standard medical items in the fixed knowledge.
  • the second medical item and the third medical item that cannot be sorted according to the order between the standard medical items in the fixed knowledge, and then the review terminal can be based on the standard medical items in the fixed knowledge.
  • the second medical items that are sorted in the order are sorted, and according to the sorting result, the second medical items that have different sorted positions are sorted according to the order between the standard medical items in the fixed knowledge.
  • the review terminal After the review terminal obtains the unsorted medical items to be processed, it inputs the unprocessed medical items and the first sorted medical item into a probability model to insert the unprocessed medical items into the sorted In a medical project, an optimal ranking sequence including the first medical project of the day is finally formed.
  • the review terminal may calculate the probability of inserting the medical item to be processed at a different position in the first medical item that has been sorted, and select a sort sequence with the highest probability as the best sort sequence. And therefore, the review of the first ranked medical item is the review of the best ranking sequence.
  • the application scenario is extended to include a third medical item without a fixed knowledge part and a second medical item that cannot be sorted according to the sequence part between the standard medical items in the fixed knowledge, that is, the above-mentioned medical item to be processed Generate the best sequence through the probability model and complete the remodeling of the path.
  • the above-mentioned method for reviewing medical documents may further include: when the first medical item is sorted according to the order between the medical items to be processed and the standard medical items in the fixed knowledge according to the probability model, multiple When there are two best sorted sequences, the difference values of the probabilities corresponding to the multiple best sorted sequences are calculated. When the difference value is less than a preset value, the different sorted sequences in the multiple best sorted sequences are obtained. Medical items, and medical items that are not sorted according to the current clinical path model, as the current medical item; after obtaining the next diagnosis result as the current diagnosis result, continue to obtain the current clinical path template corresponding to the current diagnosis result, until the document information Extracting the diagnostic results to the traversal is complete.
  • the above-mentioned method for reviewing medical documents may further include: after the traversal of the diagnosis results in the diagnosis result set is complete, and there are still unsorted current medical items, the unsorted current medical items are received; Modified order of current medical items that are ordered; when the modified order indicates that the unsorted current medical items are unreasonable, the unordered current medical items are marked as unreasonable medical items; when the modified order indicates that the unsorted current medical items are reasonable , According to the modification instruction, the unsorted current medical item is inserted into the generated best sorted sequence, and the probability model is updated according to the best sorted sequence after the unsorted current medical item is inserted.
  • FIG. 3 is a flowchart of a medical document review method in another embodiment.
  • the fifth medical item is ranked according to a probability model. Medical items with different sort positions in the best sort sequence with a value less than a preset value.
  • the sixth medical item is a medical item that fails to match the current clinical path when the clinical path is matched. In one embodiment, only the fifth medical item or the sixth medical item may be included, or both the fifth medical item and the sixth medical item may be included.
  • the seventh medical project refers to the unsorted current medical project mentioned above, that is, after all the diagnosis results are traversed, there is still a fifth medical project and / or a sixth medical project, and the existing seventh medical project requires manual intervention. For guidance.
  • the probability is re-ranked to complete the path remodeling.
  • multiple best-ranked sequences are obtained through the ranking of the probability model, that is, multiple best-ranked sequences are output by the probability model, multiple The difference value of the probability corresponding to the best sorting sequence, that is, the probability corresponding to the generated best sorting sequence is first obtained. The probability is generated and output by the probability model, and then the difference of the probability corresponding to multiple best sorting sequences is calculated.
  • the review terminal judges whether there is a sixth medical item that is not sorted by day according to the current clinical path model, that is, a sixth medical item that has not been successfully matched with the current clinical path. If it exists, the fifth medical item and the sixth medical item are regarded as current Medical items.
  • the sixth medical item may be taken as the current medical item and obtained
  • the next diagnosis result of the current diagnosis result extracted from the medical document is taken as the current diagnosis result, and then the current clinical path template corresponding to the current diagnosis result is continued to be obtained until the diagnosis result traversal is completed from the document information to complete the document information All the diagnosis results and medical items involved in the traversal are completed, and the medical items are sorted and completed according to the diagnosis results, so that all medical items can be reshaped according to time to facilitate subsequent review.
  • the remaining unsorted The medical items are output, that is, the unsorted current medical items are output in order to introduce manual intervention, that is, the user first judges whether the output unsorted current medical item, that is, the seventh medical item, is reasonable.
  • the judgment is based on human experience. That is, whether the disease requires the seventh medical item, etc. If it is not reasonable, mark the seventh medical item as unreasonable, so that the off-clinical path list template corresponding to the current clinical path can be updated, so that the next time you get the When the seventh medical item corresponding to the clinical path is directly marked as an unreasonable medical item, manual intervention can be avoided.
  • the seventh medical item is inserted into the generated optimal ranking sequence, and the probability is updated according to the optimal ranking sequence after entering and leaving the seventh medical item.
  • the model can be directly sorted by the probability model next time, without the need to introduce manual intervention to improve efficiency.
  • the out-of-clinical item list template is updated.
  • the probability model is modified according to the manual modification instruction. The reasonable items will be manually inserted into the sequence, and the probability model will be updated, and the unreasonable items will be used as a template for the list of items outside the clinical path for unreasonable management and control to improve efficiency.
  • sorting the first medical items in each preset time interval according to the order between the standard medical items in the fixed knowledge may include: The items are matched with the pre-sorted standard medical items in the fixed knowledge; if the matching is successful, the first medical item is sorted according to the successfully matched standard medical items.
  • the first medical item is first matched with the medical items in the fixed knowledge. If the matching is successful, it means that the successfully matched first medical item exists. Fix the knowledge, and then sort the first medical item according to the fixed knowledge corresponding to the successfully matched medical item.
  • the following examples are used to illustrate:
  • the following items are divided into the first day of hospitalization according to the clinical path template.
  • blood routine urine routine
  • coagulation function infection.
  • the review terminal sorts according to fixed knowledge. If the blood gas analysis needs to be completed before lung function, its serial numbers are marked as 1_a and 2_a (_a is a random marker added automatically); the same way Common sense is completed before the 3D CT examination of the clavicle, and it is marked as 1_b, 2_b (similarly _b is a random marker added automatically); according to the blood collection process, blood routine, liver function, kidney function, electrolyte, blood glucose, coagulation function, Infectious diseases are screened for sampling at the same time, so they are labeled as 1_c (similarly _c is a random marker added automatically); according to the requirements of test knowledge, blood glucose, liver function, etc.
  • the clear order situation is 1_c, 1_a, 2_a, 1_b, 2_b, or 1_c, 1_b, 2_b, 1_a, 2_a, that is, there are second medical items 1_a, 2_a, 1_b, 2_b, and unsorted urine routines that are sorted according to the order between standard medical items in fixed knowledge Chest radiograph, bilateral lower extremity vascular ultrasound, echocardiography, that is, there is a third medical item that cannot be sorted according to the order between standard medical items in fixed knowledge.
  • This third medical item can be regarded as being arbitrarily inserted into the sequence.
  • This completes the ordering of fixed knowledge in addition to the order of fixed knowledge, there are also ways of using drugs. If the drug is used 3 times a day, the drug will be 6: 00-8: 00, 12: 00-14: 00, and 18:00. -The time period of 20:00 is evenly distributed, and twice a day will be evenly divided according to 8:00, 16:00, etc., and the approximate order is determined, and then the CRF will discharge the sequence with the greatest probability based on probability).
  • the urine routine is usually completed after the blood drawing project, and the chest radiograph and electrocardiogram are usually immediately followed by, followed by 1_b, 2_b, etc.
  • “Routine, chest radiograph, electrocardiogram, 1_a, 2_a, 1_b, 2_b, double lower limb vascular ultrasound, echocardiography” are the sequences with the highest probability, which is the best sorted sequence above.
  • sequence 1 is 0.867: "1_c, urine routine, chest radiograph, electrocardiogram, 1_a, 2_a, 1_b, 2_b, double lower limb vascular ultrasound, echocardiography "; sequence 2 probability is 0.785:" 1_c, urine routine, chest radiograph, electrocardiogram, 1_a, 2_a, 1_b, 2_b, echocardiography, double lower limb vascular ultrasound ", ie Echocardiography and vascular ultrasound of both lower extremities are arranged in sequence, and the difference is less than the preset value (the preset value can be controlled manually, 10% is taken here.
  • the first medical items are first sorted according to the order between the standard medical items in the fixed knowledge. It is only necessary to match the first medical items with the fixed knowledge, which is simple and efficient.
  • the medical items to be processed and the first medical item sorted according to the order between the standard medical items in the fixed knowledge are sorted to obtain the optimal sorting sequence, which includes: The first medical item after the ordering among the standard medical items in the sequence generates a first sequence; selecting the current pending medical item from the pending medical items, and inserting the current pending medical item into different positions of the first sequence to form The second sequence, and calculate the probability of different second sequences; obtain the position corresponding to the second sequence with the highest probability as the insertion position of the current pending medical item, and select the next pending medical item from the pending medical items as the current The medical items to be processed are continued to be inserted into different positions of the first sequence to form a second sequence, until the medical items to be processed are inserted into the first sequence to obtain the best sorted sequence.
  • difference value (p1-p2) / ((p1 + p2) / 2).
  • p1 and p2 are corresponding probabilities at different positions where the current to-be-treated medical item is inserted into the first sequence.
  • the first medical item to be sorted according to the order between the medical items to be processed according to the probability model and the standard medical items in the fixed knowledge is: using the probability model to automatically generate judgment features according to the diagnosis and treatment timeline to Predetermined order gives the best ranking sequence based on the probability model.
  • the diagnosis and treatment time axis refers to the actual sequence of diagnosis and treatment items and the sequence of items after manual processing, which are formed by a time-stamped care execution order.
  • the medical items to be processed and the first medical item sorted according to the order among the standard medical items in the fixed knowledge mainly include the following steps:
  • the time sequence probability table between the items is generated from the time-stamped care execution order data.
  • the items completed at the same time are marked as 9. See table 2 below, where the column indicates the first. For example, the probability of blood routine before the chest radiograph is 0.81. .
  • the uncertain part of the original sequence is inserted as the insertion value, and the probability is calculated.
  • the insertion point with the highest probability is selected for insertion, and the sequence is completed in a circular manner.
  • the following table 3 and table 4 urine routine, chest Films, electrocardiograms, vascular ultrasound of both lower extremities, and echocardiography are to be inserted as insertion items.
  • the probability is calculated by querying the probability table, that is, the above table 2.
  • the probability of a after c is multiplied by the probability of a before b to obtain the probability of cab arrangement, and then the remaining items to be inserted are inserted by the same logic, that is, from the medical treatment to be processed.
  • the current medical item to be processed is selected, and the current medical item to be processed is inserted into different positions of the first sequence to form a second sequence, and the probability of the different second sequence is calculated; the position corresponding to the second sequence with the highest probability is obtained
  • FIG. 4 is a schematic diagram of an optimal sorting sequence insertion process in an embodiment.
  • difference value (p1-p2) / ((p1 + p2) / 2).
  • p1 and p2 are corresponding probabilities at different positions where the current to-be-treated medical item is inserted into the first sequence.
  • the probability difference that the same currently pending medical item is inserted into two different positions is less than 10% of the preset value, it is considered that a contradiction has occurred, and the preset value can be controlled manually.
