Chain collaborative processing method and system for health examination data

Abstract

The application discloses a chain cooperative processing method and system for health examination data. The method acquires an abnormal examination item set, a full-amount examination result dictionary, a processing module priority queue and a condition configuration set; filters the condition configuration according to an agency authority and forms a module execution sequence; extracts a current factor from the abnormal examination item set by the processing module, and supplements a missing examination item by the full-amount examination result dictionary; generates a module processing record according to a two-dimensional condition group, a normal range reference condition and a minimum matching factor number, removes a participating item and updates a consumption state when an output condition is met; performs ordering and factor containing deduplication, and outputs structured examination information. The variable abnormal examination item set is separated from the full-amount examination result dictionary, the exclusive consumption of the examination item and the original value supplement are considered, and the configuration processing of the condition and the agency authority is supported.

Description

Technical Field

[0001] This invention relates to the field of computer data processing and health examination report information generation technology, and in particular to a chain-based collaborative processing method and system for performing exclusive consumption, configuration condition matching, permission filtering and structured output of shared examination data. Background Technology

[0002] Health examination report data processing systems typically receive numerical, enumerated, or text-based examination results from multiple sources and invoke multiple processing modules to convert these results into structured information for a report generation interface. Because the processing scope of different modules may overlap, the same examination item may simultaneously meet the input conditions of multiple processing modules.

[0003] Existing implementations often involve each processing module independently reading the same set of anomaly checks and performing threshold comparisons or keyword matching separately. While some systems call processing modules in a fixed order, they fail to record the status of checks used by preceding processing modules; other systems directly delete used data, preventing subsequent processing modules from obtaining the original values ​​used for relation calculations or level classification.

[0004] Furthermore, existing condition configurations typically employ a single-level condition list. A single-level list struggles to simultaneously express nested relationships such as "all conditions within a group are met" and "at least one condition from multiple candidate groups is met," and it also makes it difficult to treat check results within the reference range as configurable conditions. If configuration files are copied separately for different target organizations, the applicable scope and processing level of the organization will be coupled with the condition matching logic, leading to an explosion in the number of rules and version inconsistencies.

[0005] Therefore, the existing technology has at least the following problems: First, it lacks a data organization mechanism that takes into account both exclusive consumption and full readback on the shared anomaly check item set, which can easily lead to duplicate structured information or damage the data dependencies of subsequent modules; Second, it is difficult to complete the parsing of complex condition combinations, normal range references and institutional permissions through a unified configuration structure, resulting in solidified processing logic and high maintenance costs. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to provide a chain-based collaborative processing method and system for health examination data, so as to realize orderly and traceable consumption of examination items on shared examination data, and to support nested condition combinations, normal range references and institutional permission control with a unified condition configuration structure.

[0007] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0008] This invention provides a chain-based collaborative processing method for health examination data, comprising:

[0009] Get the set of anomaly check items, the dictionary of full check results, the priority queue of processing modules, and the set of condition configurations;

[0010] Based on the target organization identifier, organization processing level, applicable object rules, and information processing level, available conditions are filtered and configured, and a module execution sequence is formed according to the priority queue of the processing modules.

[0011] The set of anomaly check items is used as the current set of anomaly check items. The processing module is called according to the module execution sequence. The current factor set is extracted from the current set of anomaly check items, and the missing check items are read from the full check result dictionary to obtain the supplementary factor set.

[0012] Based on the multiple matching conditions included in the available conditions configuration, a combined matching is performed on the current factor set and the supplementary factor set. When the multiple matching conditions are met, it is determined that the available conditions configuration is triggered, and the module processing record and the set of participating item indexes are obtained.

[0013] When the module processes a record that meets the output conditions, it removes consumed check items according to the participating item index set. The consumed check items are the abnormal check items that actually participated in the combined matching and were removed. The module also updates the consumption status of the cross-module auxiliary attributes in the full check result dictionary, changing the consumption status from unconsumed to consumed.

[0014] The module processes the records by sorting and deduplication, and factor inclusion deduplication, and outputs structured inspection information.

[0015] Furthermore, the acquisition of the anomaly check item set, the full check result dictionary, the processing module priority queue, and the condition configuration set includes:

[0016] The original inspection results were normalized in terms of name and unit according to the standard identification of the inspection items;

[0017] The inspection results that exceed the corresponding reference range are encapsulated as anomaly check items and written into the anomaly check item set in the original input order;

[0018] All inspection results that have completed name normalization and unit normalization are written into the full inspection result dictionary with the inspection item standard identifier as the key.

[0019] For each anomaly check item, save the set index, check item standard identifier, check result value, reference range, and anomaly direction;

[0020] Processing modules whose processing scopes overlap are written into the processing module priority queue according to the preset business priority;

[0021] Write the condition identifier, factor field, two-dimensional condition group, minimum number of matching factors, applicable object rules, and information processing level into the condition configuration set.

[0022] Further, when the module processes a record that meets the output condition, it removes consumed check items according to the participating item index set. These consumed check items are the abnormal check items that actually participated in the combined matching and were removed. The module also updates the cross-module auxiliary attribute consumption status in the full check result dictionary, changing the consumption status from unconsumed to consumed. This includes:

[0023] Before invoking the current processing module, record the data version and number of elements of the current anomaly check item set;

[0024] Write the index of the anomaly check item that matches the current processing module and actually participates in the combination matching into the participation item index set;

[0025] The participating item index set is deduplicated and sorted in descending order to obtain a descending index sequence;

[0026] Verify that the data version of the current anomaly check item set is consistent with the data version recorded before the current processing module was called;

[0027] Consumed check items are removed one by one according to the descending index sequence to keep the set index of unconsumed check items valid;

[0028] Generate consumption records for removed consumed items, including the consumption module identifier, consumption order, and consumption time.

[0029] When the current processing module uses cross-module auxiliary attributes, the consumption status of the corresponding cross-module auxiliary attributes will be updated from unconsumed to consumed.

[0030] The set of exception checks after the removal operation is determined as the current set of exception checks for the next processing module.

[0031] Further, based on the multiple matching conditions included in the available condition configuration, a combined matching is performed on the current factor set and the supplementary factor set. When the multiple matching conditions are met, the available condition configuration is determined to be triggered, resulting in a module processing record and a set of participating item indexes, including:

[0032] The two-dimensional condition group configured with the available conditions is parsed into multiple outer candidate groups, and each outer candidate group is parsed into multiple inner conditions, wherein the inner conditions are the matching conditions;

[0033] The inner condition is matched with the current factor set and the supplementary factor set respectively to obtain the inner condition matching result for each inner condition;

[0034] When all the inner conditions in the same outer candidate group are satisfied, the corresponding outer candidate group is marked as satisfied, and the outer candidate group marking result is obtained.

[0035] When the outer candidate group marking result of at least one outer candidate group is satisfied, the two-dimensional condition group is marked as satisfied, and the available condition configuration is determined to be triggered;

[0036] The participation item index set is formed by summing the indexes of the outlier check items that are marked as satisfying the outer candidate group and originate from the current factor set.

[0037] The module processing record is generated based on the labeling results of the two-dimensional condition group.

[0038] Further, the step of matching the inner condition with the current factor set and the supplementary factor set respectively includes:

[0039] Identify the variable reference markers in the inner conditions, and extract the referenced check item standard identifiers from the variable reference markers;

[0040] The inspection result value, unit of measurement, and reference range are read from the full inspection result dictionary using the referenced inspection item standard identifier as the key.

[0041] The inspection result value is converted to a standardized inspection result value by converting the standard unit of measurement corresponding to the reference range.

[0042] The upper limit, lower limit, and boundary inclusion identifier of the reference range are analyzed.

