Method and device for processing diagnosis and treatment event, storage medium and electronic device

Abstract

This disclosure relates to a method, apparatus, storage medium, and electronic device for processing medical events, belonging to the field of medical big data processing technology. The method includes: acquiring raw medical data corresponding to the current patient based on the current patient's patient identifier; performing structured extraction on the raw medical data to obtain current event entities included in the raw medical data; determining the current event axis of the current event entity based on its event category; mapping the current event entity onto the current event axis based on its current time node to obtain a target event axis; and displaying the target event axis so that medical personnel can analyze the current patient's treatment process based on the target event axis. This disclosure enables the visualization of medical events included in raw medical information.

Description

Technical Field

[0001] This disclosure relates to the field of medical big data processing technology, and more specifically, to a method for processing medical events, a device for processing medical events, a computer-readable storage medium, and an electronic device. Background Technology

[0002] The information recorded in existing hospital information systems is all raw medical information, which cannot be visualized.

[0003] It should be noted that the information in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] The purpose of this disclosure is to provide a method for processing medical events, a device for processing medical events, a computer-readable storage medium, and an electronic device, thereby overcoming, at least to some extent, the problem of the inability to visualize medical events due to limitations and defects in related technologies.

[0005] According to one aspect of this disclosure, a method for handling medical events is provided, comprising:

[0006] Based on the patient identifier of the current patient, obtain the original medical data corresponding to the current patient, and perform structured extraction on the original medical data to obtain the current event entity included in the original medical data;

[0007] Based on the event category of the current event entity, determine the current event axis of the current event entity, and based on the current time node of the current event entity, map the current event entity onto the current event axis to obtain the target event axis;

[0008] The target event axis is displayed so that medical staff can analyze the current patient's diagnosis and treatment process based on the target event axis.

[0009] In one exemplary embodiment of this disclosure, the original medical data includes one or more of the following: original medication prescription data, original medical order data, original medical history data, original medical imaging data, and original discharge record data;

[0010] Specifically, the original medical data is subjected to structured extraction to obtain the current event entities included in the original medical data, including:

[0011] Extract the first current event entity with a first entity category from the original medical order medication data and / or original medical order entrustment data and / or original current medical history data;

[0012] Extract a second current event entity with a second entity category from the original medical imaging data and / or original discharge record data;

[0013] The current event entity is generated based on the first current event entity and the second current event entity.

[0014] In one exemplary embodiment of this disclosure, extracting a first current event entity having a first entity category from the original medication order data and / or original medical order data and / or original current medical history data includes:

[0015] The original medication order data is filtered to obtain target medication order data. The original drug entities in the target medication order data are then screened using a preset medical knowledge dictionary and medical drug codes to obtain the first target drug entity.

[0016] The original medical order data is extracted based on a preset first entity extraction model to obtain a second target drug entity, and a first event set is constructed based on the first target drug entity and the second target drug entity.

[0017] Based on a preset second entity extraction model, the first treatment event entity is extracted from the original current medical history data, and the first current event entity is generated based on the first treatment event entity and the first event set.

[0018] In one exemplary embodiment of this disclosure, a first event set is constructed based on a first target drug entity and a second target drug entity, including:

[0019] Based on the first execution start time and the first execution end time of the first target drug entity, calculate the first time interval of the first target drug entity, and based on the second execution start time and the second execution end time of the second target drug entity, calculate the second time interval of the second target drug entity;

[0020] When it is determined that both the first time interval and the second time interval are less than a preset time threshold, drug entities with the same entity name in the first target drug entity and the second target drug entity are merged, and the first event set is constructed based on the merged first target drug entity and the second target drug entity.

[0021] In one exemplary embodiment of this disclosure, generating the first current event entity based on the first treatment event entity and the first event set includes:

[0022] Based on the event start time of the first treatment event entity and the execution start time of each target drug entity included in the first event set, the first treatment event entity and the target drug entity are merged.

[0023] Based on a pre-defined diagnostic and treatment knowledge base and a diagnostic and treatment plan dictionary, the event logic included in the first treatment event entity and the target drug entity after merging is extracted;

[0024] The event purpose of the extracted event logic is identified based on the preset event logic library, and the first current event entity is obtained based on the identification result.

[0025] In one exemplary embodiment of this disclosure, extracting a second current event entity having a second entity category from the original medical image data and / or original discharge record data includes:

[0026] Based on a preset third entity extraction model, the second current event entity with a second entity category included in the original medical image data and / or original discharge record data is extracted, and it is determined whether the extraction result is an empty set.

[0027] When the extraction result is determined to be an empty set, the treatment entities included in the original medical image data and / or original discharge record data are extracted, and the treatment entities are analyzed to obtain the treatment purpose and / or treatment effect corresponding to the treatment entity.

[0028] The second current event entity is generated based on the stated treatment objective and / or treatment effect.

[0029] In one exemplary embodiment of this disclosure, mapping the current event entity onto the current event axis based on the current time node of the current event entity to obtain the target event axis includes:

[0030] Based on the current time node of the current event entity, determine the current position of the current event entity on the current event axis, and perform abstract processing on the current event entity to obtain the event point corresponding to the current event entity;

[0031] Based on the current position, the event point is placed on the current event axis to obtain the target event axis;

[0032] The target event axis includes a normal event axis and / or a trend event axis. The current event entity corresponding to the event point on the normal event axis is displayed in the display area corresponding to the event point based on a preset display method.

[0033] According to one aspect of this disclosure, a device for processing medical events is provided, comprising:

[0034] The structured extraction module is used to obtain the original medical data corresponding to the current patient based on the current patient's patient identifier, and to perform structured extraction on the original medical data to obtain the current event entity included in the original medical data.