  • a first sequence is generated according to a first medical item sorted according to an order among standard medical items in fixed knowledge; a current pending medical item is selected from the pending medical items, and the current pending medical item is selected
  • the second sequence is inserted at different positions of the first sequence, and the probability of the different second sequence is calculated; the position corresponding to the second sequence with the highest probability is obtained as the insertion position of the current medical item to be processed, and the cycle is sequentially obtained to obtain the best Sorting sequence, simple and reliable.
  • steps in the flowchart of FIG. 2-3 are sequentially displayed in accordance with the directions of the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless explicitly stated in this document, the execution of these steps is not strictly limited, and these steps can be performed in other orders. Moreover, at least a part of the steps in Figure 2-3 may include multiple sub-steps or stages. These sub-steps or stages are not necessarily performed at the same time, but may be performed at different times. These sub-steps or stages The execution order of is not necessarily performed sequentially, but may be performed in turn or alternately with at least a part of another step or a sub-step or stage of another step.
  • a medical document review device which includes: an extraction module 100, a clinical path template acquisition module 200, a first ranking module 300, a second ranking module 400, and an audit module 500. among them:
  • the extraction module 100 is configured to obtain a medical document to be audited from a server, obtain document information corresponding to a patient to be audited from the server, and extract a diagnosis result set and a current medical item from the medical document to be audited.
  • the clinical path template acquisition module 200 is configured to extract a pre-stored clinical path template set from the server, and compare the current diagnosis result with the standard diagnosis result in the clinical path template set to obtain the current clinical path template.
  • the chronological order stores various standard medical items.
  • the first sorting module 300 is configured to sort the current medical items according to a current clinical path template according to a preset time interval to obtain a first medical item.
  • the second sorting module 400 is configured to obtain pre-stored fixed knowledge that characterizes a sequence between standard medical items, and perform a first order in each preset time interval according to the order between the standard medical items in the fixed knowledge. Sort medical items.
  • the auditing module 500 is configured to audit the sorted first medical items.
  • the above-mentioned medical document review device may further include:
  • a first determination module is configured to determine whether there are medical items to be processed that cannot be sorted according to an order among standard medical items in fixed knowledge.
  • the third sorting module is used to sort the first medical item after the medical item to be processed is sorted according to the order between the standard medical items in the fixed knowledge when there is a medical item to be processed to obtain the best Sorting sequence, the probability model uses the single data training performed by the nursing doctor's order with different timestamps carried by different diseases to generate and characterize the probability value of the sorting order of each standard medical item.
  • the review module 500 is also used to review the best sorted sequence.
  • the above-mentioned medical document review device may further include:
  • a calculation module configured to obtain the first medical item after being sorted according to the probability model, the medical item to be processed, and the first medical item sorted according to the order between standard medical items in the fixed knowledge. The difference in probabilities corresponding to multiple best-ranked sequences.
  • a second judgment module configured to obtain medical items of different rankings in the multiple optimal ranking sequences obtained when the difference value is less than a preset value, and medical items that are not ranked according to the current clinical path model As the current medical project.
  • the traversal module is used to obtain the next diagnosis result as the current diagnosis result, and then continue to obtain the current clinical path template corresponding to the current diagnosis result, until the traversal of the diagnosis result is extracted from the document information.
  • the above-mentioned medical document review device may further include:
  • An output module is configured to output the unsorted current medical items when the unsorted current medical items still exist after the diagnosis result traversal in the diagnostic result set is completed.
  • the receiving module is configured to receive an input modification instruction for an unsorted current medical item.
  • a marking module configured to mark the unsorted current medical items as unreasonable medical items when the modification instruction indicates that the unsorted current medical items are unreasonable.
  • An update module is used to insert the unsorted current medical items into the generated best sorted sequence according to the modified instructions when the unordered current medical items are reasonable, and after inserting the unsorted current medical items according to the modified instructions Update the probabilistic model for the best sorted sequence.
  • the second sorting module 400 may include:
  • the matching unit is configured to match the first medical item with a standard medical item that is pre-sorted and completed in the fixed knowledge.
  • the first sorting unit is configured to sort the first medical item according to the standard medical item that is successfully matched if the matching is successful.
  • the third sorting module may include:
  • the second sorting unit is configured to generate a first sequence by sorting the first medical item according to an order among standard medical items in the fixed knowledge.
  • the inserting unit is configured to select a current fourth medical item from the medical items to be processed, insert the current medical item to be processed into different positions of the first sequence to form a second sequence, and calculate a probability of the different second sequence.
  • a third sorting unit configured to obtain the position corresponding to the second sequence with the highest probability as the insertion position of the current pending medical item, select the next pending medical item from the pending medical items as the current pending medical item, and continue Insert the current to-be-processed medical items into different positions of the first sequence to form a second sequence, until the to-be-processed medical items are inserted into the first sequence to obtain the best-ranked sequence.
  • the difference value is calculated according to the following method:
  • p1 and p2 are corresponding probabilities at different positions where the current to-be-treated medical item is inserted into the first sequence.
  • Each module in the above-mentioned medical document auditing device may be implemented in whole or in part by software, hardware, and a combination thereof.
  • the above-mentioned modules may be embedded in the hardware form or independent of the processor in the computer device, or may be stored in the memory of the computer device in the form of software, so that the processor calls and performs the operations corresponding to the above modules.
  • a computer device is provided.
  • the computer device may be a terminal, and its internal structure diagram may be as shown in FIG. 6.
  • the computer equipment includes a processor, a memory, a network interface, a display screen, and an input device connected through a system bus.
  • the processor of the computer device is used to provide computing and control capabilities.
  • the memory of the computer device includes a non-volatile storage medium and an internal memory.
  • the non-volatile storage medium stores an operating system and computer-readable instructions.
  • the internal memory provides an environment for operating systems and computer-readable instructions in a non-volatile storage medium.
  • the network interface of the computer device is used to communicate with an external terminal through a network connection.
  • the computer-readable instructions are executed by a processor to implement a method for reviewing medical documents.
  • the display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen.
  • the input device of the computer device may be a touch layer covered on the display screen, or a button, a trackball, or a touchpad provided on the computer device casing. , Or an external keyboard, trackpad, or mouse.
  • FIG. 6 is only a block diagram of a part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer equipment to which the scheme of the present application is applied.
  • the specific computer equipment may be Include more or fewer parts than shown in the figure, or combine certain parts, or have a different arrangement of parts.
  • a computer device includes a memory and one or more processors.
  • Computer-readable instructions are stored in the memory.
  • the one or more processors execute the following steps: Review medical documents, and extract the diagnosis result set and current medical items from the pending medical documents; obtain the current diagnosis results from the diagnosis result set, and extract the pre-stored clinical path template set from the server, which will be compared with the current diagnosis results and clinical
  • the standard diagnostic results in the path template set are compared to obtain the current clinical path template.
  • the current clinical path template stores each standard medical item in chronological order; according to the order of each standard medical item in the current clinical path template, the current medical item is Sort the preset time intervals to obtain the first medical item; obtain fixed knowledge that pre-stores the characterization of the order between the standard medical items, and perform each preset time according to the order between the standard medical items in the fixed knowledge. First medical item in the interval Sort; and the first medical items sorted review.
  • the first medical items in each preset time interval are sorted according to an order between standard medical items in fixed knowledge, and may further include: Determine whether there are pending medical items that cannot be sorted according to the order between the standard medical items in the fixed knowledge; when there are pending medical items, according to the probability model, the medical items to be processed and according to the standards in the fixed knowledge
  • the first medical item after ordering among medical items is sorted to get the best sorted sequence.
  • the probability model uses single-data training performed by nursing doctors with different timestamps with different diseases to generate and characterize the ranking of each standard medical item. Sequential probability values; and when the processor executes the computer-readable instructions, the following steps are further implemented: the review of the ranked first medical item may include: reviewing the best ranked sequence.
  • the processor executes the computer-readable instructions, the following steps are further implemented: when the medical items to be processed are processed according to a probability model and the first medical items that are sorted according to the order between standard medical items in fixed knowledge are obtained When there are multiple best sort sequences, the difference values of the probabilities corresponding to the multiple best sort sequences are calculated; when the difference value is less than a preset value, the sorted, Corresponding different medical items, and medical items that are not sorted according to the current clinical path model are taken as the current medical items; and after obtaining the next diagnosis result as the current diagnosis result, continue to obtain the current clinical path corresponding to the current diagnosis result Template until the diagnosis result traversal is completed from the document information extraction.
  • the processor when the processor executes the computer-readable instructions, the processor further implements the following steps: when the unsorted current medical items still exist after the diagnosis result traversal in the diagnostic result set is completed, the unsorted current medical items are output; receiving Entered modification instructions for unsorted current medical items; when the modification instructions indicate that the unsorted current medical items are unreasonable, mark the unsorted current medical items as unreasonable medical items; and when the modification instructions indicate unsorted medical items When the current medical item is reasonable, the unsorted current medical item is inserted into the generated best sorted sequence according to the modification instruction, and the probability model is updated according to the best sorted sequence after the unsorted current medical item is inserted.
  • the fourth medical item to be processed and the first medical item sorted according to the order between the standard medical items in the fixed knowledge are obtained according to a probability model implemented when the processor executes the computer-readable instructions.
  • the optimal sorting sequence may include: generating a first sequence from a first medical item sorted according to an order among standard medical items in fixed knowledge; selecting a current pending medical item from the pending medical items, and The second sequence is formed at different positions inserted into the first sequence of the medical treatment item, and the probability of the different second sequence is calculated; and the position corresponding to the second sequence with the highest probability is taken as the current insertion position of the medical item to be processed.
  • the next pending medical item is selected as the current pending medical item, and the current pending medical item is continuously inserted into different positions of the first sequence to form a second sequence until the pending medical items are inserted into the first sequence. To get the best sorted sequence.
  • One or more non-volatile storage media storing computer-readable instructions, and when the computer-readable instructions are executed by one or more processors, the one or more processors implement the medical treatment provided in any one of the embodiments of the present application. Steps of the document review method.
  • Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory.
  • Volatile memory can include random access memory (RAM) or external cache memory.
  • RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous chain Synchlink DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
  • SRAM static RAM
  • DRAM dynamic RAM
  • SDRAM synchronous DRAM
  • DDRSDRAM dual data rate SDRAM
  • ESDRAM enhanced SDRAM
  • SLDRAM synchronous chain Synchlink DRAM
  • Rambus direct RAM
  • DRAM direct memory bus dynamic RAM
  • RDRAM memory bus dynamic RAM