[0043] Based on the boundary inclusion identifier, the standardized check result value is compared with the upper limit value and the lower limit value respectively to obtain the Boolean matching value of the normal range reference condition;

[0044] For inner conditions that do not contain the variable reference marker, the corresponding condition identifier is compared with the abnormal direction identifier in the current factor set to obtain the Boolean matching value of the abnormal condition;

[0045] Write the Boolean match value of the normal range reference condition and the Boolean match value of the abnormal condition into the condition state vector of the corresponding outer candidate group.

[0046] Further, the step of performing combined matching on the current factor set and the supplementary factor set based on the multiple matching conditions included in the available condition configuration, determining that the available condition configuration is triggered when the multiple matching conditions are met, and obtaining the module processing record and the participating item index set, also includes:

[0047] The standard identifiers of the inspection items in the current factor set are deduplicated to obtain the number of different factors;

[0048] The number of different factors is compared with the minimum number of matching factors configured by the available conditions. When the number of different factors reaches the minimum number of matching factors, the factor number matching status is marked as satisfied, and the available conditions configuration is marked as triggered.

[0049] When the factor quantity matching status is not satisfied, the two-dimensional condition group of the available condition configuration is parsed into multiple outer candidate groups, and each outer candidate group is parsed into multiple inner conditions. The inner conditions are matched with the current factor set and the supplementary factor set respectively. The normal range reference condition is used as the inner condition for matching. When all the inner conditions in the same outer candidate group are satisfied, the corresponding outer candidate group is marked as satisfied. When at least one outer candidate group is marked as satisfied, the two-dimensional condition group is marked as satisfied, and the available condition configuration is marked as triggered.

[0050] When the factor quantity matching status is marked as satisfied or the two-dimensional condition group is marked as satisfied, the available condition configuration is determined to be triggered, and the condition identifier, factor source and processing sorting level are recorded to form the module processing record;

[0051] When the factor quantity matching state is marked as satisfied, only the check item indexes that originate from the current factor set and participate in the counting of different factor quantities are written into the participation item index set; when the two-dimensional condition group is marked as satisfied, only the check item indexes that originate from the current factor set and actually participate in the matching of the two-dimensional condition group are written into the participation item index set.

[0052] Furthermore, the configuration of filtering available conditions based on target organization identifier, organization processing level, applicable object rules, and information processing level, and forming a module execution sequence according to the priority queue of the processing modules, includes:

[0053] The applicable rules are parsed into a universally applicable mode, an exclusion list mode, or a specified list mode.

[0054] In the all-applicable mode, the object matching status corresponding to the target organization identifier is set to allowed, and a first sorting weight is assigned;

[0055] In the exclusion list mode, when the target organization identifier does not belong to the exclusion identifier set, the object matching status is set to allowed;

[0056] In the specified list mode, when the target organization identifier belongs to the specified identifier set, the object matching status is set to allowed, and a second ranking weight greater than the first ranking weight is assigned.

[0057] Input the organization processing level and the information processing level into a preset level permission matrix to obtain the level matching status;

[0058] The conditional configuration in which both the object matching status and the level matching status are allowed is determined as the available conditional configuration;

[0059] Matching conflicts between multiple applicable object rules are resolved in the order that the second sorting weight takes precedence over the first sorting weight.

[0060] Further, the step of using the set of anomaly checks as the current set of anomaly checks, calling the processing module according to the module execution sequence, extracting the current factor set from the current set of anomaly checks, and reading missing checks from the full check result dictionary to obtain a supplementary factor set includes:

[0061] In the first processing round, the dedicated processing module and the configuration processing module are invoked according to the module execution sequence.

[0062] The anomaly check items matched from the current anomaly check item set are marked as primary triggering factors and written into the current factor set;

[0063] When the required check item for the available condition configuration does not exist in the current factor set, the corresponding check item is read from the full check result dictionary and marked as a reference factor to form the supplementary factor set;

[0064] Prohibit check item consumption operations from being triggered solely by a set of supplementary factors marked as reference factors;

[0065] After the first processing round ends, a second processing round is executed for the available condition configurations that have not generated module processing records;

[0066] In the second processing round, the two-dimensional condition group containing the normal range reference conditions is parsed based on the dictionary of the full check results, and the newly generated module processing record is added to the candidate record set.

[0067] Furthermore, the process of sorting and deduplicating records and factoring for deduplication in the module, and outputting structured inspection information, includes:

[0068] Sort the processing records of the multiple modules according to the result identifier and processing sorting level;

[0069] For module processing records with the same result identifier, retain the module processing records whose processing sorting level meets the preset priority conditions;

[0070] Construct a corresponding set of factor identifiers from the standard identifiers of the inspection items referenced in each retained module processing record;

[0071] When the first factor identifier set is included in the second factor identifier set and the corresponding module processing record satisfies the preset substitution relationship, the module processing record corresponding to the first factor identifier set is removed.

[0072] The result identifiers, evidence item identifiers, matching paths, consumption records, and applicable permissions from the remaining module processing records are encapsulated into structured data objects;

[0073] The structured data objects are sorted according to the priority queue of the processing module to obtain the structured inspection information;

[0074] The structured examination information is sent to the health examination report generation interface, which then generates editable report prompts.

[0075] This invention also provides a chain-based collaborative processing system for health checkup data, comprising:

[0076] The data acquisition module is used to acquire the set of anomaly check items, the dictionary of full check results, the priority queue of processing modules, and the set of condition configurations.

[0077] The permission resolution module is used to filter available condition configurations based on the target organization identifier, organization processing level, applicable object rules, and information processing level, and form a module execution sequence according to the priority queue of the processing modules.

[0078] The factor extraction module is used to take the set of anomaly checks as the current set of anomaly checks, call the processing module according to the module execution sequence, extract the current factor set from the current set of anomaly checks, and read the missing checks from the full check result dictionary to obtain the supplementary factor set.

[0079] The condition matching module is used to perform combined matching on the current factor set and the supplementary factor set based on multiple matching conditions included in the available condition configuration, and to determine that the available condition configuration is triggered when the multiple matching conditions are met, so as to obtain the module processing record and the set of participating item indexes.

[0080] The consumption update module is used to remove consumed check items according to the participation item index set when the processed record of the module meets the output conditions. The consumed check items are the abnormal check items that actually participated in the combined matching and were removed. The module also updates the consumption status of the cross-module auxiliary attribute in the full check result dictionary, and the consumption status is updated from non-consumed to consumed.

[0081] The information output module is used to perform sorting and deduplication and factor inclusion deduplication on the records processed by the module, and output structured inspection information.

[0082] The beneficial effects of this invention are as follows: Since the set of anomaly checks assumes a variable exclusive consumption state, and the dictionary of full check results serves as a read-only supplementary data source, the preceding processing module can prevent the same anomaly check item from becoming the primary triggering factor for subsequent modules through removal operations. Simultaneously, subsequent modules can still obtain the original values ​​required for relational calculations, thus balancing output deduplication and data dependency integrity. Because the two-dimensional condition group represents intra-group and inter-group logic hierarchically, and the normal range comparison results from the dictionary of full check results are injected into the condition state vector by variable reference markers, complex combinations of conditions can be completed through data configuration without rewriting processing code for each condition combination. Permission resolution generates available condition configurations before condition matching, decoupling the applicable scope of the organization from the data processing logic, which helps reduce inconsistencies between multiple organization configuration copies. Attached Figure Description

[0083] Figure 1 This is a flowchart illustrating the health check data chain collaborative processing method provided in an embodiment of the present invention;

[0084] Figure 2 This is a schematic diagram of the module structure of the health check data chain collaborative processing system provided in an embodiment of the present invention. Detailed Implementation

[0085] The present invention will be further described below with reference to the accompanying drawings and embodiments. The following embodiments are used to explain the technical concept of the present invention and are not intended to limit the scope of protection. The target institutions, inspection item standard identifiers, values, and configuration contents in the embodiments are exemplary data and do not point to any specific institution or individual. The structured inspection information is the intermediate data processing result generated by the computer system, for editing, reviewing, or displaying by the report generation interface, and does not directly provide health conclusions for specific target objects.