[0035] The event entity mapping module is used to determine the current event axis of the current event entity according to the event category of the current event entity, and to map the current event entity onto the current event axis according to the current time node of the current event entity to obtain the target event axis;

[0036] The target event axis display module is used to display the target event axis so that medical staff can analyze the current patient's diagnosis and treatment process based on the target event axis.

[0037] According to one aspect of this disclosure, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the method for processing diagnostic events as described in any of the preceding claims.

[0038] According to one aspect of this disclosure, an electronic device is provided, comprising:

[0039] Processor; and

[0040] Memory for storing the executable instructions of the processor;

[0041] The processor is configured to execute the method for processing any of the above-described diagnostic events by executing the executable instructions.

[0042] This disclosure provides a method for processing medical events. On one hand, it extracts structured data from raw medical data to obtain current event entities. Then, based on the event category of the current event entity, it determines the current event axis of the current event entity and maps the current event entity onto the current event axis based on the current time node of the current event entity to obtain a target event axis. Finally, it displays the target event axis so that medical staff can analyze the current patient's treatment process based on the target event axis, solving the problem in the prior art that it is impossible to visualize the medical events included in the raw medical information. On the other hand, since the current event entity can be mapped onto the current event axis based on the current time node of the current event entity to obtain the target event axis, medical staff can analyze the specific treatment process of the current patient based on the chronological order on the time axis and the current timestamp entity included at the corresponding time point when viewing the target medical event, thus improving the analysis efficiency of medical staff.

[0043] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0044] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure. It is obvious that the drawings described below are merely some embodiments of this disclosure, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0045] Figure 1 The flowchart illustrates a method for handling a medical event according to an example embodiment of the present disclosure.

[0046] Figure 2 The illustration schematically shows a flowchart of a method for extracting a first current event entity having a first entity category from the original medication order data and / or original medical order data and / or original medical history data, according to an example embodiment of the present disclosure.

[0047] Figure 3 The diagram illustrates a scenario example of a current event entity generation process according to an exemplary embodiment of the present disclosure.

[0048] Figure 4 The diagram schematically illustrates a structural example of a first entity extraction model based on BiLSTM-CRF according to an exemplary embodiment of the present disclosure.

[0049] Figure 5An example diagram illustrating a current event entity according to an exemplary embodiment of this disclosure is shown schematically.

[0050] Figure 6 An example diagram illustrating a time-referenced current event axis according to an exemplary embodiment of this disclosure is shown.

[0051] Figure 7 The illustration schematically shows an example diagram of a target event axis generated based on a common event axis according to an exemplary embodiment of the present disclosure.

[0052] Figure 8 The illustration schematically shows an example diagram of a target event axis generated based on a trend event axis according to an exemplary embodiment of the present disclosure.

[0053] Figure 9 The diagram schematically illustrates a block diagram of a medical event processing apparatus according to an exemplary embodiment of the present disclosure.

[0054] Figure 10 An electronic device for implementing the above-described method for processing medical events, according to an example embodiment of the present disclosure, is illustrated schematically. Detailed Implementation

[0055] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided to make this disclosure more comprehensive and complete, and to fully convey the concept of the example embodiments to those skilled in the art. The described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a full understanding of embodiments of this disclosure. However, those skilled in the art will recognize that the technical solutions of this disclosure can be practiced with one or more of the specific details omitted, or other methods, components, apparatus, steps, etc., can be employed. In other instances, well-known technical solutions are not shown or described in detail to avoid obscuring various aspects of this disclosure.

[0056] Furthermore, the accompanying drawings are merely illustrative of this disclosure and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0057] Electronic medical records (EMRs) primarily refer to the digitized medical service records used by medical personnel in clinical diagnosis, treatment, and intervention guidance, including text, symbols, icons, graphics, data, and images generated by the system. An EMR system is an integrated information system with functions for recording, storing, and accessing medical information, assisting doctors in clinical decision-making, and reusing information for public health and research. With the development and application of internet technology in the healthcare field, EMR systems have largely replaced paper medical records, becoming the foundation of modern hospital information infrastructure. EMR systems are not merely static medical records; they represent a deeper level of digitization. By recording relevant medical services, doctors can browse patient clinical information, examination and test results, nursing procedures, and medical images online for medical history review and clinical research.

[0058] Taking cancer patients as an example, cancer treatment requires a long-term process. During the diagnosis and treatment of cancer patients, doctors need to understand all the patient's medical events from the first visit to the present, which often span multiple visits, including outpatient and inpatient care. Currently, data storage in hospital information systems is relatively fragmented, generally using each visit as the basic data aggregation unit, requiring doctors to access data from multiple visits. Furthermore, patient medical information is stored in different information systems (e.g., medication information is stored in the HIS (Hospital Information System), test results in the LIS (Laboratory Information Management System), examination reports in the RIS (Radiology Information System), and electronic medical records in the EMR (Electronic Medical Record) system), requiring doctors to traverse multiple systems to access different dimensions of patient information.

[0059] However, the information recorded in hospital information systems is raw data, not the specific medical events related to cancer patients that doctors are directly interested in. For example, doctors are concerned with the TNM stage of a cancer patient, the overall treatment strategy, the chemotherapy regimen, the number of chemotherapy cycles performed, and the specific chemotherapy drugs used in each cycle; they also need to know when the patient experienced recurrence, metastasis, or death. These cancer-related events need to be extracted from the raw records using natural language processing and medical logic operations.