Landscapes

  • Health & Medical Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Epidemiology (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Primary Health Care (AREA)
  • Public Health (AREA)
  • Business, Economics & Management (AREA)
  • Biomedical Technology (AREA)
  • General Business, Economics & Management (AREA)
  • Medical Treatment And Welfare Office Work (AREA)

Abstract

本申请涉及提供一种医疗单据审核方法，包括：从服务器获取到待审核医疗单据，并从所述待审核医疗单据中提取到诊断结果集合和当前医疗项目；从所述诊断结果集合中获取当前诊断结果，并获取与所述当前诊断结果对应的当前临床路径模板；根据所述当前临床路径模板将所述当前医疗项目按照预设时间间隔进行排序得到第一医疗项目；获取固定知识，并根据所述固定知识中的标准医疗项目之间的顺序对每一预设时间间隔内的所述第一医疗项目进行排序；对所述排序后的第一医疗项目进行审核。

Description

医疗单据审核方法、装置、计算机设备和存储介质
相关申请的交叉引用
本申请要求于2018年8月14日提交中国专利局，申请号为2018109243134，申请名称为“医疗单据审核方法、装置、计算机设备和存储介质”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及一种医疗单据审核方法、装置、计算机设备和存储介质。
背景技术
随着计算机技术的发展，出现了线上医疗结算技术，即根据用户所提交的医疗单据对所产生的医疗费用进行结算。
然而，发明人意识到，目前的医疗结算管控系统多依赖于对某一个指标的阈值把控或基于数据模型对于结算单据进行风险分类。这样的方式将所有的医疗项目放在同一个时间点进行审核，进而可能造成审核出现错误。
发明内容
基于此，有必要针对上述技术问题，提供一种能够提高审核准确率的医疗单据审核方法、装置、计算机设备和存储介质。
一种医疗单据审核方法，所述方法包括：
从服务器获取到待审核医疗单据，并从所述待审核医疗单据中提取到诊断结果集合和当前医疗项目；
从所述诊断结果集合中获取当前诊断结果，并从服务器提取到预先存储的临床路径模板集合，将所述当前诊断结果与所述临床路径模板集合中的标准诊断结果进行比对得到当前临床路径模板，所述当前临床路径模板中按照时间顺序存储了各个标准医疗项目；
根据所述当前临床路径模板中各个标准医疗项目的顺序，将所述当前医疗项目按照预设时间间隔进行排序得到第一医疗项目；
获取预先存储的表征了所述标准医疗项目之间的顺序的固定知识，并根据所述固定知识中的标准医疗项目之间的顺序对每一预设时间间隔内的所述第一医疗项目进行排序；及
对所述排序后的第一医疗项目进行审核根据所述固定知识中的标准医疗项目之间的顺序。
根据所述固定知识中的标准医疗项目之间的顺序根据所述固定知识中的标准医疗项目之间的顺序根据所述固定知识中的标准医疗项目之间的顺序根据所述固定知识中的标 准医疗项目之间的顺序根据所述固定知识中的标准医疗项目之间的顺序根据所述固定知识中的标准医疗项目之间的顺序根据所述固定知识中的标准医疗项目之间的顺序一种医疗单据审核装置，所述装置包括：
提取模块，用于从服务器获取到待审核医疗单据，并从所述待审核医疗单据中提取到诊断结果集合和当前医疗项目；
临床路径模板获取模块，用于从所述诊断结果集合中获取当前诊断结果，从服务器提取到预先存储的临床路径模板集合，将所述当前诊断结果与所述临床路径模板集合中的标准诊断结果进行比对得到当前临床路径模板，所述当前临床路径模板中按照时间顺序存储了各个标准医疗项目；
第一排序模块，用于根据所述当前临床路径模板中各个标准医疗项目的顺序，将所述当前医疗项目按照预设时间间隔进行排序得到第一医疗项目；
第二排序模块，用于获取预先存储的表征了所述标准医疗项目之间的顺序的固定知识，并根据所述固定知识中的标准医疗项目之间的顺序对每一预设时间间隔内的所述第一医疗项目进行排序；及
审核模块，用于对所述排序后的第一医疗项目进行审核根据所述固定知识中的标准医疗项目之间的顺序。
一种计算机设备，包括存储器和一个或多个处理器，存储器中存储有计算机可读指令，计算机可读指令被处理器执行时实现本申请任意一个实施例中提供的医疗单据审核方法的步骤。
一个或多个存储有计算机可读指令的非易失性存储介质，计算机可读指令被一个或多个处理器执行时，使得一个或多个处理器实现本申请任意一个实施例中提供的医疗单据审核方法的步骤。
根据所述固定知识中的标准医疗项目之间的顺序根据所述固定知识中的标准医疗项目之间的顺序根据固定知识中的标准医疗项目之间的顺序本申请的一个或多个实施例的细节在下面的附图和描述中提出。本申请的其它特征和优点将从说明书、附图以及权利要求书变得明显。
附图说明
为了更清楚地说明本申请实施例中的技术方案，下面将对实施例中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其它的附图。
图1为根据一个或多个实施例中医疗单据审核方法的应用场景图。
图2为根据一个或多个实施例中医疗单据审核方法的流程示意图。
图3为根据另一个或多个实施例中医疗单据审核方法的流程示意图。
图4为根据一个或多个实施例中的最佳排序序列插入过程的示意图。
图5为根据一个或多个实施例中医疗单据审核装置的框图。
图6为根据一个或多个实施例中计算机设备的框图。
具体实施方式
为了使本申请的技术方案及优点更加清楚明白，以下结合附图及实施例，对本申请进行进一步详细说明。应当理解，此处描述的具体实施例仅仅用以解释本申请，并不用于限定本申请。
本申请提供的医疗单据审核方法，可以应用于如图1所示的应用环境中。审核终端102通过网络与服务器104进行通信。审核终端102从服务器104获取到待审核医疗单据，然后从待审核医疗单据中提取诊断结果和当前医疗项目，依次获取诊断结果作为当前诊断结果，然后获取与当前诊断结果对应的当前临床路径模板，根据当前临床路径模板将当前医疗项目按照预设时间间隔进行排序得到第一医疗项目，其中第一医疗项目可以是按天排序得到每一天中的医疗项目，然后根据固定知识中的标准医疗项目之间的顺序对每一预设时间间隔内的，例如每一天中的第一医疗项目进行排序，从而对排序后的第一医疗项目进行审核以作为待审核医疗单据的审核结果，而并不是将所有的医疗项目均按照在同一时间节点进行审核，提高了审核的准确率。审核终端102可以但不限于是各种个人计算机、笔记本电脑、智能手机、平板电脑和便携式可穿戴设备，服务器104可以用独立的服务器或者是多个服务器组成的服务器集群来实现。
在其中一个实施例中，如图2所示，提供了一种医疗单据审核方法，以该方法应用于图1中的审核终端为例进行说明，包括以下步骤：
S202：从服务器获取到待审核医疗单据，并从待审核医疗单据中提取到诊断结果和当前医疗项目。
具体地，待审核医疗单据是待审核患者在一次就诊过程中产生的，其可以包括诊断结果和医疗项目，其中医疗项目可以包括收费项目名称、项目数量以及金额等。
一般地，在医院看病过程中，医院的服务器会将在医院部署的各个医院终端中产生的待审核医疗单据进行收集，从而服务器中存储有每一个患者对应的待审核医疗单据，从而审核终端可以从服务器获取到待审核患者对应的待审核医疗单据，并从该单据信息中提取到诊断结果集合和当前医疗项目，诊断结果集合和当前医疗项目即是待审核患者在一次就诊过程中产生的。
S204：从诊断结果集合中获取当前诊断结果，并从服务器提取到预先存储的临床路径模板集合，将所述当前诊断结果与所述临床路径模板集合中的标准诊断结果进行比对得到当前临床路径模板，当前临床路径模板中按照时间顺序存储了各个标准医疗项目。
具体地，临床路径(Clinical pathway)是指针对某一疾病建立一套标准化治疗模式与治疗程序，是一个有关临床治疗的综合模式，以循证医学证据和指南为指导来促进治疗组织和疾病管理的方法，最终起到规范医疗行为，减少变异，降低成本，提高质量的作用。 每一疾病对应的临床路径模板中按时间存储有各个医疗项目。
诊断结果集合中的诊断结果是指患者所患的疾病，例如其可以是疾病的名称或者是疾病编码，根据诊断结果可以匹配到与诊断结果对应的临床路径模板。且可选地，如果存在多个诊断结果，则可以首先选取其中一个诊断结果作为当前诊断结果，然后通过当前诊断结果匹配到对应的当前临床路径模板。
S206：根据当前临床路径模板中各个标准医疗项目的顺序，将当前医疗项目按照预设时间间隔进行排序得到第一医疗项目。
具体地，由于临床路径模板中按时间存储有各个医疗项目，因此可以将各个医疗项目按照临床路径模板中划分的时间进行排序，例如按天排序以得到每一天中的第一医疗项目，例如当前医疗项目存在10个，经过当前临床路径排序后，将10个当前医疗项目的3个排序到第一天中，则这3个当前医疗项目作为第一天的第一医疗项目，当前医疗项目的4个排序到第二天中，则这4个当前医疗项目作为第二天的第一医疗项目，剩余的3个则无法根据当前临床路径进行排序。
可选地，可以通过NLP(Neuro-Linguistic Programming，神经语言程序学)将所收集的医疗项目的名称与临床路径模板中的医疗项目的名称进行匹配，并将匹配成功的医疗项目根据临床路径模板中的划分时间按天进行划分以得到每一天中的第一医疗项目。
S208：获取预先存储的表征了所述标准医疗项目之间的顺序的固定知识，并根据固定知识中的标准医疗项目之间的顺序对每一预设时间间隔内的第一医疗项目进行排序根据固定知识中的标准医疗项目之间的顺序。