[0086] Example 1

[0087] like Figure 1 As shown, this embodiment provides a chain-based collaborative processing method for health examination data, applied to a health examination report data processing system, including the following steps.

[0088] S1. Obtain the set of exception check items, the dictionary of full check results, the priority queue of processing modules, and the set of condition configurations.

[0089] The health examination report data processing system can receive structured examination results through a data access interface, map examination item names to standard identifiers, and copy the set of abnormal examination items into a variable working set. The full examination result dictionary serves as a read-only data source, storing both normal and abnormal items during this processing; only the consumption status of cross-module auxiliary attributes is allowed to be updated according to step S5. The processing module priority queue can be implemented using an array, linked list, or queue with a priority field, and the condition configuration set can be stored using JSON, a relational table, or a key-value document.

[0090] Specifically, the data access interface can generate a unique processing task identifier for each processing task and write the same processing task identifier into the anomaly check item set, the full check result dictionary, and the condition configuration snapshot. Each check result carries at least the check item standard identifier, check result value, unit of measurement, reference range, collection time, and data source identifier. For check results that do not carry a reference range or cannot complete unit normalization, the system retains the original fields and writes the data quality status, so that subsequent condition matching can keep the corresponding condition in an unresolved state, rather than mistakenly treating incomplete input as a condition met. By establishing the association between data in the same batch through processing task identifiers and configuration snapshots, the mixing of check results from different batches or different versions of condition configurations during a single execution can be avoided.

[0091] Optionally, the set of anomaly check items and the dictionary of full check results can be formed by splitting the same normalized result. Before splitting, duplicate records are sorted according to the check item standard identifier and collection time; for multiple records with the same check item standard identifier, one main record can be retained according to the preset data source priority, and the other records can be saved as auxiliary records. The splitting only changes the data organization form and does not change the original check result value. Thus, the set of anomaly check items is suitable for orderly removal operations, and the dictionary of full check results is suitable for key-based supplementary reading operations, and the two maintain consistency in data source.

[0092] S2. Based on the target organization identifier, organization processing level, applicable object rules, and information processing level, filter available condition configurations and form a module execution sequence according to the priority queue of the processing modules.

[0093] Permission resolution is performed before the processing module is invoked. The target organization identifier can be an organization code that has been hashed or mapped. The organization processing level and information processing level are both internal data processing permission parameters of the system. The resolution result only determines whether the conditional configuration can be loaded; it does not change the value of the check items or the condition matching logic. In other implementations, the same configuration filtering function can also be achieved using role-based access control lists or attribute policy expressions.

[0094] S3. Using the set of anomaly check items as the current set of anomaly check items, the processing module is called according to the module execution sequence to extract the current factor set from the current set of anomaly check items and read the missing check items from the full check result dictionary to obtain the supplementary factor set.

[0095] The current factor set stores unconsumed anomaly check items that can serve as primary trigger conditions, while the supplementary factor set stores reference factors read from the full check result dictionary. When the processing module reads supplementary factors, it does not re-insert them into the current anomaly check item set, thereby maintaining the consumption boundary of the variable working set. In one implementation, each factor contains a source type field, whose values ​​include the primary trigger source and the reference source, to facilitate subsequent determination of the participating item index set.

[0096] In this embodiment of the invention, chained collaborative processing refers to multiple processing modules sharing the same current set of anomaly checks, and sequentially obtaining input, generating module processing records, and changing the range of main triggering factors visible to subsequent processing modules according to the module execution sequence. This mechanism differs from multiple processing modules performing parallel matching on the same static copy: the preceding processing module only consumes the checks that actually participated in the matching after forming a module processing record that meets the output conditions; checks that did not participate in the matching are retained, and processing modules that have not formed a record that meets the output conditions do not change the current set of anomaly checks.

[0097] The underlying principle of this mechanism is to map "whether a new module can be triggered to process the record" and "whether the original value can still be read" to two separate data structures. The current set of anomaly checks determines the primary trigger qualification, while the dictionary of full check results provides read-only values ​​and reference ranges. After a preceding processing module removes a check, subsequent processing modules can no longer use that check as a new primary trigger factor, but can still use the corresponding original value as a supplementary factor for relational calculations if other primary trigger factors already exist. This prevents the same anomaly check from repeatedly dominating multiple outputs, and avoids data gaps in combined conditions due to the physical deletion of original values.

[0098] In one example, the initial set of anomaly checks sequentially contains four anomaly checks: Item A is too high, Item B is too low, Item C is too high, and Item D is too low. The full check result dictionary also contains the result within the reference range for Item E. The first processing module matches Item A and Item B and forms a module processing record, with the participating item index set being {0,1}. After consumption, the current set of anomaly checks retains only Item C and Item D. The second processing module can use Item C as the primary triggering factor and read the original value of Item A and the result within the reference range of Item E from the full check result dictionary as supplementary factors. Item A will not re-enter the current set of anomaly checks and therefore will not become a consumption object again. This example demonstrates that the consumption boundary is jointly determined by the factor source type and the current set membership.

[0099] S4. Based on the multiple matching conditions included in the available conditions configuration, perform combined matching on the current factor set and the supplementary factor set. When the multiple matching conditions are met, determine that the available conditions configuration is triggered to obtain the module processing record and the participation item index set.

[0100] The "multiple matching conditions" mentioned in step S4 are not simply listed in a flat, single-layer condition list, but are structured and parsed through a "two-dimensional condition group" with a two-layer nested logical relationship. Specifically, this two-dimensional condition group divides multiple matching conditions into two levels: outer candidate groups and inner conditions. Each outer candidate group contains one or more inner conditions. All inner conditions within the same outer candidate group are logically ANDed, meaning that all conditions must be met simultaneously for the candidate group to be considered valid. Different outer candidate groups are logically ORed, meaning that if at least one outer candidate group is determined to be satisfied, the entire two-dimensional condition group is determined to be triggered. Through this hierarchical nested structure, the complex combination matching logic of "all conditions within the group are satisfied, and at least one group among multiple candidate groups is satisfied" can be accurately expressed in a unified data configuration method, avoiding the need to write separate processing code for each combination relationship. To facilitate a quantitative description of the matching determination process, the two-dimensional condition group and its Boolean matching results can be further formalized as the following candidate group array structure:

[0101] A two-dimensional condition set can be represented as an array of candidate groups, where each candidate group contains multiple condition objects. Let the first... The available condition configurations include The outer candidate group, the first The outer layer candidate groups include The inner condition, the first The Boolean matching value of each inner condition is The labeling result of the two-dimensional condition group can be expressed as:

[0102] (1)

[0103] in, Indicates the first The results of marking two-dimensional condition groups corresponding to each available condition configuration; Indicates the first The number of outer candidate groups that can be configured with available conditions; Indicates the first The first available condition configuration The number of inner conditions for each outer candidate group; Indicates the first The first available condition configuration Within the outer candidate group, the first Boolean matching values ​​for each inner condition; Indicates the availability of conditional index configuration; Indicates the outer candidate group index; The inner condition index is represented. In equation (1), the inner product corresponds to all conditions being met within the same group, and the outer complement product corresponds to at least one condition being met among multiple candidate groups.

[0104] The minimum number of matching factors is used to provide a fast matching path alongside the two-dimensional condition group. Both matching paths must record the actual referenced factors, source type, and condition identifier, allowing the module processing record to trace back to the participant index set. In one example, a condition configuration contains two candidate groups. The condition state vector of the first candidate group is [1,1,0], and the condition state vector of the second candidate group is [1,1]. In this case, the first candidate group does not satisfy the condition, while the second candidate group does. The value is 1.

[0105] S5. When the module processes a record that meets the output conditions, it removes consumed check items according to the participating item index set. The consumed check items are the abnormal check items that actually participated in the combined matching and were removed. It also updates the consumption status of the cross-module auxiliary attributes in the full check result dictionary, changing the consumption status from unconsumed to consumed.