[0060] Of course, some IT vendors integrate data from multiple patient visits and various systems to create a comprehensive patient view. However, due to the difficulty in extracting and inferring important clinical events from raw records in these systems, these views typically only present the raw data without much processing. Doctors cannot directly obtain the information they need from such comprehensive patient views. Furthermore, integrating and displaying raw information across different patient visits and systems only solves part of the problem and does not yet constitute an efficient intelligent data product to assist doctors in their clinical work. Doctors still need to use their reasoning skills based on the raw information they see to obtain the information they truly want.

[0061] Based on this, this exemplary embodiment first provides a method for handling medical events, which can run on terminal devices, servers, server clusters, or cloud servers, etc. Of course, those skilled in the art can also run the method disclosed herein on other platforms as needed, and this exemplary embodiment does not impose any special limitations on this. Reference Figure 1 As shown, the method for handling this medical event may include the following steps:

[0062] Step S110. Based on the patient identifier of the current patient, obtain the original medical data corresponding to the current patient, and perform structured extraction on the original medical data to obtain the current event entity included in the original medical data;

[0063] Step S120. Determine the current event axis of the current event entity according to the event category of the current event entity, and map the current event entity onto the current event axis according to the current time node of the current event entity to obtain the target event axis;

[0064] Step S130. Display the target event axis so that medical staff can analyze the current patient's diagnosis and treatment process based on the target event axis.

[0065] In the above-mentioned method for processing medical events, on the one hand, the current event entities included in the original medical data are obtained by performing structured extraction on the original medical data; then, the current event axis of the current event entity is determined according to the event category of the current event entity, and the current event entity is mapped onto the current event axis according to the current time node of the current event entity to obtain the target event axis; finally, the target event axis is displayed so that medical staff can analyze the current patient's diagnosis and treatment process according to the target event axis, which solves the problem that the existing technology cannot visualize the diagnosis and treatment events included in the original medical information; on the other hand, since the current event entity can be mapped onto the current event axis according to the current time node of the current event entity to obtain the target event axis, medical staff can analyze the specific diagnosis and treatment process of the current patient based on the chronological order on the time axis and the current timestamp entity included at the corresponding time point when viewing the target diagnosis and treatment event, which improves the analysis efficiency of medical staff.

[0066] The following will provide a detailed explanation and description of the method for handling medical events according to the exemplary embodiments of this disclosure, in conjunction with the accompanying drawings.

[0067] First, the application scenarios and inventive objectives of the exemplary embodiments of this disclosure will be explained and described. Specifically, the method for processing diagnostic and therapeutic events provided in the exemplary embodiments of this disclosure can be used in oncology scenarios to visualize and define the multi-source data processing logic for important diagnostic and therapeutic events that occur during the treatment of cancer patients, such as surgery, chemotherapy, radiotherapy, targeted therapy, immunotherapy, pathological examination, abnormal tumor markers, recurrence, metastasis, and death. Simultaneously, the exemplary embodiments of this disclosure also provide a current event axis group with time reference as the core oncology specialty view, allowing doctors to fully understand the entire diagnostic and therapeutic process of cancer patients through a single page, and to more easily analyze the causal relationships between different events through the chronological order of events.

[0068] Secondly, in a method for handling medical events provided in the exemplary embodiments of this disclosure:

[0069] In step S110, based on the patient identifier of the current patient, the original medical data corresponding to the current patient is obtained, and the original medical data is extracted in a structured manner to obtain the current event entity included in the original medical data.

[0070] In this example embodiment, firstly, the original medical data of the current patient is obtained based on the patient identifier. During the acquisition of the original medical data, based on the given patient identifier (patient ID, such as an ID card number or phone number), all original medical records (original medical data) corresponding to the current patient can be extracted from the hospital's big data platform. These original medical records may include electronic medical records from previous visits, diagnostic information, treatment records such as surgeries and medications, examination and testing records, etc. That is, the original medical records may include original medication orders, original medical instructions, original medical history, original medical imaging data, and original discharge records; of course, other data may also be included, and this example does not impose any special limitations on this.

[0071] Secondly, after obtaining the original medical data, a structured extraction can be performed on the original medical data to obtain the current event entities included in the original medical data; wherein, the current event entities include a first current event entity with a first entity category, and / or a second current event entity with a second entity category. Specifically, the extraction of current event entities can be achieved in the following way: First, extract the first current event entity with the first entity category from the original medication order data and / or original medical order data and / or original medical history data; extract the second current event entity with the second entity category from the original medical imaging data and / or original discharge record data; generate the current event entity based on the first current event entity and the second current event entity. For example, core diagnostic and treatment events can be extracted from the original medical records, including the time and name of diagnosis, relevant symptoms at the onset of illness, disease course, and tumor stage, extracted using structured techniques; chemotherapy, targeted therapy, and endocrine therapy plans can be inferred from medication orders based on ATC codes; the purpose of chemotherapy can be inferred from the timeline of surgery and chemotherapy; and historical treatment plans can be extracted from the initial hospitalization history. This step relies on a disease knowledge graph, including the drugs of interest for different diseases, which drugs formed which treatment plans, and which examinations and tests are relevant to the disease, etc.

[0072] Among them, reference Figure 2 As shown, the extraction of the first current event entity can be achieved through the following steps:

[0073] Step S210: Filter the original medical order medication data to obtain target medical order medication data, and filter the original drug entities in the target medical order medication data using a preset medical knowledge dictionary and medical drug codes to obtain the first target drug entity.