具体地，固定知识是指给出了医疗项目之间的先后顺序的知识，例如其可以指出血糖、肝功能等需要空腹检查等，药品的用药需要按照时间段进行平均分配等。
因此，审核终端可以先获取到每一天中的第一医疗项目，然后对该第一医疗项目根据固定知识中的标准医疗项目之间的顺序进行排序，具体地，其排序方式也是将第一医疗项目的名称与固定知识中的医疗项目的名称进行匹配，在匹配成功后，则获取到对应的固定知识对应的医疗项目的先后顺序，从而根据该先后顺序对第一医疗项目进行排序。
S210：对排序后的第一医疗项目进行审核。
具体地，为了避免医疗单据审核时，对所有的医疗单据的都按照同一时间点发生的进行审核，因此首先从医疗单据中提取了诊断结果和当前医疗项目，然后根据诊断结果得到临床路径模板，从而对当前医疗项目按天排序，并且为了进一步地提高排序准确性，又通过固定知识进行排序，从而将医疗单据所对应的医疗项目按照时间点进行排序后，再对各个排序后的医疗项目进行审核，从而按照时间节点的顺序依次对医疗项目进行审核，可以提高审核的准确性。
上述医疗单据审核方法，在获取到单据信息后，并不直接对单据信息进行审核，而是从单据信息中提取诊断结果和当前医疗项目，获取到诊断结果对应的当前临床路径模板，根据当前临床路径模板对当前医疗项目按天进行排序，并根据固定知识中的标准医疗项目 之间的顺序将按天排序后的第一医疗项目进行排序，从而可以对排序后的第一医疗项目进行审核，而并不是将所有的医疗项目均按照在同一时间节点进行审核，提高了审核的准确率。
在其中一个实施例中，根据固定知识中的标准医疗项目之间的顺序中标准医疗项目之间的顺序对每一预设时间间隔内的第一医疗项目进行排序之后，还可以包括：判断是否存在无法根据固定知识中的标准医疗项目之间的顺序中标准医疗项目之间的顺序进行排序的待处理医疗项目；当待处理医疗项目时，则根据概率模型，对待处理医疗项目以及根据固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序，得到最佳排序序列概率模型利用不同疾病带有的不同时间戳的护理医嘱执行单数据训练生成且表征了各个标准医疗项目的排序顺序的概率值。从而对排序后的第一医疗项目进行审核，包括：对最佳排序序列进行审核。
具体地，概率模型是利用不同疾病带有的不同时间戳的护理医嘱执行单数据训练生成的，其给出了各个诊断结果，即疾病对应的各个医疗项目的排序顺序，将不同的医疗项目输入到概率模型中可以得到该不同的医疗项目的排序顺序以及不同的排序顺序对应的概率。
其中待处理医疗项目可以包括两种，一种是在将第一医疗项目根据固定知识中的标准医疗项目之间的顺序进行排序时，存在的与固定知识匹配成功，但是无法根据固定知识中的标准医疗项目之间的顺序进行排序的第二医疗项目，另外一种是没有与固定知识匹配成功的第三医疗项目。其中第二和第三仅是为了区分，在此不做具体限制，即审核终端将第一医疗项目与固定知识进行匹配时，存在第二医疗项目与固定知识中的医疗项目匹配成功，但是根据固定知识中的标准医疗项目之间的顺序其不能确定与其他的第一医疗项目之间的先后顺序，即该第二医疗项目可能存在多个不同的排序位置，或者是无法与固定知识匹配成功的第三医疗项目。因此审核终端在将第一医疗项目根据固定知识中的标准医疗项目之间的顺序进行排序后，审核终端需要判断是否存在根据固定知识中的标准医疗项目之间的顺序排序后存在不同排序位置的第二医疗项目以及是否存在无法根据固定知识中的标准医疗项目之间的顺序进行排序的第三医疗项目，也即判断是否存在无法根据固定知识中的标准医疗项目之间的顺序进行排序的待处理医疗项目。
具体地，在实际应用中，审核终端可以首先将第一医疗项目的名称与固定知识中的医疗项目的名称进行匹配，以得到可以根据固定知识中的标准医疗项目之间的顺序进行排序的第二医疗项目和不能根据固定知识中的标准医疗项目之间的顺序进行排序的第三医疗项目，然后审核终端根据固定知识中的标准医疗项目之间的顺序对可以根据固定知识中的标准医疗项目之间的顺序进行排序的第二医疗项目进行排序，并根据排序结果得到根据固定知识中的标准医疗项目之间的顺序排序后存在不同排序位置的第二医疗项目。其中可选地，在某个实施例中可能仅存在根据固定知识中的标准医疗项目之间的顺序排序后存在不同排序位置的第二医疗项目，或者仅存在不能根据固定知识中的标准医疗项目之间的顺序 进行排序的第三医疗项目，或者既存在不能根据固定知识中的标准医疗项目之间的顺序进行排序的第三医疗项目，还存在根据固定知识中的标准医疗项目之间的顺序排序后存在不同排序位置的第二医疗项目。
在审核终端获取到没有排序的待处理医疗项目后，则将该待处理医疗项目和已经排序的第一医疗项目输入至概率模型中，以将待处理医疗项目按照一定顺序插入至已经排序的第一医疗项目中，最后形成包含该一天中的第一医疗项目的最佳排序序列。可选地，审核终端可以计算将待处理医疗项目插入至已经排序的第一医疗项目中的不同位置处的概率，选取概率最大的一个排序序列作为最佳排序序列。且从而对排序后的第一医疗项目进行审核即是对该最佳排序序列进行审核。
上述实施例中，扩展了应用场景，将没有固定知识部分的第三医疗项目，以及根据固定知识中的标准医疗项目之间的顺序部分无法排序的第二医疗项目，即上述的待处理医疗项目通过概率模型生成最佳序列，完成路径的重塑。
在其中一个实施例中，上述医疗单据审核方法还可以包括：当根据概率模型，对待处理医疗项目以及根据固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序，得到多个最佳排序序列时，则计算所得到的多个最佳排序序列对应的概率的差异值；当差异值小于预设值时，则获取所得到的多个最佳排序序列中的、排序不同的医疗项目，以及未根据当前临床路径模型进行排序的医疗项目，作为当前医疗项目；获取下一诊断结果作为当前诊断结果后，继续获取与当前诊断结果对应的当前临床路径模板，直至从单据信息中提取到诊断结果遍历完成。
在其中一个实施例中，上述医疗单据审核方法还可以包括：当诊断结果集合中的诊断结果遍历完成后，仍存在未排序的当前医疗项目，则未排序的当前医疗项目；接收输入的针对未排序的当前医疗项目的修改指令；当修改指令表示未排序的当前医疗项目不合理时，则将未排序的当前医疗项目标记为不合理医疗项目；当修改指令表示未排序的当前医疗项目合理时，则根据修改指令，将未排序的当前医疗项目插入至所生成的最佳排序序列中，并根据插入了未排序的当前医疗项目后的最佳排序序列更新概率模型。
具体地，参见图3，图3为另一实施例中的医疗单据审核方法的流程图，首先对下文出现的名词进行解释：第五医疗项目是根据概率模型进行排序后，两个概率的差异值小于预设值的最佳排序序列中排序位置不同的医疗项目。第六医疗项目是在于临床路径进行匹配的时候，与当前临床路径匹配失败的医疗项目。其中在一个实施例中，可以仅包含第五医疗项目或第六医疗项目，或者还可以既包含第五医疗项目又包含第六医疗项目。第七医疗项目是指上文中的未排序的当前医疗项目，即在所有诊断结果遍历完成后，仍存在第五医疗项目和/或第六医疗项目，该存在的第七医疗项目需要引入人工干预以进行指导。
在该实施例中不仅包含上文中的内容，还可以包含以下内容：
当审核终端对每一天中的第一医疗项目排序后，存在待处理医疗项目，即没有固定知识部分的第三医疗项目，以及根据固定知识中的标准医疗项目之间的顺序部分无法排序的 第二医疗项目时，则通过概率模型再次进行排序以完成路径重塑，但是如果通过概率模型进行排序得到多个最佳排序序列，即概率模型输出了多个最佳排序序列，则可以计算多个最佳排序序列对应的概率的差异值，即首先获取到所生成的最佳排序序列对应的概率，该概率是由概率模型生成并输出的，然后计算多个最佳排序序列对应的概率的差异值，当差异值小于预设值时，则所获得的多个最佳排序序列存在排序不相同的部分，因此可以获取所得到的多个最佳排序序列中排序不同的第五医疗项目，以在后续将该第五医疗项目进行重新排序。
审核终端判断是否存在未根据当前临床路径模型按天排序的第六医疗项目，即未与当前临床路径匹配成功的第六医疗项目，如果存在，则将第五医疗项目和第六医疗项目作为当前医疗项目。可选地，如果不存在第六医疗项目，则仅将第五医疗项目作为当前医疗项目，且此外，如果不存在第五医疗项目，则可以仅将第六医疗项目作为当前医疗项目，并获取从医疗单据中所提取的当前诊断结果的下一诊断结果作为当前诊断结果，然后继续获取与当前诊断结果对应的当前临床路径模板，直至从单据信息中提取到诊断结果遍历完成，以将单据信息中所涉及的诊断结果和医疗项目全部遍历完成，根据诊断结果将医疗项目进行排序完成，从而可以将所有的医疗项目按照时间进行路径的重塑，以便于后续的审核。
且可以选地，在上述实施例中，当当从单据信息中提取到诊断结果遍历完成后，仍存在第五医疗项目和第六医疗项目其中至少任意一种，则将仍存在的没有排序完成的医疗项目进行输出，即将未排序完成的当前医疗项目输出，以便引入人工干预，即审核用户首先判断输出的未排序的当前医疗项目即第七医疗项目是否合理，该判断是根据人工经验进行的，即该疾病是否需要进行该第七医疗项目等，如果不合理，则将第七医疗项目标记为不合理，从而可以更新与当前临床路径对应的临床路径外清单模板，以便于下次在得到该临床路径对应的该第七医疗项目时，直接标记该第七医疗项目为不合理医疗项目，从而可以避免人工干预。此外，如果该第七医疗项目是合理的，则根据人工修改指令，将该第七医疗项目插入至所生成的最佳排序序列中，并根据出入第七医疗项目后的最佳排序序列更新概率模型，从而可以在下一次通过概率模型直接进行排序，而不需要再引入人工干预，提高效率。
上述实施例中，当人工审核的结果为项目不合理时，则更新临床路径外项目清单模板。当人工审核的结果为项目合理但是概率模型存在问题时，则根据人工修改指令对概率模型进行修改。即将合理的项目人工插入序列中，并更新概率模型，将不合理项目作为临床路径外项目清单模板，用于不合理管控，以提高效率。