[0106] The removal operation is performed in descending index order to avoid elements at smaller indices being moved prematurely and causing subsequent indices to become invalid. Cross-module auxiliary attributes can be derived from historical question-and-answer text, previous record summaries, or external interface tags, but in this method, they are only used as structured attribute values. The consumption status is used to notify subsequent processing modules that the auxiliary attribute has been included in the module processing record of the preceding module, and is not used to generate conclusions for the target object.

[0107] Furthermore, consumption updates can be performed as an atomic operation within the scope of the processing task. Before a consumption update begins, the data version, number of elements, and summary of the participating item index set of the current anomaly check item set are saved. After the consumption update is completed, the number of remaining elements, the number of consumption records, and the data version increment are checked. If any check result does not meet the preset consistency condition, the system does not commit the update and restores the current anomaly check item set based on the set snapshot before the consumption update. By incorporating module processing record writing, check item removal, and cross-module auxiliary attribute consumption status updates into the same commit boundary, it can prevent situations where only output records are written without removing checks, or where checks are removed without leaving a partial completion status of the consumption records.

[0108] It should be noted that the consumption status of cross-module auxiliary attributes only indicates whether the attribute has been used in the processing chain; it does not change the original content carried by the cross-module auxiliary attribute. When a subsequent processing module reads a consumed status, it can decide to skip the cross-module auxiliary attribute or use it only as reference information based on the corresponding availability configuration. Thus, the consumption status can be passed along with the processing module sequence while preserving the traceability of consumption records for status changes.

[0109] S6. Perform sorting and deduplication and factor inclusion deduplication on the records processed by the module, and output structured inspection information.

[0110] Same-identity sorting deduplication eliminates records with the same result identifier but different sorting levels. Factor inclusion relation deduplication eliminates records that have been replaced by records with more complete information. Structured inspection information can be presented as a data object containing fields such as result_id, evidence_ids, match_path, consume_trace, and permission_tag, and passed to the report generation program via an interface. The report generation program can convert the structured data object into editable report prompts, which can be reviewed by authorized personnel.

[0111] Specifically, the result identifier in the structured inspection information is used to associate with a preset prompt template; the evidence item identifier is used to point to the inspection item standard identifier participating in the combined matching; the matching path is used to record the candidate groups and condition states that generate the combined matching results; the consumption record is used to record the removal trajectory of the main triggering factor; and the applicable permissions are used to identify the target institution scope corresponding to the structured data object. These fields together constitute a machine-readable intermediate data object, enabling the report generation program to reconstruct the data source and processing path without re-executing condition matching.

[0112] In this embodiment of the invention, if a configuration parsing failure, a full check result dictionary addressing failure, or a data version inconsistency occurs during the execution of a processing module, the system can mark the execution status of the corresponding processing module as incomplete and continue to retain the relevant abnormal check items to avoid consumption when there is a lack of valid module processing records. The incomplete status can be written to the execution log along with the processing task identifier and error location for subsequent retries. During retries, the current set of unconsumed abnormal check items and the same configuration version are still used as input, thereby ensuring that consistent structured check information is obtained when the same processing task is executed repeatedly.

[0113] In this embodiment, exclusive consumption occurs in the set of anomaly check items, while full readout occurs in the dictionary of full check results. These two data structures fulfill different responsibilities. Since consumed check items will not be used as primary triggering factors again, but their values ​​can still be used as reference factors in subsequent relationship calculations, the data processing flow reduces redundant output without disrupting the dependency of subsequent modules on the original values.

[0114] Example 2

[0115] This embodiment refines the construction of the anomaly check item set and the full check result dictionary.

[0116] S1. The acquisition of the anomaly check item set, the full check result dictionary, the processing module priority queue, and the condition configuration set includes:

[0117] S1.1. Normalize the names and units of the original inspection results according to the standard identifiers of the inspection items.

[0118] S1.2. Encapsulate the inspection results that exceed the corresponding reference range into anomaly inspection items, and write them into the anomaly inspection item set in the original input order.

[0119] S1.3. Write all the inspection results that have completed name normalization and unit normalization into the full inspection result dictionary, using the inspection item standard identifier as the key.

[0120] S1.4. Save the set index, standard identifier of the check item, check result value, reference range and abnormal direction for the abnormal check item.

[0121] S1.5. Write the processing modules whose processing scopes overlap into the processing module priority queue according to the preset business priority.

[0122] S1.6. Write the condition identifier, factor field, two-dimensional condition group, minimum number of matching factors, applicable object rules, and information processing level into the condition configuration set.

[0123] Name normalization can be achieved using an alias mapping table, while unit normalization can be achieved using the conversion function corresponding to the check item standard identifier. The set of anomaly checks retains its input order so that the set index can be used as a removal address; the dictionary of full check results is directly addressed using the check item standard identifier. In other implementations, the set of anomaly checks can also use an ordered mapping structure with stable element identifiers, in which case the set of participating item indexes is replaced by the set of participating item identifiers.

[0124] Specifically, the alias mapping table may include the original name, the inspection item standard identifier, the standard unit of measurement, and the conversion rule identifier. After the data access interface obtains the original inspection result, it first looks up the inspection item standard identifier based on the original name, and then converts the inspection result value to the standard unit of measurement based on the conversion rule identifier. If the same original name can be mapped to multiple inspection item standard identifiers, a secondary matching is performed by combining the data source identifier and the original project code; if it still cannot be uniquely determined, the data quality status of the corresponding inspection result is set to pending confirmation, and it is not written to the abnormal inspection item set for the time being.

[0125] In this embodiment of the invention, the abnormal direction may include below the reference range, above the reference range, and non-numerical state mismatch. For numerical inspection results, the abnormal direction is determined according to the upper and lower limits of the reference range and the boundary inclusion identifier; for enumeration-type inspection results, the original text is converted into standard state values ​​according to a preset enumeration mapping table, and then compared with the reference state set. The reference range in the abnormal inspection item stores the reference range after unit normalization, thereby ensuring that the subsequent normal range reference conditions and abnormal direction identifiers use the same units.

[0126] Preferably, each queue element in the processing module priority queue includes a processing module identifier, a service priority, a configuration version, and an activation status. The service priority can be determined based on the number of factors required by the processing module, the inclusion relationship of its processing scope, and a preset output priority; processing modules requiring more factors or whose processing scope includes the processing scope of other modules can be ranked higher. For processing modules with the same service priority, the execution order can be determined according to a fixed sorting value of the processing module identifier, ensuring a stable module execution sequence under the same input.

[0127] In one example, the original inspection results contain two alias records for project A and one record for project B. The system normalizes the two alias records to the same inspection item standard identifier A001 and retains the master record according to the collection time and data source priority. When the standardized inspection result value of A001 exceeds the corresponding reference range, A001 is written to the abnormal inspection item set, and the master record and the auxiliary record of A001 are written to the same key-value item in the full inspection result dictionary. Thus, the subsequent processing module obtains the unique master triggering factor from the abnormal inspection item set, while still being able to trace the original record in the full inspection result dictionary.

[0128] Example 3

[0129] This embodiment describes the consumption update mechanism for anomaly check items.

[0130] S5. When the module processes a record that meets the output conditions, removing consumed inspection items according to the participating item index set and updating the cross-module auxiliary attribute consumption status in the full inspection result dictionary includes:

[0131] S5.1. Before calling the current processing module, record the data version and number of elements of the current exception check item set.

[0132] S5.2. Write the index of the anomaly check item that matches the current processing module and actually participates in the combination matching into the participation item index set.

[0133] S5.3. Perform deduplication and descending order sorting on the participating item index set to obtain a descending index sequence.

[0134] S5.4. Verify that the data version of the current anomaly check item set is consistent with the data version recorded before calling the current processing module.

[0135] S5.5. Remove consumed check items one by one according to the descending index sequence to keep the set index of unconsumed check items valid.

[0136] S5.6. Generate consumption records for the removed consumed items, including the consumption module identifier, consumption order, and consumption time.