[0074] Specifically, with Figure 3The extraction of chemotherapy event entities is illustrated using the example shown. A chemotherapy event entity can include core fields such as chemotherapy regimen name, start time, end time, chemotherapy purpose, and efficacy evaluation. (Reference) Figure 3 As shown, firstly, the original medication order data needs to be filtered to obtain the target medication order data (i.e., valid medication orders). Then, the original drug entities included in the target medication order data are screened in a medical knowledge dictionary (e.g., a chemotherapy drug dictionary) and a medical drug code (e.g., an ATC code) to obtain the first target drug entity (i.e., a valid chemotherapy record). During the filtering of the original medication order data, orders with abnormal status (such as voided, unpaid, etc.) can be filtered out. Secondly, based on the chemotherapy drug dictionary name and chemotherapy drug ATC code, it is determined which drugs are chemotherapy drugs, and the start and end times of these drug administrations are recorded. The extracted first target drug entity can be represented by events 1 and 2 as follows:

[0075] Event 1: Name: Paclitaxel; Start Time: 2020-03-10 08:00:00; End Time: 2020-03-15 14:00:00;

[0076] Event 2: Name: Cisplatin; Start Time: 2020-03-12 08:00:00; End Time: 2020-03-17 14:00:00.

[0077] Step S220: Extract the original medical order data based on the preset first entity extraction model to obtain the second target drug entity, and construct the first event set based on the first target drug entity and the second target drug entity.

[0078] In this example embodiment, firstly, the second target drug entity included in the original medical order data is extracted based on a preset first entity extraction model; wherein, the first entity extraction model can be an entity extraction model based on LSTM+CRF or BiLSTM-CRF, or it can be an entity extraction model based on CNN+CRF, and this example does not impose any special restrictions on it. Meanwhile, continue to refer to... Figure 3 As shown, during the extraction of the second target drug entity, some chemotherapy drugs in the original medical order data were brought by the patient and not prescribed during this visit. This medication information is usually issued in the form of a medical order, with the content in unstructured text, such as: "Oxaliplatin and leucovorin calcium infusion for no less than 2 hours". Therefore, the chemotherapy drugs can be extracted using Named Entity Recognition (NER) technology in Natural Language Processing (NLP), with the start and end times of the medical order execution serving as the start and end times of medication administration.

[0079] For example, taking the first entity extraction model as an example, which is an entity extraction model based on BiLSTM-CRF, the specific extraction process of the second target drug entity will be explained and illustrated. For details, please refer to... Figure 4 As shown, the entity extraction model based on BiLSTM-CRF can include a word embedding layer 401, a bidirectional long short-term memory network (BiLSTM) layer 402, and a conditional random field (CRF) layer 403; wherein the word embedding layer, BiLSTM layer, and CRF layer are connected sequentially; the embedding layer can be used to convert the original medical order data into word vectors to obtain the input embedding; then the embedding is input into the BiLSTM layer for feature extraction (encoding) to obtain the feature representation of the sequence, logits, and the logits are decoded to obtain the labeled sequence; finally, the labeled sequence is input into the decoding CRF layer to obtain the sequence of each word, and finally the second target drug entity is output through the output layer; wherein the obtained second target drug entity can be as shown in events 3 and 4 below:

[0080] Event 3: Name: Paclitaxel; Start Time: 2020-03-24 08:00:00; End Time: 2020-03-29 14:00:00;

[0081] Event 4: Name: Cisplatin, Start Time: 2020-03-26 08:00:00; End Time: 2020-03-31 14:00:00.

[0082] Furthermore, once the first target drug entity and the second target drug entity are obtained, a first event set can be constructed based on the first target drug entity and the second target drug entity. The specific construction process of the first event set can be implemented as follows: First, based on the first execution start time and the first execution end time of the first target drug entity, a first time interval of the first target drug entity is calculated; and based on the second execution start time and the second execution end time of the second target drug entity, a second time interval of the second target drug entity is calculated. Second, when it is determined that both the first time interval and the second time interval are less than a preset time threshold, drug entities with the same entity name in the first target drug entity and the second target drug entity are merged, and the first event set is constructed based on the merged first target drug entity and second target drug entity.

[0083] That is, when merging the first target drug entity and the second target drug entity, events with an interval less than a specific interval (e.g., 48 hours) can be merged to form a drug administration event set DES. For each cluster DES(i), chemotherapy drug administration events are merged, drug names are combined, and the chemotherapy regimen name is obtained by querying the chemotherapy regimen dictionary. The earliest drug administration start time in DES(i) is used as the regimen start time, and the latest drug administration end time is used as the regimen end time. The resulting first event set CES1 can be represented by events 5 and 6 as follows:

[0084] Event 5: Protocol Name: TP (Paclitaxel + Cisplatin); Start Time: 2020-03-10; End Time: 2020-03-17;

[0085] Event 6: Protocol Name: TP (Paclitaxel + Cisplatin); Start Time: 2020-03-24; End Time: 2020-03-31.

[0086] It should be further explained here that by merging the first target drug entity and the second target drug entity, the problem of too many event points on the target event axis due to too many event entities can be reduced, making it difficult for medical staff to view. At the same time, by merging the first target drug entity and the second target drug entity, medical staff can view more comprehensive medical information at the same event point, which facilitates their logical inference based on the medical information.

[0087] Step S230: Extract the first treatment event entity from the original current medical history data based on the preset second entity extraction model, and generate the first current event entity based on the first treatment event entity and the first event set.