在其中一个实施例中，根据固定知识中的标准医疗项目之间的顺序中标准医疗项目之间的顺序对每一预设时间间隔内的第一医疗项目进行排序，可以包括：将第一医疗项目与固定知识中的预先排序完成的标准医疗项目进行匹配；如果匹配成功，则按照匹配成功的标准医疗项目对第一医疗项目进行排序。
具体地，在根据固定知识中的标准医疗项目之间的顺序进行排序时，首先将第一医疗项目与固定知识中的医疗项目进行匹配，如果匹配成功，则说明匹配成功的第一医疗项目存在固定知识，然后根据匹配成功的医疗项目对应的固定知识对第一医疗项目进行排序。
在实际应用中，以以下例子进行说明：以下都是按照临床路径模板划分到住院第一天的项目，目前要将其做一天内按时间轴的排序：血常规、尿常规、凝血功能、感染性疾病筛查、肝功能、肾功能、电解质、血糖、胸片、心电图、锁骨正侧位X线片、锁骨三维CT检查、双下肢血管超声、肺功能、超声心动图、血气分析。
首先，审核终端按照固定知识进行排序，如血气分析需要在肺功能之前完成，则标记其序号分别为1_a、2_a(_a为自动添加的随机标记符)；同理锁骨正侧位X线片按常理在锁骨三维CT检查前完成，则标记其为1_b、2_b(同理_b为自动添加的随机标记符)；按照采血流程，血常规、肝功能、肾功能、电解质、血糖、凝血功能、感染性疾病筛查均同时采样，由此均标记为1_c(同理_c为自动添加的随机标记符)；按照检验知识的要求，血糖、肝功能等需要空腹检查，因此将标记为c的项目统一前置于b、a，其余无序号项目用临时5位随机编号编写表明其自由插入的属性，如下表1，此时明确的顺序情况为1_c、1_a、2_a、1_b、2_b或1_c、1_b、2_b、1_a、2_a，即存在根据固定知识中的标准医疗项目之间的顺序排序后存在不同排序位置的第二医疗项目1_a、2_a、1_b、2_b及未能排序的尿常规、胸片、双下肢血管超声、超声心动图，即存在不能根据固定知识中的标准医疗项目之间的顺序进行排序的第三医疗项目，该第三医疗项目可视为可以任意插入到序列中，这种有两种以上排序方式就形成矛盾部分。到此完成按照固定知识的排序(固定知识除先后顺序外还有药品使用方式，如1日3次则会将药品按6:00-8:00，12:00-14:00，18:00-20:00时间段平均分配，1日2次会按8:00，16:00均分等等，大致顺序确定后再由CRF基于概率排出最大概率的序列)。
表1 根据固定知识中的标准医疗项目之间的顺序排序后的结果。
Figure PCTCN2019099425-appb-000001
可选地，从而审核终端在按照固定知识对第一医疗项目排序后，还存在待处理医疗项目，即没有固定知识的部分的第三医疗项目以及存在矛盾的部分的第二医疗项目，从而可 以基于概率模型进行进一步的排序，假设基于概率模型分析，尿常规通常于抽血项目后完成，胸片、心电图通常在紧随其后，接下来是1_b、2_b等等，即序列“1_c、尿常规、胸片、心电图、1_a、2_a、1_b、2_b、双下肢血管超声、超声心动图”为概率最大序列，即上文中的最佳排序序列。
可选地，如产生两个概率误差差异小于10％的最佳排序序列，此时称出现了矛盾部分，如序列1概率为0.867：“1_c、尿常规、胸片、心电图、1_a、2_a、1_b、2_b、双下肢血管超声、超声心动图”；序列2概率为0.785：“1_c、尿常规、胸片、心电图、1_a、2_a、1_b、2_b、超声心动图、双下肢血管超声”，即超声心动图和双下肢血管超声先后顺序排列，差异值小于预设值(该预设值可人为控制，此处取10％，差异值计算(0.867-0.785)/((0.867+0.785)/2))＝9.93％，则认为根据概率模型计算得到了多个差异值小于预设值的最佳排序序列，排序失败。
上述实施例中，首先根据固定知识中的标准医疗项目之间的顺序对第一医疗项目进行排序，仅需要将第一医疗项目与固定知识进行匹配，简单且效率高。
在其中一个实施例中，根据概率模型，对待处理医疗项目以及根据固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序，得到最佳排序序列，包括：将根据固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目生成第一序列；从待处理医疗项目中选取当前待处理医疗项目，并将当前待处理医疗项目插入第一序列的不同位置处形成第二序列，并计算不同的第二序列的概率；获取概率最大的第二序列所对应的位置作为当前待处理医疗项目的插入位置，从待处理医疗项目中选取下一待处理医疗项目作为当前待处理医疗项目，并继续将当前待处理医疗项目插入第一序列的不同位置处形成第二序列，直至待处理医疗项目均插入至第一序列中得到最佳排序序列。
在其中一个实施例中，差异值是根据以下方式计算得到的：差异值＝(p1-p2)/((p1+p2)/2)。其中，p1和p2为当前待处理医疗项目插入第一序列的不同位置处对应的概率。
具体地，根据概率模型对待处理医疗项目以及根据固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序是：根据诊疗时间轴利用概率模型自动生成判断特征以对固定知识未予明确的顺序给出基于概率模型的最佳排序序列。其中诊疗时间轴是指通过带时间戳的护理执行单形成的有先后顺序的诊疗项目实际序列及人工处理后的项目序列。
具体地，据概率模型对待处理医疗项目以及根据固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序主要包括以下步骤：
首先由带时间戳的护理执行单数据生成项目间的先后顺序概率表，同时完成的项目标注为9，可以参见如下表2，其中列表示在前，如血常规在胸片前的概率为0.81。
表2 医疗项目先后顺序概率表
Figure PCTCN2019099425-appb-000002
Figure PCTCN2019099425-appb-000003
其次，将原序列不确定的部分作为插入值插入，并计算概率，选择概率最大的插入点插入，如此循环完成排序；如前述的序列已排序的内容如下表3和表4，尿常规、胸片、心电图、双下肢血管超声、超声心动图作为插入项目待插入。
表3 已确定的第一序列
空位 1_c 空位 1_a 空位 2_a 空位 1_b 空位 2_b 空位
或表4已确定的第一序列
空位 1_c 空位 1_b 空位 2_b 空位 1_a 空位 2_a 空位
通过概率计算确定(1_a、2_a，1_b、2_b)作为两个组的先后顺序，可视为c-b确定，a插入位置判断的问题。如c-a-b的顺序概率大于c-b-a的概率则保留c-a-b的排序。
其中概率计算为查询概率表，即上述表2，a在c后的概率乘以a在b前的概率得到c-a-b排列的概率，然后以同样逻辑将剩余的待插入项插入，即从待处理医疗项目中选取当前待处理医疗项目，并将当前待处理医疗项目插入第一序列的不同位置处形成第二序列，并计算不同的第二序列的概率；获取概率最大的第二序列所对应的位置作为当前待处理医疗项目的插入位置，具体可以参见图4所示，图4为一个实施例中的最佳排序序列插入过程的示意图。
且可选地，差异值是根据以下方式计算得到的：差异值＝(p1-p2)/((p1+p2)/2)。其中，p1和p2为当前待处理医疗项目插入第一序列的不同位置处对应的概率。当同一个当前待处理医疗项目插入到两个不同位置的概率差值小于预设值10％时，则认为出现了矛盾，其中该预设值可以人工进行控制。
上述实施例中，将根据固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目生成第一序列；从待处理医疗项目中选取当前待处理医疗项目，并将当前待处理医疗项目插入第一序列的不同位置处形成第二序列，并计算不同的第二序列的概率；获取概率最大的第二序列所对应的位置作为当前待处理医疗项目的插入位置，依次循环以得到最佳排序序列，简单可靠。
应该理解的是，虽然图2-3的流程图中的各个步骤按照箭头的指示依次显示，但是这些步骤并不是必然按照箭头指示的顺序依次执行。除非本文中有明确的说明，这些步骤的执行并没有严格的顺序限制，这些步骤可以以其它的顺序执行。而且，图2-3中的至少一部分步骤可以包括多个子步骤或者多个阶段，这些子步骤或者阶段并不必然是在同一时刻执行完成，而是可以在不同的时刻执行，这些子步骤或者阶段的执行顺序也不必然是依次进行，而是可以与其它步骤或者其它步骤的子步骤或者阶段的至少一部分轮流或者交替地 执行。
在一个实施例中，如图5所示，提供了一种医疗单据审核装置，包括：提取模块100、临床路径模板获取模块200、第一排序模块300、第二排序模块400和审核模块500，其中：
提取模块100，用于从服务器获取到待审核医疗单据，从服务器获取到待审核患者对应的单据信息，并从待审核医疗单据中提取到诊断结果集合和当前医疗项目。
临床路径模板获取模块200，用于从服务器提取到预先存储的临床路径模板集合，将当前诊断结果与临床路径模板集合中的标准诊断结果进行比对得到当前临床路径模板，当前临床路径模板中按照时间顺序存储了各个标准医疗项目。
第一排序模块300，用于根据当前临床路径模板将当前医疗项目按照预设时间间隔进行排序得到第一医疗项目。
第二排序模块400，用于获取预先存储的表征了标准医疗项目之间的顺序的固定知识，并根据固定知识中所述标准医疗项目之间的顺序对每一预设时间间隔内的第一医疗项目进行排序。
审核模块500，用于对排序后的第一医疗项目进行审核。
在其中一个实施例中，上述医疗单据审核装置还可以包括：
第一判断模块，用于判断是否存在无法根据固定知识中的标准医疗项目之间的顺序进行排序的待处理医疗项目。