[0137] S5.7. When the current processing module uses cross-module auxiliary attributes, the consumption status of the corresponding cross-module auxiliary attributes is updated from unconsumed to consumed.

[0138] S5.8. Determine the set of exception checks after the removal operation as the current set of exception checks for the next processing module.

[0139] Data version verification is used to avoid race conditions caused by concurrent modules on the same mutable set. If the versions are inconsistent, the removal can be abandoned and factor matching can be re-executed based on the new version. The principle of descending removal is that removing a larger index element does not change the position of a smaller index element. Consumption records can also contain condition configuration versions and participant summaries for processing chain tracing. In one example, the participant index set is {2,5,2,7}, which, after deduplication and descending order, becomes [7,5,2]. Removing these elements sequentially will not invalidate any indexes that have not yet been executed.

[0140] In this embodiment of the invention, the participant item index set only collects the indexes of anomaly check items that actually generate combined matching results and originate from the current factor set. Anomaly check items that are read during condition parsing but do not affect the combined matching results are not included in the participant item index set, nor are supplementary factors read from the full check result dictionary. By distinguishing between "being read" and "being consumed," the processing module can be prevented from expanding its consumption scope due to pre-reading multiple candidate check items.

[0141] Specifically, data versions can be represented using monotonically increasing integers. When the current processing module starts execution, it reads the first data version, and after completing the combination matching, it reads the second data version. Submission of the descending index sequence is only allowed if the first and second data versions match. Upon successful submission, the data version is incremented by a preset increment. If the two versions do not match, it indicates that another update operation has changed the current anomaly check item set. The current processing module discards the old participant index set and re-extracts the current factor set based on the updated current anomaly check item set.

[0142] It is worth noting that descending removal is used not only to maintain the validity of the set index, but also to ensure that the consumption records are consistent with the actual removal order. Before each removal, it can be verified whether the index in the descending index sequence falls within the current number of elements, and whether the check item standard identifier at the index is consistent with the participating item summary; if the verification fails, the submission is stopped and the original set is retained. Each time a consumed check item is removed, a consumption record is generated. The consumption record is associated with the processing task identifier, processing module identifier, and configuration version, thus enabling the consumption source to be located backward from the final structured check information.

[0143] Optionally, when processing tasks are executed sequentially by a single thread, data version verification can still be used to identify abnormal reentrancy or duplicate submissions. When processing tasks are coordinated by multiple service processes, version verification and descending order removal can be encapsulated into a conditional update operation, ensuring that only the processing module holding the desired data version can complete the submission. None of the above implementations change the logical execution order formed according to the priority queue of processing modules.

[0144] Through the aforementioned consumption and update mechanism, the preprocessing module's occupation of the main triggering factor can generate verifiable state changes, rather than relying solely on implicit agreements between modules. Data versioning ensures that the updated object is not concurrently modified, the participant item summary ensures that the removed object is consistent with the matched object, the descending index sequence ensures the address stability of the ordered set, consumption records ensure the update process is traceable, and cross-module auxiliary attribute consumption status ensures that non-checked attributes can also pass on their usage status along the processing chain. These multiple constraints work together to reduce the probability of duplicate consumption, misaligned removal, and partial commits.

[0145] Example 4

[0146] This embodiment refines the combination matching of two-dimensional condition groups.

[0147] S4. Based on the multiple matching conditions included in the available condition configuration, perform combined matching on the current factor set and the supplementary factor set. When the multiple matching conditions are met, determine that the available condition configuration is triggered, and obtain the module processing record and the participating item index set, including:

[0148] S4.1. The two-dimensional condition group configured with the available conditions is parsed into multiple outer candidate groups, and each outer candidate group is parsed into multiple inner conditions, wherein the inner conditions are the matching conditions.

[0149] S4.2. Match the inner condition with the current factor set and the supplementary factor set respectively to obtain the inner condition matching result of each inner condition.

[0150] S4.3. When all the inner conditions in the same outer candidate group are satisfied, the corresponding outer candidate group is marked as satisfied, and the outer candidate group marking result is obtained.

[0151] S4.4. When the outer candidate group marking result of at least one outer candidate group is satisfied, mark the two-dimensional condition group as satisfied, and determine that the available condition configuration is triggered.

[0152] S4.5. Summarize the anomaly check item indices that participate in the outer candidate group that is marked as satisfied and originate from the current factor set to form the participation item index set.

[0153] S4.6. Generate the module processing record based on the marking results of the two-dimensional condition group.

[0154] Two-dimensional condition groups can be stored as arrays of arrays, condition trees, or equivalent directed acyclic condition graphs. When using arrays of arrays, the parser traverses the outer candidate groups, using short-circuit AND operations within each outer candidate group. After finding a satisfying outer candidate group, the parser can decide whether to stop matching or continue recording other satisfying paths based on the configured exhaustive field. Continuing to record other satisfying paths helps form a complete match_path, while stopping matching reduces computational overhead.

[0155] In this embodiment of the invention, a two-dimensional condition group refers to a data structure that arranges multiple candidate condition sets that can independently generate combined matching results in the outer layer, and arranges the conditions that must be commonly satisfied in the same candidate condition set in the inner layer. The outer candidate groups are related by any satisfied condition, while the inner conditions are related by all satisfied conditions. This structure separates the condition logic from the processing code, allowing the same processing module to complete different combined matching by loading different two-dimensional condition groups without changing the execution interface of the processing module.

[0156] The parsing process of the two-dimensional condition group may include structure verification, condition addressing, state calculation, and path recording. Structure verification confirms that the outer layer contains at least one candidate group, each candidate group contains at least one inner condition, and that the condition identifier does not contain null values ​​or circular references. Condition addressing retrieves the corresponding data from the current factor set or supplementary factor set according to the condition identifier. State calculation generates the satisfied, unsatisfied, or unresolved state of each inner condition. Path recording stores the outer candidate group identifier and inner condition identifier that generated the two-dimensional condition group labeling result. When an unresolved state exists and no satisfied outer candidate group exists, the entire two-dimensional condition group can be kept unresolved to avoid equating missing data with unsatisfied conditions.

[0157] Specifically, the calculation order of inner-layer conditions can be determined according to the data source: first, match the abnormal conditions in the current factor set, and then parse the normal range reference conditions in the supplementary factor set. If the current factor set lacks a check item standard identifier corresponding to an abnormal condition, the abnormal condition is marked as unsatisfied; if the check result corresponding to the normal range reference condition is missing or the reference range is invalid, the normal range reference condition is marked as unresolved. Only when all inner-layer conditions of an outer-layer candidate group are satisfied is the outer-layer candidate group marked as satisfied.

[0158] In one example, the two-dimensional condition set contains outer candidate sets G1 and G2. G1 includes three inner conditions: "Item A is high," "Item B is low," and "Item C is within the reference range." G2 includes two inner conditions: "Item A is high" and "Item D is high." The current factor set contains "Item A is high" and "Item D is high," and the supplementary factor set contains the results for "Item C is within the reference range." The parser obtains the conditional state vector [1,0,1] for G1 and the conditional state vector [1,1] for G2. Therefore, G2 is marked as satisfied, and the two-dimensional condition set is marked as satisfied. The participating item index set only summarizes the indices of items A and D in the current anomaly check item set.

[0159] Through the hierarchical condition structure described above, multiple alternative combination paths can be expressed in the same condition configuration, and necessary conditions within a group will not be replaced by simple factor counting. The matching path simultaneously saves the condition state and factor source, enabling the subsequent deduplication, consumption update, and review interfaces to identify which candidate group generated the module's processed record and accurately determine the current factor that should be consumed.

[0160] Example 5

[0161] This example illustrates the unified matching of normal range reference conditions and abnormal conditions.

[0162] S4.2. The step of matching the inner condition with the current factor set and the supplementary factor set respectively includes:

[0163] S4.2.1. Identify the variable reference markers in the inner condition and extract the referenced check item standard identifier from the variable reference markers.