[0088] In this example embodiment, firstly, a first treatment event entity is extracted from the original current medical history data based on a preset second entity extraction model. This second entity extraction model can be an entity extraction model implemented based on LSTM+CRF or BiLSTM-CRF, or it can be an entity extraction model implemented based on CNN+CRF; this example does not impose any special restrictions on this. The second entity extraction model can be the same as or different from the first entity extraction model; this example does not impose any special restrictions on this. Meanwhile, continue to refer to... Figure 3As shown, since many cancer patients receive treatment at other hospitals, this information is recorded in the original medical history data in natural language. A typical description is: "The patient was diagnosed with xxx cancer on xxx month xxx day, and started TP regimen 3 cycles on February 1, 2019, March 2, 2019, and April 1, 2019, etc." By extracting the first at least one event entity based on the second entity extraction model, chemotherapy event CES2 can be extracted. The extracted first treatment event entity can be represented by events 7, 8, and 9 as follows:

[0089] Event 7: Project Name: TP; Start Time: February 1, 2019; End Time: Unknown;

[0090] Event 8: Project Name: TP; Start Time: March 1, 2019; End Time: Unknown;

[0091] Event 9: Project Name: TP:; Start Time: April 1, 2019; End Time: Unknown.

[0092] Furthermore, once the first treatment event entity is obtained, a first current event entity can be generated based on the first treatment event entity and the first event set. The generation process of the first current event entity can be as follows: First, the first treatment event entity and the target drug entities are merged based on the event start time of the first treatment event entity and the execution start time of each target drug entity included in the first event set; second, the event logic included in the merged first treatment event entity and target drug entities is extracted based on a preset diagnostic knowledge base and diagnostic protocol dictionary; then, the event purpose of the extracted event logic is identified based on a preset event logic library, and the first current event entity is obtained based on the identification result. For details, please refer to [link / reference]. Figure 3 As shown, firstly, CES1 (the first event set) and CES2 (the first treatment event entity) are merged. The specific merging rule is: events with the same start time are merged (i.e., events with the same start time are merged). Figure 3 The records shown are merged), and then, based on the preset medical knowledge base and medical plan dictionary, the event logic included in the first treatment event entity and the target drug entity is extracted (that is, Figure 3 The process involves scheme identification (as shown in the diagram), followed by purpose identification, to obtain the first current event entity. It should be noted that by performing record merging, scheme identification, and purpose identification, the logical relationships between the various drug entities and treatment events included in the first current event entity become clearer, facilitating rapid review by medical staff.

[0093] Furthermore, after obtaining the first current event entity, a second current event entity with a second entity category can be extracted from the original medical image data and / or original discharge record data. Specifically, the extraction process of the second current event entity can be implemented as follows: First, based on a preset third entity extraction model, the second current event entities with a second entity category included in the original medical image data and / or original discharge record data are extracted, and it is determined whether the extraction result is an empty set; second, when it is determined that the extraction result is an empty set, the treatment entities included in the original medical image data and / or original discharge record data are extracted, and the treatment entities are analyzed to obtain the treatment purpose and / or treatment effect corresponding to the treatment entity; finally, the second current event entity is generated according to the treatment purpose and / or treatment effect.

[0094] Specifically, the aforementioned third entity extraction model can be an entity extraction model based on LSTM+CRF or BiLSTM-CRF, or it can be an entity extraction model based on CNN+CRF. This example does not impose any special restrictions on this. This third entity extraction model can be consistent with the first entity extraction model and / or the geothermal entity extraction model, or it can be different. This example does not impose any special restrictions on this. Meanwhile, please refer to... Figure 3 As shown, two additional fields, chemotherapy purpose and chemotherapy evaluation, need to be added to the chemotherapy event. Their calculation logic is as follows: Chemotherapy purpose: Generally, values ​​include neoadjuvant chemotherapy, concurrent chemotherapy, and adjuvant chemotherapy. During the extraction process, NLP extraction can be prioritized from the treatment plan description in the electronic medical record to determine the chemotherapy purpose. If no value is extracted (empty set), logical reasoning is used based on whether the patient underwent other radical treatments (such as surgical resection, radiotherapy, etc.) before and after this chemotherapy session to determine the treatment purpose of this chemotherapy. Chemotherapy evaluation: Generally, values ​​include CR (complete response), PR (partial response), NC (no change), and PD (progression). During the extraction process, NLP extraction is prioritized from the electronic medical record or imaging report to determine the chemotherapy efficacy. If no value is extracted (empty set), the tumor size is extracted from the imaging reports before and after chemotherapy based on the definition of chemotherapy evaluation (e.g., PR represents a lesion reduction of more than 50% but not complete disappearance), and numerical comparisons are performed to obtain the chemotherapy efficacy evaluation.

[0095] At this point, the first and second current event entities have been fully obtained, and a current event entity can be generated based on these entities. The specific details of the obtained current event entity can be found in [reference needed]. Figure 5 As shown.

[0096] In step S120, the current event axis of the current event entity is determined according to the event category of the current event entity, and the current event entity is mapped onto the current event axis according to the current time node of the current event entity to obtain the target event axis.

[0097] In this example embodiment, firstly, the current event axis of the current event entity is determined according to its event category. This event category can include ordinary events or trend events. An ordinary event is one where there is no continuous numerical change, while a trend event is one where different numerical points exist over a certain period, and the trend of the event can be determined by the changes in these points. Secondly, based on the current time node of the current event entity, it is mapped onto the current event axis to obtain the target event axis. The specific process for generating the target event axis is as follows: Firstly, based on the current time node of the current event entity, the current position of the current event entity on the current event axis is determined, and the current event entity is abstracted to obtain the event point corresponding to it. Secondly, based on the current position, the event point is placed on the current event axis to obtain the target event axis. The target event axis includes ordinary event axes and / or trend event axes. The current event entity corresponding to the event point on the ordinary event axis is displayed in the display area corresponding to the event point based on a preset display method.