第三排序模块，用于当存在待处理医疗项目时，则根据概率模型，对待处理医疗项目以及根据固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序，得到最佳排序序列，概率模型利用不同疾病带有的不同时间戳的护理医嘱执行单数据训练生成且表征了各个标准医疗项目的排序顺序的概率值。
审核模块500还用于对最佳排序序列进行审核。
在其中一个实施例中，上述医疗单据审核装置还可以包括：
计算模块，用于当根据概率模型，对待处理医疗项目以及根据固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序得到，多个最佳排序序列时，则计算所得到的多个最佳排序序列对应的概率的差异值。
第二判断模块，用于当差异值小于预设值时，则获取所得到的多个最佳排序序列中的、排序不同的医疗项目，以及未根据所述当前临床路径模型进行排序的医疗项目，作为当前医疗项目。
遍历模块，用于获取下一诊断结果作为当前诊断结果后，继续获取与当前诊断结果对应的当前临床路径模板，直至从单据信息中提取到诊断结果遍历完成。
在其中一个实施例中，上述医疗单据审核装置还可以包括：
输出模块，用于当诊断结果集合中的诊断结果遍历完成后，仍存在未排序的当前医疗项目，则输出未排序的当前医疗项目。
接收模块，用于接收输入的针对未排序的当前医疗项目的修改指令。
标记模块，用于当修改指令表示未排序的当前医疗项目不合理时，则将未排序的当前医疗项目标记为不合理医疗项目。
更新模块，用于当修改指令表示未排序的当前医疗项目合理时，则根据修改指令将未排序的当前医疗项目插入至所生成的最佳排序序列中，并根据插入未排序的当前医疗项目后的最佳排序序列更新概率模型。
在其中一个实施例中，第二排序模块400可以包括：
匹配单元，用于将第一医疗项目与固定知识中的预先排序完成的标准医疗项目进行匹配。
第一排序单元，用于如果匹配成功，则按照匹配成功的标准医疗项目对第一医疗项目进行排序。
在其中一个实施例中，第三排序模块可以包括：
第二排序单元，用于将根据固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目生成第一序列。
插入单元，用于从待处理医疗项目中选取当前第四医疗项目，并将当前待处理医疗项目插入第一序列的不同位置处形成第二序列，并计算不同的第二序列的概率。
第三排序单元，用于获取概率最大的第二序列所对应的位置作为当前待处理医疗项目的插入位置，从待处理医疗项目中选取下一待处理医疗项目作为当前待处理医疗项目，并继续将当前待处理医疗项目插入第一序列的不同位置处形成第二序列，直至待处理医疗项目均插入至第一序列中得到最佳排序序列。
在其中一个实施例中，差异值是根据以下方式计算得到的：
差异值＝(p1-p2)/((p1+p2)/2)
其中，p1和p2为当前待处理医疗项目插入第一序列的不同位置处对应的概率。
关于医疗单据审核装置的具体限定可以参见上文中对于医疗单据审核方法的限定，在此不再赘述。上述医疗单据审核装置中的各个模块可全部或部分通过软件、硬件及其组合来实现。上述各模块可以硬件形式内嵌于或独立于计算机设备中的处理器中，也可以以软件形式存储于计算机设备中的存储器中，以便于处理器调用执行以上各个模块对应的操作。
在一个实施例中，提供了一种计算机设备，该计算机设备可以是终端，其内部结构图可以如图6所示。该计算机设备包括通过系统总线连接的处理器、存储器、网络接口、显示屏和输入装置。其中，该计算机设备的处理器用于提供计算和控制能力。该计算机设备的存储器包括非易失性存储介质、内存储器。该非易失性存储介质存储有操作系统和计算机可读指令。该内存储器为非易失性存储介质中的操作系统和计算机可读指令的运行提供环境。该计算机设备的网络接口用于与外部的终端通过网络连接通信。该计算机可读指令被处理器执行时以实现一种医疗单据审核方法。该计算机设备的显示屏可以是液晶显示屏 或者电子墨水显示屏，该计算机设备的输入装置可以是显示屏上覆盖的触摸层，也可以是计算机设备外壳上设置的按键、轨迹球或触控板，还可以是外接的键盘、触控板或鼠标等。
本领域技术人员可以理解，图6中示出的结构，仅仅是与本申请方案相关的部分结构的框图，并不构成对本申请方案所应用于其上的计算机设备的限定，具体的计算机设备可以包括比图中所示更多或更少的部件，或者组合某些部件，或者具有不同的部件布置。
一种计算机设备，包括存储器和一个或多个处理器，存储器中储存有计算机可读指令，计算机可读指令被处理器执行时，使得一个或多个处理器执行以下步骤：从服务器获取到待审核医疗单据，并从待审核医疗单据中提取到诊断结果集合和当前医疗项目；从诊断结果集合中获取当前诊断结果，从服务器提取到预先存储的临床路径模板集合，将与当前诊断结果与临床路径模板集合中的标准诊断结果进行比对得到当前临床路径模板，当前临床路径模板中按照时间顺序存储了各个标准医疗项目；根据当前临床路径模板中各个标准医疗项目的顺序，将当前医疗项目按照预设时间间隔进行排序得到第一医疗项目；获取预先存储的表征了所述标准医疗项目之间的顺序的固定知识，并根据固定知识中的标准医疗项目之间的顺序对每一预设时间间隔内的第一医疗项目进行排序；及对排序后的第一医疗项目进行审核。
在一个实施例中，处理器执行计算机可读指令时所实现的根据固定知识中的标准医疗项目之间的顺序对每一预设时间间隔内的第一医疗项目进行排序之后，还可以包括：判断是否存在无法根据所述固定知识中的标准医疗项目之间的顺序进行排序的待处理医疗项目；当存在待处理医疗项目时，则根据概率模型，对待处理医疗项目以及根据固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目，进行排序得到最佳排序序列，概率模型利用不同疾病带有的不同时间戳的护理医嘱执行单数据训练生成且表征了各个标准医疗项目的排序顺序的概率值；及处理器执行计算机可读指令时还实现以下步骤：对排序后的第一医疗项目进行审核，可以包括：对最佳排序序列进行审核。
在一个实施例中，处理器执行计算机可读指令时还实现以下步骤：当根据概率模型对待处理医疗项目以及根据固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序得到多个最佳排序序列时，则计算所得到的多个最佳排序序列对应的概率的差异值；当差异值小于预设值时，则获取所得到的多个最佳排序序列中排序的、对应的不同的医疗项目，以及未根据所述当前临床路径模型进行排序的医疗项目，作为当前医疗项目；及获取下一诊断结果作为当前诊断结果后，继续获取与当前诊断结果对应的当前临床路径模板，直至从单据信息中提取到诊断结果遍历完成。
在一个实施例中，处理器执行计算机可读指令时还实现以下步骤：当诊断结果集合中的诊断结果遍历完成后，仍存在未排序的当前医疗项目，则输出未排序的当前医疗项目；接收输入的针对未排序的当前医疗项目的修改指令；当修改指令表示未排序的当前医疗项目不合理时，则将未排序的当前医疗项目标记为不合理医疗项目；及当修改指令表示未排序的当前医疗项目合理时，则根据修改指令，将未排序的当前医疗项目插入至所生成的最 佳排序序列中，并根据插入了未排序的当前医疗项目后的最佳排序序列更新概率模型。
在一个实施例中，处理器执行计算机可读指令时所实现的根据概率模型对第四待处理医疗项目以及根据固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序得到最佳排序序列，可以包括：将根据固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目生成第一序列；从待处理医疗项目中选取当前待处理医疗项目，并将当前待处理医疗项目插入第一序列的不同位置处形成第二序列，并计算不同的第二序列的概率；及获取概率最大的第二序列所对应的位置作为当前待处理医疗项目的插入位置，从待处理医疗项目中选取下一待处理医疗项目作为当前待处理医疗项目，并继续将当前待处理医疗项目插入第一序列的不同位置处形成第二序列，直至待处理医疗项目均插入至第一序列中得到最佳排序序列。
一个或多个存储有计算机可读指令的非易失性存储介质，计算机可读指令被一个或多个处理器执行时，使得一个或多个处理器实现本申请任意一个实施例中提供的医疗单据审核方法的步骤。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程，是可以通过计算机可读指令来指令相关的硬件来完成，所述的计算机可读指令可存储于一非易失性计算机可读取存储介质中，该计算机可读指令在执行时，可包括如上述各方法的实施例的流程。其中，本申请所提供的各实施例中所使用的对存储器、存储、数据库或其它介质的任何引用，均可包括非易失性和/或易失性存储器。非易失性存储器可包括只读存储器(ROM)、可编程ROM(PROM)、电可编程ROM(EPROM)、电可擦除可编程ROM(EEPROM)或闪存。易失性存储器可包括随机存取存储器(RAM)或者外部高速缓冲存储器。作为说明而非局限，RAM以多种形式可得，诸如静态RAM(SRAM)、动态RAM(DRAM)、同步DRAM(SDRAM)、双数据率SDRAM(DDRSDRAM)、增强型SDRAM(ESDRAM)、同步链路(Synchlink)DRAM(SLDRAM)、存储器总线(Rambus)直接RAM(RDRAM)、直接存储器总线动态RAM(DRDRAM)、以及存储器总线动态RAM(RDRAM)等。
以上实施例的各技术特征可以进行任意的组合，为使描述简洁，未对上述实施例中的各个技术特征所有可能的组合都进行描述，然而，只要这些技术特征的组合不存在矛盾，都应当认为是本说明书记载的范围。
以上所述实施例仅表达了本申请的几种实施方式，其描述较为具体和详细，但并不能因此而理解为对发明专利范围的限制。应当指出的是，对于本领域的普通技术人员来说，在不脱离本申请构思的前提下，还可以做出若干变形和改进，这些都属于本申请的保护范围。因此，本申请专利的保护范围应以所附权利要求为准。