[0164] S4.2.2. Read the inspection result value, unit of measurement, and reference range from the full inspection result dictionary using the referenced inspection item standard identifier as the key.

[0165] S4.2.3. Convert the inspection result value to a standardized inspection result value according to the standard unit of measurement corresponding to the reference range.

[0166] S4.2.4. Parse the upper limit, lower limit, and boundary inclusion identifier of the reference range.

[0167] S4.2.5. Based on the boundary inclusion identifier, compare the standardized check result value with the upper limit value and the lower limit value respectively to obtain the Boolean matching value of the normal range reference condition.

[0168] S4.2.6. For inner conditions that do not contain the variable reference marker, compare the corresponding condition identifier with the abnormal direction identifier in the current factor set to obtain the Boolean matching value of the abnormal condition.

[0169] S4.2.7. Write the Boolean matching value of the normal range reference condition and the Boolean matching value of the abnormal condition into the condition state vector of the corresponding outer candidate group.

[0170] Let the standardized inspection result value be... The lower limit of the reference range is The upper limit of the reference range is The lower boundary contains parameters. The upper boundary contains parameters as The Boolean match value for a normal range reference condition can then be expressed as:

[0171] (2)

[0172] in, Indicates the first Boolean match value for a normal range of reference conditions; An indicator function that takes 1 when the logical expression within the parentheses is true and 0 otherwise; Indicates the first The standardized check result value of each referenced check item; and They represent the first The lower and upper limits of the reference range for each referenced check item; and These indicate whether the lower and upper boundaries contain endpoints, respectively. This indicates a normal range reference to a conditional index.

[0173] Variable reference markers can be enclosed in pairs of delimiters to enclose the check item standard identifier. During parsing, escape characters are processed first, and then the existence of the referenced check item standard identifier in the full check result dictionary is verified. If the check result value is missing, the unit cannot be converted, or the reference range is incomplete, the corresponding Boolean matching value is set to a pending state instead of being directly set to satisfied. Using a pending state can avoid generating deterministic output when the input is incomplete.

[0174] Specifically, variable reference markers essentially convert the keys in the full check result dictionary into condition objects that can be evaluated by the condition parser. After recognizing the delimiters, the condition parser does not directly treat the text inside the markers as the displayed text. Instead, it first performs a validity check on the check item standard identifier, and then accesses the full check result dictionary using the check item standard identifier. Therefore, check results that are within the reference range and not written into the abnormal check item set can also participate in the calculation of the two-dimensional condition group as Boolean matching values.

[0175] When performing unit conversions, the condition parser reads the unit of measurement of the inspection result value and the standard unit of measurement corresponding to the reference range. If they match, the inspection result value is used directly; if a conversion relationship exists, the standardized inspection result value is obtained according to the conversion rule corresponding to the standard identifier of the inspection item; if no conversion relationship is configured, the condition status is set to pending. The reference range parser can also distinguish between closed intervals, open intervals, and one-sided intervals. For reference ranges with only a lower or upper limit value, the missing side is not included in the comparison, while the existing side is still compared according to the boundary inclusion identifier.

[0176] It should be noted that the condition state vector can be stored using a combination of binary values ​​and pending flags. A Boolean matching value of 1 indicates that the corresponding inner condition is met, and 0 indicates that the corresponding inner condition is not met. The pending flag indicates that the current input is insufficient to complete the comparison. When calculating the outer candidate group, the two-dimensional condition group parser first checks whether there is an inner condition with a value of 0. If so, the outer candidate group is marked as unmet. If there is no inner condition with a value of 0 but a pending flag exists, the outer candidate group remains unmet. Only when all values ​​are 1 is it marked as met.

[0177] In one example, the variable reference marker points to item C. The full check result dictionary records the check result value of item C as 5.2 units, with a reference range of [3.0 units, 6.0 units], including endpoints at both the upper and lower boundaries. After completing the unit consistency verification, 5.2 falls within the reference range, and the Boolean match value for the corresponding normal range reference condition is 1. If the same result is input in another convertible unit, it is first converted to the standard unit of measurement before comparison; if the conversion cannot be completed, the corresponding condition remains unresolved, and no satisfied state is generated.

[0178] The variable referencing mechanism allows abnormal conditions and normal range reference conditions to enter the same condition state vector. Abnormal conditions are primarily triggered by the current factor set, while normal range reference conditions are supplemented by constraints provided by the dictionary of full check results. Both participate in the full satisfaction calculation within the outer candidate group. This allows for the expression of a combination relationship where "at least one unconsumed abnormal check item serves as the primary trigger, while several check items are within the reference range," without mistakenly writing check results within the reference range into the abnormal check item set.

[0179] Example 6

[0180] This embodiment illustrates the combination relationship between the minimum number of matching factors and the two-dimensional condition set.

[0181] S4. Based on the multiple matching conditions included in the available condition configuration, perform combined matching on the current factor set and the supplementary factor set. When the multiple matching conditions are met, determine that the available condition configuration is triggered, and obtain the module processing record and the participating item index set, which also includes:

[0182] S4.1'. Remove duplicate check item standard identifiers from the current factor set to obtain the number of different factors.

[0183] S4.2'. Compare the number of different factors with the minimum number of matching factors configured in the available conditions. When the number of different factors reaches the minimum number of matching factors, mark the factor number matching status as satisfied and mark the available conditions configuration as triggered.

[0184] S4.3'. When the factor quantity matching state is not satisfied, the two-dimensional condition group of the available condition configuration is parsed into multiple outer candidate groups, and each outer candidate group is parsed into multiple inner conditions. The inner conditions are matched with the current factor set and the supplementary factor set, respectively. The normal range reference condition is used as the inner condition for matching. When all the inner conditions in the same outer candidate group are satisfied, the corresponding outer candidate group is marked as satisfied. When at least one outer candidate group is marked as satisfied, the two-dimensional condition group is marked as satisfied, and the available condition configuration is marked as triggered.

[0185] S4.4'. When the factor quantity matching state is marked as satisfied or the two-dimensional condition group is marked as satisfied, the available condition configuration is determined to be triggered, and the condition identifier, factor source and processing sorting level are recorded to form the module processing record.

[0186] S4.5'. When the factor quantity matching state is marked as satisfied, only the check item indexes that originate from the current factor set and participate in the counting of different factor quantities are written into the participation item index set; when the two-dimensional condition group is marked as satisfied, only the check item indexes that originate from the current factor set and actually participate in the matching of the two-dimensional condition group are written into the participation item index set.

[0187] Let the first The number of different factors corresponding to each condition configuration is The minimum number of matching factors is Then the factor quantity matching status is:

[0188] (3)

[0189] in, Indicates the first Each condition configuration corresponds to the number of factors matching status; Indicates an indicator function; Indicates the first The number of different factors corresponding to each condition configuration; Indicates the first The minimum number of matching factors corresponding to each condition configuration; This indicates the condition configuration index. Equation (3) is not calculated when the minimum number of matching factors is null. In this case, the factor count matching path is invalid, and the process directly proceeds to step S4.3' for two-dimensional condition group matching. Factor deduplication uses the check item standard identifier as the key to prevent duplicate data rows of the same check item from increasing the count.

[0190] In this embodiment of the invention, the minimum number of matching factors is used to quickly match different anomaly check items that do not require a fixed combination path but require a certain number, corresponding to the factor number matching path in step S4.2'. Different factor numbers only count the different check item standard identifiers in the current factor set, excluding reference factors in the supplementary factor set, and also excluding duplicate counting of multiple data records with the same check item standard identifier. Therefore, the factor number matching status is always consistent with the range of the main triggering factors capable of performing consumption updates.