[0098] For details, please refer to Figure 6 As shown, a time-referenced event axis group-based oncology specialty view is proposed, consisting of multiple event axes. Each event axis corresponds to a type of diagnostic and treatment event (such as chemotherapy). The event axes are divided into ordinary event axes and trend axes. Each point on the axis represents an event, and the occurrence time of the event corresponds to the corresponding time on the time reference axis. The ordinary event axis is suitable for non-numerical events and can define four behaviors: displaying content above the event point when at rest, displaying content below the event point when at rest, displaying content above the event point when the mouse hovers over it, and displaying content below the event point when the mouse hovers over it. These four behaviors define the default and hover information displayed for different events. The trend axis is suitable for numerical events, such as tumor marker detection values. Each point on the axis represents a tumor marker detection result, and the height of the point on the axis represents the magnitude of the value. Thus, the trend axis represents the changing trend of a certain indicator.

[0099] Furthermore, by comparing multiple points on a single axis horizontally, changes in a certain type of event can be analyzed as the disease and treatment progress. By comparing points on multiple axes vertically, the correlations and even causal relationships between different events during the patient's diagnosis and treatment process can be quickly compared. For example, changes in tumor markers, tumor size, etc., with treatment.

[0100] In step S130, the target event axis is displayed so that medical staff can analyze the current patient's diagnosis and treatment process based on the target event axis.

[0101] Specifically, the obtained target event axis can be referenced. Figure 7 as well as Figure 8 As shown. It can be understood that the method for processing medical events provided in this exemplary embodiment can, based on original medical records, perform multi-source natural language extraction and medical logic reasoning with the assistance of a knowledge base to form a technology for generating various medical events of greatest concern to cancer patients; simultaneously, it visualizes and displays cancer patient medical events in a tumor specialty view centered on an event axis group with time reference.

[0102] This disclosure also provides a device for handling medical events. (See reference) Figure 9 As shown, the device for processing medical events may include a structured extraction module 910, an event entity mapping module 920, and a target event axis display module 930. Wherein:

[0103] The structured extraction module 910 can be used to obtain the original medical data corresponding to the current patient based on the current patient's patient identifier, and perform structured extraction on the original medical data to obtain the current event entity included in the original medical data.

[0104] The event entity mapping module 920 can be used to determine the current event axis of the current event entity according to the event category of the current event entity, and map the current event entity onto the current event axis according to the current time node of the current event entity to obtain the target event axis;

[0105] The target event axis display module 930 can be used to display the target event axis so that medical staff can analyze the current patient's diagnosis and treatment process based on the target event axis.

[0106] In one exemplary embodiment of this disclosure, the original medical data includes one or more of the following: original medication prescription data, original medical order data, original medical history data, original medical imaging data, and original discharge record data;

[0107] Specifically, the original medical data is subjected to structured extraction to obtain the current event entities included in the original medical data, including:

[0108] Extract the first current event entity with a first entity category from the original medical order medication data and / or original medical order entrustment data and / or original current medical history data;

[0109] Extract a second current event entity with a second entity category from the original medical imaging data and / or original discharge record data;

[0110] The current event entity is generated based on the first current event entity and the second current event entity.

[0111] In one exemplary embodiment of this disclosure, extracting a first current event entity having a first entity category from the original medication order data and / or original medical order data and / or original current medical history data includes:

[0112] The original medication order data is filtered to obtain target medication order data. The original drug entities in the target medication order data are then screened using a preset medical knowledge dictionary and medical drug codes to obtain the first target drug entity.

[0113] The original medical order data is extracted based on a preset first entity extraction model to obtain a second target drug entity, and a first event set is constructed based on the first target drug entity and the second target drug entity.

[0114] Based on a preset second entity extraction model, the first treatment event entity is extracted from the original current medical history data, and the first current event entity is generated based on the first treatment event entity and the first event set.

[0115] In one exemplary embodiment of this disclosure, a first event set is constructed based on a first target drug entity and a second target drug entity, including:

[0116] Based on the first execution start time and the first execution end time of the first target drug entity, calculate the first time interval of the first target drug entity, and based on the second execution start time and the second execution end time of the second target drug entity, calculate the second time interval of the second target drug entity;

[0117] When it is determined that both the first time interval and the second time interval are less than a preset time threshold, drug entities with the same entity name in the first target drug entity and the second target drug entity are merged, and the first event set is constructed based on the merged first target drug entity and the second target drug entity.

[0118] In one exemplary embodiment of this disclosure, generating the first current event entity based on the first treatment event entity and the first event set includes:

[0119] Based on the event start time of the first treatment event entity and the execution start time of each target drug entity included in the first event set, the first treatment event entity and the target drug entity are merged.

[0120] Based on a pre-defined diagnostic and treatment knowledge base and a diagnostic and treatment plan dictionary, the event logic included in the first treatment event entity and the target drug entity after merging is extracted;

[0121] The event purpose of the extracted event logic is identified based on the preset event logic library, and the first current event entity is obtained based on the identification result.

[0122] In one exemplary embodiment of this disclosure, extracting a second current event entity having a second entity category from the original medical image data and / or original discharge record data includes:

[0123] Based on a preset third entity extraction model, the second current event entity with a second entity category included in the original medical image data and / or original discharge record data is extracted, and it is determined whether the extraction result is an empty set.

[0124] When the extraction result is determined to be an empty set, the treatment entities included in the original medical image data and / or original discharge record data are extracted, and the treatment entities are analyzed to obtain the treatment purpose and / or treatment effect corresponding to the treatment entity.

[0125] The second current event entity is generated based on the stated treatment objective and / or treatment effect.