Claims (20)

  1. 一种医疗单据审核方法，包括：
    从服务器获取到待审核医疗单据，并从所述待审核医疗单据中提取到诊断结果集合和当前医疗项目；
    从所述诊断结果集合中获取当前诊断结果，并从服务器提取到预先存储的临床路径模板集合，将所述当前诊断结果与所述临床路径模板集合中的标准诊断结果进行比对得到当前临床路径模板，所述当前临床路径模板中按照时间顺序存储了各个标准医疗项目；
    根据所述当前临床路径模板中各个标准医疗项目的顺序，将所述当前医疗项目按照预设时间间隔进行排序得到第一医疗项目；
    获取预先存储的表征了所述标准医疗项目之间的顺序的固定知识，并根据所述固定知识中的标准医疗项目之间的顺序对每一预设时间间隔内的所述第一医疗项目进行排序；及
    对所述排序后的第一医疗项目进行审核。
  2. 根据权利要求1所述的方法，其特征在于，所述根据所述固定知识中的标准医疗项目之间的顺序对每一预设时间间隔内的所述第一医疗项目进行排序之后，还包括：
    判断是否存在无法根据所述固定知识中的标准医疗项目之间的顺序进行排序的待处理医疗项目；
    当存在所述待处理医疗项目时，则根据概率模型，对所述待处理医疗项目以及根据所述固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序，得到最佳排序序列，所述概率模型利用不同疾病带有的不同时间戳的护理医嘱执行单数据训练生成且表征了各个标准医疗项目的排序顺序的概率值；及
    所述对所述排序后的第一医疗项目进行审核，包括：
    对所述最佳排序序列进行审核。
  3. 根据权利要求2所述的方法，其特征在于，还包括：
    当根据所述概率模型，对所述待处理医疗项目以及根据所述固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序，得到多个最佳排序序列时，则计算所得到的多个最佳排序序列对应的概率的差异值；
    当所述差异值小于预设值时，则获取所得到的多个最佳排序序列中的、对应的排序不同的医疗项目，以及未根据所述当前临床路径模型进行排序的医疗项目，作为当前医疗项目；及
    获取下一诊断结果作为当前诊断结果后，继续获取与所述当前诊断结果对应的当前临床路径模板，直至从所述单据信息中提取到诊断结果遍历完成。
  4. 根据权利要求3所述的方法，其特征在于，还包括：
    当所述诊断结果集合中的诊断结果遍历完成后，仍存在未排序的当前医疗项目，则输出未排序的当前医疗项目；
    接收输入的针对所述未排序的当前医疗项目的修改指令；
    当所述修改指令表示所述未排序的当前医疗项目不合理时，则将所述未排序的当前医疗项目标记为不合理医疗项目；及
    当所述修改指令表示所述未排序的当前医疗项目合理时，则根据所述修改指令，将所述未排序的当前医疗项目插入至所生成的最佳排序序列中，并根据插入了所述未排序的当前医疗项目后的最佳排序序列更新所述概率模型。
  5. 根据权利要求1至4任意一项所述的方法，其特征在于，所述根据所述固定知识中的标准医疗项目之间的顺序对每一预设时间间隔内的所述第一医疗项目进行排序，包括：
    将所述第一医疗项目与所述固定知识中的预先排序完成的标准医疗项目进行匹配；及
    如果匹配成功，则按照匹配成功的标准医疗项目对所述第一医疗项目进行排序。
  6. 根据权利要求3或4所述的方法，其特征在于，所述根据概率模型，对所述待处理医疗项目以及根据所述固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序，得到最佳排序序列，包括：
    将根据所述固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目生成第一序列；
    从所述待处理医疗项目中选取当前待处理医疗项目，并将所述当前待处理医疗项目插入所述第一序列的不同位置处形成第二序列，并计算不同的第二序列的概率；及
    获取概率最大的第二序列所对应的位置作为所述当前待处理医疗项目的插入位置，从所述待处理医疗项目中选取下一待处理医疗项目作为当前待处理医疗项目，并继续将所述当前待处理医疗项目插入所述第一序列的不同位置处形成第二序列，直至所述待处理医疗项目均插入至所述第一序列中得到最佳排序序列。
  7. 根据权利要求6所述的方法，其特征在于，所述差异值是根据以下方式计算得到的：
    差异值＝(p1-p2)/((p1+p2)/2)
    其中，p1和p2为所述当前待处理医疗项目插入所述第一序列的不同位置处对应的概率。
  8. 一种医疗单据审核装置，其特征在于，所述装置包括：
    提取模块，用于从服务器获取到待审核医疗单据，并从所述待审核医疗单据中提取到诊断结果集合和当前医疗项目；
    临床路径模板获取模块，用于从所述诊断结果集合中获取当前诊断结果，从服务器提取到预先存储的临床路径模板集合，将所述当前诊断结果与所述临床路径模板集合中的标准诊断结果进行比对得到当前临床路径模板，所述当前临床路径模板中按照时间顺序存储了各个标准医疗项目；
    第一排序模块，用于根据所述当前临床路径模板中各个标准医疗项目的顺序，将所述当前医疗项目按照预设时间间隔进行排序得到第一医疗项目；
    第二排序模块，用于获取预先存储的表征了所述标准医疗项目之间的顺序的固定知识，并根据所述固定知识中的标准医疗项目之间的顺序对每一预设时间间隔内的所述第一医疗项目进行排序根据所述固定知识中的标准医疗项目之间的顺序；及
    审核模块，用于对所述排序后的第一医疗项目进行审核。
  9. 根据权利要求8所述的装置，其特征在于，还包括：
    第一判断模块，用于判断是否存在无法根据所述固定知识中的标准医疗项目之间的顺序进行排序的待处理医疗项目；
    第三排序模块，用于当存在所述待处理医疗项目时，则根据概率模型，对所述待处理医疗项目以及根据所述固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序，得到最佳排序序列；
    所述审核模块还用于对所述最佳排序序列进行审核。
  10. 一种计算机设备，包括存储器及一个或多个处理器，所述存储器中储存有计算机可读指令，所述计算机可读指令被所述一个或多个处理器执行时，使得所述一个或多个处理器执行以下步骤：从服务器获取到待审核医疗单据，并从所述待审核医疗单据中提取到诊断结果集合和当前医疗项目；从所述诊断结果集合中获取当前诊断结果，并从服务器提取到预先存储的临床路径模板集合，将所述当前诊断结果与所述临床路径模板集合中的标准诊断结果进行比对得到当前临床路径模板，所述当前临床路径模板中按照时间顺序存储了各个标准医疗项目；根据所述当前临床路径模板中各个标准医疗项目的顺序，将所述当前医疗项目按照预设时间间隔进行排序得到第一医疗项目；获取预先存储的表征了所述标准医疗项目之间的顺序的固定知识，并根据所述固定知识中的标准医疗项目之间的顺序对每一预设时间间隔内的所述第一医疗项目进行排序；及对所述排序后的第一医疗项目进行审核根据所述固定知识中的标准医疗项目之间的顺序。
  11. 根据权利要求10所述的计算机设备，其特征在于，所述处理器执行所述计算机可读指令时所执行的所述根据所述固定知识中的标准医疗项目之间的顺序对每一预设时间间隔内的所述第一医疗项目进行排序之后，还包括：判断是否存在无法根据所述固定知识中的标准医疗项目之间的顺序进行排序的待处理医疗项目；当存在所述待处理医疗项目时，则根据概率模型，对所述待处理医疗项目以及根据所述固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序，得到最佳排序序列，所述概率模型利用不同疾病带有的不同时间戳的护理医嘱执行单数据训练生成且表征了各个标准医疗项目的排序顺序的概率值；及所述处理器执行所述计算机可读指令时所执行的所述对所述排序后的第一医疗项目进行审核，包括：对所述最佳排序序列进行审核。
  12. 根据权利要求11所述的计算机设备，其特征在于，所述处理器执行所述计算机可读指令时还执行以下步骤：当根据所述概率模型对所述待处理医疗项目以及根据所述固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序得到多个最佳排序序列时，则计算所得到的多个最佳排序序列对应的概率的差异值；当所述差异值小于预设 值时，则获取所得到的多个最佳排序序列中的、排序不同的医疗项目，以及未根据所述当前临床路径模型进行排序的医疗项目，作为当前医疗项目；及获取下一诊断结果作为当前诊断结果后，继续获取与所述当前诊断结果对应的当前临床路径模板，直至从所述单据信息中提取到诊断结果遍历完成。
  13. 根据权利要求12所述的计算机设备，其特征在于，所述处理器执行所述计算机可读指令时还执行以下步骤：当所述诊断结果集合中的诊断结果遍历完成后，仍存在未排序的当前医疗项目，则输出未排序的当前医疗项目；接收输入的针对所述未排序的当前医疗项目的修改指令；当所述修改指令表示所述未排序的当前医疗项目不合理时，则将所述未排序的当前医疗项目标记为不合理医疗项目；及当所述修改指令表示所述未排序的当前医疗项目合理时，则根据所述修改指令，将所述未排序的当前医疗项目插入至所生成的最佳排序序列中，并根据插入了所述未排序的当前医疗项目后的最佳排序序列更新所述概率模型。
  14. 根据权利要求10至13任意一项所述的计算机设备，其特征在于，所述处理器执行所述计算机可读指令时所执行的所述根据所述固定知识中的标准医疗项目之间的顺序对每一预设时间间隔内的所述第一医疗项目进行排序，包括：将所述第一医疗项目与所述固定知识中的预先排序完成的标准医疗项目进行匹配；及如果匹配成，则按照匹配成功的标准医疗项目对所述第一医疗项目进行排序。
  15. 根据权利要求12或13所述的计算机设备，其特征在于，所述处理器执行所述计算机可读指令时所执行的所述根据概率模型对所述待处理医疗项目以及根据所述固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序得到最佳排序序列，包括：将根据所述固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目生成第一序列；从所述待处理医疗项目中选取当前待处理医疗项目，并将所述当前待处理医疗项目插入所述第一序列的不同位置处形成第二序列，并计算不同的第二序列的概率；及获取概率最大的第二序列所对应的位置作为所述当前待处理医疗项目的插入位置，从所述待处理医疗项目中选取下一待处理医疗项目作为当前待处理医疗项目，并继续将所述当前待处理医疗项目插入所述第一序列的不同位置处形成第二序列，直至所述待处理医疗项目均插入至所述第一序列中得到最佳排序序列。
  16. 根据权利要求12或13所述的计算机设备，其特征在于，所述处理器执行所述计算机可读指令时所涉及的所述差异值是根据以下方式计算得到的：
    差异值＝(p1-p2)/((p1+p2)/2)
    其中，p1和p2为所述当前待处理医疗项目插入所述第一序列的不同位置处对应的概率。
  17. 一个或多个存储有计算机可读指令的非易失性计算机可读存储介质，所述计算机可读指令被一个或多个处理器执行时，使得所述一个或多个处理器执行以下步骤：从服务器获取到待审核医疗单据，并从所述待审核医疗单据中提取到诊断结果集合和当前医疗项 目；从所述诊断结果集合中获取当前诊断结果，并从服务器提取到预先存储的临床路径模板集合，将所述当前诊断结果与所述临床路径模板集合中的标准诊断结果进行比对得到当前临床路径模板，所述当前临床路径模板中按照时间顺序存储了各个标准医疗项目；根据所述当前临床路径模板中各个标准医疗项目的顺序，将所述当前医疗项目按照预设时间间隔进行排序得到第一医疗项目；获取预先存储的表征了所述标准医疗项目之间的顺序的固定知识，并根据所述固定知识中的标准医疗项目之间的顺序对每一预设时间间隔内的所述第一医疗项目进行排序；及对所述排序后的第一医疗项目进行审核根据所述固定知识中的标准医疗项目之间的顺序。
  18. 根据权利要求17所述的存储介质，其特征在于，所述计算机可读指令被所述处理器执行时所执行的所述根据所述固定知识中的标准医疗项目之间的顺序对每一预设时间间隔内的所述第一医疗项目进行排序之后，还包括：判断是否存在无法根据所述固定知识中的标准医疗项目之间的顺序进行排序的待处理医疗项目；当存在所述待处理医疗项目时，则根据概率模型，对所述待处理医疗项目以及根据所述固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序，得到最佳排序序列，所述概率模型利用不同疾病带有的不同时间戳的护理医嘱执行单数据训练生成且表征了各个标准医疗项目的排序顺序的概率值；及所述计算机可读指令被所述处理器执行时所执行的所述对所述排序后的第一医疗项目进行审核，包括：对所述最佳排序序列进行审核。
  19. 根据权利要求18所述的存储介质，其特征在于，所述计算机可读指令被所述处理器执行时还执行以下步骤：当根据所述概率模型对所述待处理医疗项目以及根据所述固定知识中的标准医疗项目之间的顺序排序后的第一医疗项目进行排序得到多个最佳排序序列时，则计算所得到的多个最佳排序序列对应的概率的差异值；当所述差异值小于预设值时，则获取所得到的多个最佳排序序列中的、排序不同的医疗项目，以及未根据所述当前临床路径模型进行排序的医疗项目，作为当前医疗项目；及获取下一诊断结果作为当前诊断结果后，继续获取与所述当前诊断结果对应的当前临床路径模板，直至从所述单据信息中提取到诊断结果遍历完成。
  20. 根据权利要求19所述的存储介质，其特征在于，所述计算机可读指令被所述处理器执行时还执行以下步骤：当所述诊断结果集合中的诊断结果遍历完成后，仍存在未排序的当前医疗项目，则输出未排序的当前医疗项目；接收输入的针对所述未排序的当前医疗项目的修改指令；当所述修改指令表示所述未排序的当前医疗项目不合理时，则将所述未排序的当前医疗项目标记为不合理医疗项目；及当所述修改指令表示所述未排序的当前医疗项目合理时，则根据所述修改指令，将所述未排序的当前医疗项目插入至所生成的最佳排序序列中，并根据插入了所述未排序的当前医疗项目后的最佳排序序列更新所述概率模型。
PCT/CN2019/099425 2018-08-14 2019-08-06 医疗单据审核方法、装置、计算机设备和存储介质 WO2020034874A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810924313.4A CN109102845B (zh) 2018-08-14 2018-08-14 医疗单据审核方法、装置、计算机设备和存储介质
CN201810924313.4 2018-08-14