[0191] Specifically, the combination matching first verifies whether the minimum number of matching factors is a valid value according to step S4.2'. A valid value should be a positive integer and not greater than the number of different check item standard identifiers listed in the factor field of the available condition configuration. When the number of different factors reaches the threshold, a factor number matching path is formed, the factor number matching status is marked as satisfied, and the available condition configuration is marked as triggered. At the same time, according to the rules of the factor number matching path in step S4.5', only the check item indexes from the current factor set that participate in the counting of different factor numbers are written into the participating item index set. When the number of different factors does not reach the threshold, the factor number matching status is not satisfied. Then, according to step S4.3', the matching of the two-dimensional condition group and the normal range reference condition continues. When at least one outer candidate group is marked as satisfied, the two-dimensional condition group is marked as satisfied, and the available condition configuration is marked as triggered. Finally, step S4.4' performs a logical OR operation on the factor number matching status and the marking result of the two-dimensional condition group. As long as either one is marked as satisfied, the available condition configuration is determined to be triggered, and the condition identifier, factor source, and processing sorting level are recorded to form a module processing record.

[0192] In one example, a factor field configured for a certain available condition includes item A, item B, item C, and item D, with a minimum matching factor count of 3. The current factor set contains two duplicate records for item A, one record for item B, and one record for item C. After deduplication according to the check item standard identifier, the number of different factors is 3, and the factor count matching status is marked as satisfied. Following the rules in step S4.5', the participating item index set only records check item indices from the current factor set that participate in the different factor counts under the factor count matching path, i.e., the main record for item A, and the indices corresponding to items B and C. Duplicate records of item A do not increase the count, nor do they change the combined matching result. If a two-dimensional condition group path is triggered, then according to the rules in step S4.5' regarding two-dimensional condition group paths, only check item indices from the current factor set that actually participate in the two-dimensional condition group matching are written to the participating item index set.

[0193] By juxtaposing the minimum number of matching factors path with the two-dimensional condition group path, two types of condition configurations can be handled separately: "meeting a certain number is sufficient" and "a specified combination must be met." These correspond to the two decision paths in steps S4.2' and S4.3', and are then uniformly summarized and determined in step S4.4'. After the module processes and records the path type, condition identifier, and factor source, the consumption update can only apply to the anomaly check items that actually participate in the current matching path, avoiding the removal of candidate factors that did not participate in the calculation. The path type includes at least the factor number matching path and the two-dimensional condition group matching path, which are used to distinguish the writing range of the participating item index set in step S4.5' and assist in determining the processing sorting level in subsequent steps.

[0194] Example 7

[0195] This embodiment explains the parsing of applicable object rules and information processing levels.

[0196] S2. The configuration of available conditions for filtering based on target organization identifier, organization processing level, applicable object rules, and information processing level, forming a module execution sequence according to the priority queue of the processing modules, includes:

[0197] S2.1. Parse the applicable object rules into a universally applicable mode, an exclusion list mode, or a specified list mode.

[0198] S2.2. In the all-applicable mode, set the object matching status corresponding to the target organization identifier to allowed and assign a first sorting weight.

[0199] S2.3. In the exclusion list mode, when the target organization identifier does not belong to the exclusion identifier set, the object matching status is set to allowed.

[0200] S2.4. In the specified list mode, when the target organization identifier belongs to the specified identifier set, the object matching status is set to allowed, and a second sorting weight greater than the first sorting weight is assigned.

[0201] S2.5. Input the organization processing level and the information processing level into a preset level permission matrix to obtain the level matching status.

[0202] S2.6. The conditional configuration in which both the object matching status and the level matching status are allowed is determined as the available conditional configuration.

[0203] S2.7. Resolve matching conflicts between multiple applicable object rules in order of priority of the second sorting weight over the first sorting weight.

[0204] The "Applicable to All" pattern can be represented by a wildcard tag, the "Exclusion List" pattern can be represented by a combination of wildcard tags and exclusion tags, and the "Specified List" pattern can consist of one or more organization tags. To avoid delimiter conflicts in the configuration text, the parser can first perform escaping and syntax checking before generating the object matching state. Let the object matching state be... The level matching status is Then the first The available state of each condition configuration can be represented as:

[0205] (4)

[0206] in, Indicates the first Available status of each condition configuration; Indicates the first Each condition configuration represents the object matching status relative to the target organization identifier; Indicates the first The level matching status of each condition configuration; This indicates a conditional configuration index. The precisely specified second sorting weight is higher than the first sorting weight corresponding to the wildcard, allowing precise rules to cover general rules. In other implementations, the preset level permission matrix can be replaced with a configurable permission policy table.

[0207] In this embodiment of the invention, institution-level configuration filtering refers to determining whether a certain conditional configuration can be included in the current processing task by utilizing the target institution identifier, applicable object rules, institution processing level, and information processing level, without replicating the conditional matching logic. The applicable object rules address which target institutions the conditional configuration applies to, and the level permission matrix addresses whether the target institution's current processing level allows loading the corresponding information processing level. Only when both types of judgments are satisfied is an available conditional configuration generated.

[0208] The technical principle behind the institution-level configuration filtering is to express institution differences as metadata for conditional configurations, rather than writing them into the processing module. The processing module only receives the already filtered available conditional configurations and is unaware of the target institution name or list syntax. Therefore, when adding or adjusting the scope of application for a target institution, only the applicable object rules or level permission matrix need to be updated; the factor extraction and combination matching process of the processing module remains unchanged, which helps maintain the consistency of configuration versions in multi-target institution scenarios.

[0209] Specifically, the parser first performs lexical decomposition and syntax verification on the applicable rules, extracting the pattern type and the corresponding set of organization identifiers. For patterns applicable to all, an allowed state is generated if no higher-ranking rule exists; for exclusion list patterns, a general allowed state is generated first, followed by a rejected state for target organizations in the exclusion identifier set; for specified list patterns, an allowed state is generated only for target organizations in the specified identifier set. When multiple rules match simultaneously, the first and second ranking weights are compared first, and then a unique object matching state is determined according to a preset rejection-first or precise rule-first strategy.

[0210] The permission level matrix uses the organization processing level as the row index and the information processing level as the column index. Each matrix cell stores the allowed or denied status. During parsing, it first verifies that both level parameters are within the preset value range, and then reads the corresponding matrix cell. If the target organization has not configured an organization processing level, the conditional configuration does not configure a valid information processing level, or the matrix cell is missing, the level matching status is set to denied, and the configuration verification information is recorded, thereby avoiding expanding the scope of application of the conditional configuration when the permission data is incomplete.

[0211] In one example, conditional configuration C1 uses an all-applicable mode with a first sorting weight, while conditional configuration C2 uses a specified list mode, specifies the target organization M1, and has a second sorting weight. When the target organization for the processing task is identified as M1, both C1 and C2 generate an allowed object matching state; since the second sorting weight is higher than the first sorting weight, the system prioritizes the precise rule corresponding to C2. Subsequently, the organization processing level of M1 and the information processing level of C2 are input into the level permission matrix. When the matrix cell is allowed, C2 is determined to be an available conditional configuration; when the matrix cell is rejected, C2 does not enter the module execution sequence.

[0212] Through the above methods, a continuous filtering chain is formed by matching the scope of applicable object rules, matching the capabilities of the level permission matrix, and resolving conflicts in the ranking weights. This filtering chain is completed before the combined matching, ensuring that conditional configurations without permissions will not enter the factor extraction stage, nor will it generate a set of participating item indexes or consumption records. At the same time, the structured inspection information can still retain the applicable permission fields, which are used by the report generation interface to perform consistency verification on the output content.

[0213] Example 8

[0214] This embodiment illustrates the main triggering factor, reference factor, and dual-round processing mechanism.

[0215] S3. The step of using the set of anomaly checks as the current set of anomaly checks, calling the processing module according to the module execution sequence, extracting the current factor set from the current set of anomaly checks, and reading missing checks from the full check result dictionary to obtain a supplementary factor set, including:

[0216] S3.1. In the first processing round, the dedicated processing module and the configuration processing module are called according to the module execution sequence.

[0217] S3.2. Mark the anomaly check items matched from the current anomaly check item set as the main triggering factors and write them into the current factor set.