[0126] In one exemplary embodiment of this disclosure, mapping the current event entity onto the current event axis based on the current time node of the current event entity to obtain the target event axis includes:

[0127] Based on the current time node of the current event entity, determine the current position of the current event entity on the current event axis, and perform abstract processing on the current event entity to obtain the event point corresponding to the current event entity;

[0128] Based on the current position, the event point is placed on the current event axis to obtain the target event axis;

[0129] The target event axis includes a normal event axis and / or a trend event axis. The current event entity corresponding to the event point on the normal event axis is displayed in the display area corresponding to the event point based on a preset display method.

[0130] The specific details of each module in the above-mentioned medical event processing device have been described in detail in the corresponding medical event processing methods, so they will not be repeated here.

[0131] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0132] Furthermore, although the steps of the method in this disclosure are described in a specific order in the accompanying drawings, this does not require or imply that the steps must be performed in that specific order, or that all the steps shown must be performed to achieve the desired result. Additional or alternative steps may be omitted, multiple steps may be combined into one step, and / or a step may be broken down into multiple steps.

[0133] In an exemplary embodiment of this disclosure, an electronic device capable of implementing the above-described method is also provided.

[0134] Those skilled in the art will understand that various aspects of this disclosure can be implemented as a system, method, or program product. Therefore, various aspects of this disclosure can be specifically implemented in the following forms: a completely hardware implementation, a completely software implementation (including firmware, microcode, etc.), or a combination of hardware and software aspects, collectively referred to herein as a "circuit," "module," or "system."

[0135] The following reference Figure 10 To describe an electronic device 1000 according to such an embodiment of the present disclosure. Figure 10 The electronic device 1000 shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments disclosed herein.

[0136] like Figure 10 As shown, the electronic device 1000 is manifested in the form of a general-purpose computing device. The components of the electronic device 1000 may include, but are not limited to: at least one processing unit 1010, at least one storage unit 1020, a bus 1030 connecting different system components (including storage unit 1020 and processing unit 1010), and a display unit 1040.

[0137] The storage unit stores program code that can be executed by the processing unit 1010, causing the processing unit 1010 to perform the steps described in the "Exemplary Methods" section of this specification according to various exemplary embodiments of this disclosure. For example, the processing unit 1010 can perform actions such as... Figure 1 Step S110: Based on the patient identifier of the current patient, obtain the original medical data corresponding to the current patient, and perform structured extraction on the original medical data to obtain the current event entity included in the original medical data; Step S120: Based on the event category of the current event entity, determine the current event axis of the current event entity, and based on the current time node of the current event entity, map the current event entity onto the current event axis to obtain the target event axis; Step S130: Display the target event axis so that medical staff can analyze the diagnosis and treatment process of the current patient based on the target event axis.

[0138] Storage unit 1020 may include readable media in the form of volatile storage units, such as random access memory (RAM) 10201 and / or cache memory 10202, and may further include read-only memory (ROM) 10203.

[0139] Storage unit 1020 may also include a program / utility 10204 having a set (at least one) program module 10205, such program module 10205 including but not limited to: operating system, one or more application programs, other program modules and program data, each or some combination of these examples may include an implementation of a network environment.

[0140] Bus 1030 can represent one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of the multiple bus structures.

[0141] Electronic device 1000 can also communicate with one or more external devices 1100 (e.g., keyboard, pointing device, Bluetooth device, etc.), one or more devices that enable a user to interact with electronic device 1000, and / or any device that enables electronic device 1000 to communicate with one or more other computing devices (e.g., router, modem, etc.). This communication can be performed via input / output (I / O) interface 1050. Furthermore, electronic device 1000 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 1060. As shown, network adapter 1060 communicates with other modules of electronic device 1000 via bus 1030. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with electronic device 1000, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0142] From the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions according to the embodiments of this disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, terminal device, or network device, etc.) to execute the methods according to the embodiments of this disclosure.

[0143] In exemplary embodiments of this disclosure, a computer-readable storage medium is also provided, on which a program product capable of implementing the methods described above is stored. In some possible implementations, various aspects of this disclosure may also be implemented as a program product including program code that, when the program product is run on a terminal device, causes the terminal device to perform the steps of the various exemplary embodiments of this disclosure described in the "Exemplary Methods" section above.

[0144] The program product for implementing the above-described method according to embodiments of the present disclosure may employ a portable compact disc read-only memory (CD-ROM) and include program code, and may run on a terminal device, such as a personal computer. However, the program product of the present disclosure is not limited thereto. In this document, the readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system, apparatus, or device.

[0145] The program product may employ any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0146] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable signal medium may also be any readable medium other than a readable storage medium, capable of sending, propagating, or transmitting programs for use by or in conjunction with an instruction execution system, apparatus, or device.

[0147] The program code contained on the readable medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof.

[0148] Program code for performing the operations of this disclosure can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can execute entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0149] Furthermore, the above figures are merely illustrative of the processes included in the method according to exemplary embodiments of this disclosure and are not intended to be limiting. It is readily understood that the processes shown in the above figures do not indicate or limit the temporal order of these processes. Additionally, it is readily understood that these processes may be executed synchronously or asynchronously, for example, in multiple modules.

[0150] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention described herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not invented by this disclosure. The specification and embodiments are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the claims.