Publications (1)

Publication Number Publication Date
WO2020034874A1 true WO2020034874A1 (zh) 2020-02-20

Family

ID=64849685

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/099425 WO2020034874A1 (zh) 2018-08-14 2019-08-06 医疗单据审核方法、装置、计算机设备和存储介质

Country Status (2)

Country Link
CN (1) CN109102845B (zh)
WO (1) WO2020034874A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112036750A (zh) * 2020-08-31 2020-12-04 平安医疗健康管理股份有限公司 适用于医疗风控的异常识别方法、装置、设备及存储介质

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109102845B (zh) * 2018-08-14 2023-07-04 深圳平安医疗健康科技服务有限公司 医疗单据审核方法、装置、计算机设备和存储介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150134349A1 (en) * 2013-11-13 2015-05-14 Koninklijke Philips N.V. System and method for quality review of healthcare reporting and feedback
CN107609980A (zh) * 2017-09-07 2018-01-19 平安医疗健康管理股份有限公司 医疗数据处理方法、装置、计算机设备及存储介质
CN107610779A (zh) * 2017-10-25 2018-01-19 医渡云（北京）技术有限公司 疾病评价及患病风险评估方法及装置
CN107729376A (zh) * 2017-09-13 2018-02-23 平安科技（深圳）有限公司 保险数据审核方法、装置、计算机设备及存储介质
CN108388580A (zh) * 2018-01-24 2018-08-10 平安医疗健康管理股份有限公司 融合医学知识及应用病例的动态知识图谱更新方法
CN109102845A (zh) * 2018-08-14 2018-12-28 平安医疗健康管理股份有限公司 医疗单据审核方法、装置、计算机设备和存储介质

Family Cites Families (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020049612A1 (en) * 2000-03-23 2002-04-25 Jaeger Scott H. Method and system for clinical knowledge management
JP4072103B2 (ja) * 2003-07-15 2008-04-09 株式会社日立製作所 診療報酬明細書の審査支援方法、診療報酬明細書の審査支援装置および診療報酬明細書の審査支援装置用プログラム
JP4814909B2 (ja) * 2008-06-04 2011-11-16 みずほ情報総研株式会社 レセプト審査支援システム、レセプト審査支援方法及びレセプト審査支援プログラム
US20100174555A1 (en) * 2009-01-05 2010-07-08 Klaus Abraham-Fuchs System for automatic clinical pathway optimization
JP5709456B2 (ja) * 2010-10-15 2015-04-30 キヤノン株式会社 処理装置、処理方法、医療支援システム及びプログラム
CN102567607B (zh) * 2010-12-28 2014-12-31 广州中医药大学第二附属医院 临床路径信息系统
US20150302153A1 (en) * 2014-04-22 2015-10-22 Cerner Innovation, Inc. Reverse document quality review
CN104134157B (zh) * 2014-08-08 2018-04-13 平安养老保险股份有限公司 一种医疗保险报销过程中可疑行为审核系统及审核方法
GB2542374A (en) * 2015-09-16 2017-03-22 Fujitsu Ltd Computer apparatus and method for enhancing clinical pathways
US20170262605A1 (en) * 2016-03-11 2017-09-14 Geneva Healthcare, Inc. Remote monitoring of medical devices
CN106469255B (zh) * 2016-08-31 2019-04-09 杭州逸曜信息技术有限公司 医疗信息的整合处理方法
WO2018057639A1 (en) * 2016-09-20 2018-03-29 Nuance Communications, Inc. Method and system for sequencing medical billing codes
CN107833621A (zh) * 2017-03-06 2018-03-23 平安医疗健康管理股份有限公司 医疗费用控制方法和装置
CN107451401A (zh) * 2017-07-11 2017-12-08 武汉金豆医疗数据科技有限公司 一种医保智能审核方法和系统
CN108198591A (zh) * 2017-12-28 2018-06-22 泰康保险集团股份有限公司 用于远程审核单据的方法与装置

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150134349A1 (en) * 2013-11-13 2015-05-14 Koninklijke Philips N.V. System and method for quality review of healthcare reporting and feedback
CN107609980A (zh) * 2017-09-07 2018-01-19 平安医疗健康管理股份有限公司 医疗数据处理方法、装置、计算机设备及存储介质
CN107729376A (zh) * 2017-09-13 2018-02-23 平安科技（深圳）有限公司 保险数据审核方法、装置、计算机设备及存储介质
CN107610779A (zh) * 2017-10-25 2018-01-19 医渡云（北京）技术有限公司 疾病评价及患病风险评估方法及装置
CN108388580A (zh) * 2018-01-24 2018-08-10 平安医疗健康管理股份有限公司 融合医学知识及应用病例的动态知识图谱更新方法
CN109102845A (zh) * 2018-08-14 2018-12-28 平安医疗健康管理股份有限公司 医疗单据审核方法、装置、计算机设备和存储介质

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112036750A (zh) * 2020-08-31 2020-12-04 平安医疗健康管理股份有限公司 适用于医疗风控的异常识别方法、装置、设备及存储介质

Also Published As

Publication number Publication date
CN109102845A (zh) 2018-12-28
CN109102845B (zh) 2023-07-04

Similar Documents

Publication Publication Date Title
US10541052B2 (en) Retroactive extraction of clinically relevant information from patient sequencing data for clinical decision support
US20190156947A1 (en) Automated information collection and evaluation of clinical data
US20160042134A1 (en) Method of calculating a score of a medical suggestion as a support in medical decision making
JP2018060529A (ja) コンテキストベースの患者類似性の方法及び装置
CN111382275A (zh) 医疗知识图谱的构建方法、装置、介质及电子设备
EP2862109A1 (en) System and method for generating textual report content using macros
KR20220038017A (ko) 임상 워크플로 결정을 자동화하고 우선 순위 판독 표시자를 생성하기 위한 시스템 및 방법
US11984227B2 (en) Automatically determining a medical recommendation for a patient based on multiple medical images from multiple different medical imaging modalities
WO2020034874A1 (zh) 医疗单据审核方法、装置、计算机设备和存储介质
CN112447270A (zh) 一种用药推荐方法、装置、设备及存储介质
US10282516B2 (en) Medical imaging reference retrieval
US20140046694A1 (en) Systems and methods for synoptic element structured reporting
Raghu et al. Validation of a deep learning–based model to predict lung cancer risk using chest radiographs and electronic medical record data
CN111061835A (zh) 查询方法及装置、电子设备和计算机可读存储介质
US20150066535A1 (en) System and method for reporting multiple medical procedures
US20140188511A1 (en) Systems and methods for stratification and management of medical conditions
US20200176127A1 (en) Systems and methods for guideline concordance
EP3659150B1 (en) Device, system, and method for optimizing image acquisition workflows
Andreychenko et al. A methodology for selection and quality control of the radiological computer vision deployment at the megalopolis scale
US20200043583A1 (en) System and method for workflow-sensitive structured finding object (sfo) recommendation for clinical care continuum
TW201835938A (zh) 資料整合方法
US10665343B1 (en) Medical treatment record integration
CN111128330A (zh) 电子病例报告表自动录入方法、装置以及相关设备
US10521552B2 (en) Method and computing device for implementing multiple matching strategies
US20240136068A1 (en) Method and device for providing medical prediction by using artificial intelligence model

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19849132

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19849132

Country of ref document: EP

Kind code of ref document: A1