Claims

1. A method for handling medical events, characterized in that, include: Based on the current patient's patient identifier, the original medical data corresponding to the current patient is obtained, and the original medical data is extracted in a structured manner to obtain the current event entity included in the original medical data; wherein, the original medical data includes one or more of the following: original medication order data, original medical order instructions data, original medical history data, original medical imaging data, and original discharge record data; Based on the event category of the current event entity, determine the current event axis of the current event entity, and based on the current time node of the current event entity, map the current event entity onto the current event axis to obtain the target event axis; The target event axis is displayed so that medical staff can analyze the current patient's diagnosis and treatment process based on the target event axis; Specifically, the original medical data is subjected to structured extraction to obtain the current event entities included in the original medical data, including: Extracting a first current event entity with a first entity category from the original medication order data and / or original medical order data and / or original medical history data; wherein, determining the first current event entity includes: filtering the original medication order data to obtain target medication order data, and screening the target medication order data for original drug entities in a preset medical knowledge dictionary and medical drug codes to obtain a first target drug entity; extracting the original medical order data based on a preset first entity extraction model to obtain a second target drug entity, and constructing a first event set based on the first target drug entity and the second target drug entity; extracting a first treatment event entity from the original medical history data based on a preset second entity extraction model, and generating the first current event entity based on the first treatment event entity and the first event set; Extracting a second current event entity with a second entity category from the original medical image data and / or original discharge record data; wherein, determining the second current event entity includes: extracting second current event entities with a second entity category from the original medical image data and / or original discharge record data based on a preset third entity extraction model, and determining whether the extraction result is an empty set; when the extraction result is determined to be an empty set, extracting treatment entities from the original medical image data and / or original discharge record data, and analyzing the treatment entities to obtain the treatment purpose and / or treatment effect corresponding to the treatment entity; generating the second current event entity based on the treatment purpose and / or treatment effect; The current event entity is generated based on the first current event entity and the second current event entity; In cases where no second current event entity is extracted from the imaging report, the tumor size is extracted from the imaging reports before and after chemotherapy according to the definition of chemotherapy evaluation, and numerical comparison is performed to obtain the efficacy evaluation of chemotherapy.

2. The method for handling medical events according to claim 1, characterized in that, Construct a first event set based on the first target drug entity and the second target drug entity, including: Based on the first execution start time and the first execution end time of the first target drug entity, calculate the first time interval of the first target drug entity, and based on the second execution start time and the second execution end time of the second target drug entity, calculate the second time interval of the second target drug entity; When it is determined that both the first time interval and the second time interval are less than a preset time threshold, drug entities with the same entity name in the first target drug entity and the second target drug entity are merged, and the first event set is constructed based on the merged first target drug entity and the second target drug entity.

3. The method for handling medical events according to claim 2, characterized in that, Based on the first treatment event entity and the first event set, the first current event entity is generated, including: Based on the event start time of the first treatment event entity and the execution start time of each target drug entity included in the first event set, the first treatment event entity and the target drug entity are merged. Based on a pre-defined diagnostic and treatment knowledge base and a diagnostic and treatment plan dictionary, the event logic included in the first treatment event entity and the target drug entity after merging is extracted; The event purpose of the extracted event logic is identified based on the preset event logic library, and the first current event entity is obtained based on the identification result.

4. The method for handling medical events according to claim 1, characterized in that, Based on the current time node of the current event entity, the current event entity is mapped onto the current event axis to obtain the target event axis, including: Based on the current time node of the current event entity, determine the current position of the current event entity on the current event axis, and perform abstract processing on the current event entity to obtain the event point corresponding to the current event entity; Based on the current position, the event point is placed on the current event axis to obtain the target event axis; The target event axis includes a normal event axis and / or a trend event axis. The current event entity corresponding to the event point on the normal event axis is displayed in the display area corresponding to the event point based on a preset display method.

5. A device for processing medical events, characterized in that, include: The structured extraction module is used to obtain the original medical data corresponding to the current patient based on the current patient's patient identifier, and to perform structured extraction on the original medical data to obtain the current event entity included in the original medical data; wherein, the original medical data includes one or more of the following: original medication order data, original medical order instructions data, original medical history data, original medical imaging data, and original discharge record data; The event entity mapping module is used to determine the current event axis of the current event entity according to the event category of the current event entity, and to map the current event entity onto the current event axis according to the current time node of the current event entity to obtain the target event axis; The target event axis display module is used to display the target event axis so that medical staff can analyze the current patient's diagnosis and treatment process based on the target event axis; Specifically, the original medical data is subjected to structured extraction to obtain the current event entities included in the original medical data, including: Extracting a first current event entity with a first entity category from the original medication order data and / or original medical order data and / or original medical history data; wherein, determining the first current event entity includes: filtering the original medication order data to obtain target medication order data, and screening the target medication order data for original drug entities in a preset medical knowledge dictionary and medical drug codes to obtain a first target drug entity; extracting the original medical order data based on a preset first entity extraction model to obtain a second target drug entity, and constructing a first event set based on the first target drug entity and the second target drug entity; extracting a first treatment event entity from the original medical history data based on a preset second entity extraction model, and generating the first current event entity based on the first treatment event entity and the first event set; Extracting a second current event entity with a second entity category from the original medical image data and / or original discharge record data; wherein, determining the second current event entity includes: extracting second current event entities with a second entity category from the original medical image data and / or original discharge record data based on a preset third entity extraction model, and determining whether the extraction result is an empty set; when the extraction result is determined to be an empty set, extracting treatment entities from the original medical image data and / or original discharge record data, and analyzing the treatment entities to obtain the treatment purpose and / or treatment effect corresponding to the treatment entity; generating the second current event entity based on the treatment purpose and / or treatment effect; The current event entity is generated based on the first current event entity and the second current event entity; In cases where no second current event entity is extracted from the imaging report, the tumor size is extracted from the imaging reports before and after chemotherapy according to the definition of chemotherapy evaluation, and numerical comparison is performed to obtain the efficacy evaluation of chemotherapy.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for processing diagnostic events as described in any one of claims 1-4.

7. An electronic device, characterized in that, include: processor; as well as Memory for storing the executable instructions of the processor; The processor is configured to execute the method for processing diagnostic events according to any one of claims 1-4 by executing the executable instructions.

Images