[0218] S3.3. When the required check item for the available condition configuration does not exist in the current factor set, the corresponding check item is read from the full check result dictionary and marked as a reference factor to form the supplementary factor set.

[0219] S3.4. Prohibit check item consumption operations from being triggered only by the set of supplementary factors marked as reference factors.

[0220] S3.5. After the first processing round ends, execute the second processing round for the available condition configurations that have not generated module processing records.

[0221] S3.6. In the second processing round, based on the dictionary parsing of the full check result, a two-dimensional condition group containing the normal range reference condition is parsed, and the newly generated module processing record is added to the candidate record set.

[0222] Dedicated processing modules can be used to perform numerical relationship calculations for specific combinations of check items, while configurable processing modules can be used to perform general condition configurations. The first processing round prioritizes completing the cross-item calculations required by the dedicated processing modules and reduces subsequent main trigger candidates through consumption updates. The second processing round is used to parse configurations that were not completed in the first processing round and depend on normal range reference conditions. If the second processing round only obtains reference factors but has no unconsumed main trigger factors, it can generate reference records without consumption actions, or mark the path as not output according to the configuration.

[0223] Specifically, at the start of the first processing round, a round state is established for each available condition configuration. This round state includes at least four states: not executed, module processing record formed, no module processing record formed, and execution incomplete. After a dedicated processing module completes its task, the configuration-based processing module only reads the current set of exception checks; checks consumed by previous modules no longer appear as primary triggering factors. At the end of the first processing round, available condition configurations with a status of "no module processing record formed" and containing normal range reference conditions are added to the second round candidate set.

[0224] The second processing round does not rebuild the initial set of anomaly checks; instead, it continues to use the current set of anomaly checks from the end of the first processing round, and allows reference factors to be read from the full check result dictionary. To avoid duplicate output of the same available condition configuration, before the second processing round begins, it checks whether a module processing record with the same condition identifier and configuration version already exists in the candidate record set; if it already exists, the corresponding configuration is skipped. The second processing round ends after completing one traversal of the candidate set, without generating a new third-round candidate set based on the results of this round, thus ensuring that the two-round processing has a clear termination condition.

[0225] It should be noted that the reference factor only provides supplementary values ​​for combined conditions and does not independently qualify for consumption of check items. When the second processing round forming module processes records, at least one primary triggering factor from the current set of anomaly check items is required, or the corresponding path must be marked as not output and not subject to consumption updates. This restriction ensures that the full check result dictionary does not replace the primary triggering role of the current set of anomaly check items and maintains the same consumption boundaries for the first and second processing rounds.

[0226] In one example, the dedicated processing module in the first processing round consumes items A and B. The configuration processing module C1 does not generate a module processing record because it lacks results within the reference range for item E. C1 contains normal range reference conditions, therefore it enters the candidate set for the second round. In the second processing round, item E is read from the full check result dictionary and combined with item C from the current anomaly check item set to complete a combination match; item C enters the participating item index set as the primary triggering factor, while item E, as a reference factor, does not enter the participating item index set. After this round, C1 is no longer scheduled repeatedly.

[0227] The dual-round processing establishes a clear temporal order between dedicated processing logic and configurable processing logic. The first round prioritizes combinations requiring the collaborative participation of multiple anomaly checks, thereby narrowing down the current set of anomaly checks. The second round then uses the full check result dictionary to supplement the normal range reference conditions. This reduces the loss of subsequent combinations caused by the configurable processing module consuming checks first, while avoiding infinite polling or supplementary factors triggering new consumption chains in reverse.

[0228] Example 9

[0229] This embodiment describes the deduplication and structured output of the module processing records.

[0230] S6. The process of sorting and deduplicating records processed by the module, and outputting structured inspection information, includes:

[0231] S6.1. Sort the processing records of the multiple modules according to the result identifier and processing sorting level.

[0232] S6.2. For module processing records with the same result identifier, retain the module processing record whose processing sorting level meets the preset priority condition.

[0233] S6.3. Construct a set of corresponding factor identifiers from the check item standard identifiers referenced by each retained module processing record.

[0234] S6.4. When the first factor identifier set is included in the second factor identifier set and the corresponding module processing record satisfies the preset substitution relationship, the module processing record corresponding to the first factor identifier set is removed.

[0235] S6.5. Encapsulate the result identifier, evidence item identifier, matching path, consumption record, and applicable permissions from the remaining module processing records into structured data objects.

[0236] S6.6. Sort the structured data objects according to the priority queue of the processing module to obtain the structured inspection information.

[0237] S6.7. Send the structured examination information to the health examination report generation interface, so that the health examination report generation interface can generate editable report prompt content.

[0238] For deduplication based on the same identifier, a stable sorting method can be used, ensuring that records processed by the module that arrived first are retained when processing order levels are the same. Deduplication based on factor inclusion relationships does not involve deleting records that are included in any set; instead, it simultaneously checks for predefined substitution relationships to prevent the erroneous removal of semantically non-substitutable records. Structured data objects do not contain deterministic health conclusions oriented towards the target object; result identifiers are used to retrieve predefined prompt templates, while evidence item identifiers and matching paths are used to display the data source to the review interface.

[0239] In this embodiment of the invention, sorting deduplication and factor inclusion deduplication are performed sequentially. Sorting deduplication first groups records by result identifier, then within each group, sorts them according to processing priority levels, from records meeting preset priority conditions to other records, and retains the order of execution of modules that maintain records at the same level through stable sorting. After compressing records with the same result identifier, a set of factor identifiers is constructed from the remaining processed records to reduce the number of records compared in factor inclusion relationships.

[0240] The preset substitution relationship can be determined jointly by the result identifier pair and the processing module identifier pair. The module processing record corresponding to the first factor identifier set is removed only if the first factor identifier set is contained within the second factor identifier set, and a preset substitution relationship exists between the two module processing records. If there is only set inclusion without a preset substitution relationship, both module processing records are retained. This dual condition avoids accidental deletion when different processing semantics happen to reference the same check item.

[0241] Specifically, each time a record is removed from a module, the result identifier, factor identifier set, and replacement record identifier of the removed record are written into the deduplication trajectory. The consumption records in the structured data object retain the original consumption process, while the deduplication trajectory retains the record compression process during the output stage. Together, they explain when a particular anomaly check item is consumed by a processing module and why a record processed by a module is not included in the final structured check information.

[0242] In one example, modules process records R1 and R2 with the same result identifier, but their processing sorting levels are 2 and 5, respectively. The preset priority condition is to retain the processing sorting level with the smaller value; therefore, after sorting and deduplication, R1 is retained. Another module processes record R3 with a factor identifier set of {A001, B002}, and R4 with a factor identifier set of {A001, B002, C003}. Since there is a preset substitution relationship between R3 and R4, R3 is removed, and R4 is recorded as replacing R3 in the deduplication trajectory. If there is no preset substitution relationship between R3 and R4, both records are retained even if the former set is contained within the latter set.

[0243] Structured data objects can be output using a unified field type. Result and condition identifiers are strings, evidence item identifiers are ordered arrays, matching paths are objects containing candidate group identifiers and condition states, consumption records are arrays of objects arranged in consumption order, and applicable permissions are permission tags composed of the target organization scope and information processing level. After receiving the above objects, the report generation interface selects a preset prompt template only based on the result identifier and displays the evidence item identifiers and matching paths as verifiable data evidence, thereby maintaining the boundary between the data processing layer and the text editing layer.

[0244] Example 10

[0245] like Figure 2 As shown, this embodiment provides a chain-based collaborative processing system for health check data, including:

[0246] The data acquisition module is used to acquire the set of anomaly check items, the dictionary of full check results, the priority queue of processing modules, and the set of condition configurations.

[0247] The permission resolution module is used to filter available condition configurations based on the target organization identifier, organization processing level, applicable object rules, and information processing level, and form a module execution sequence according to the priority queue of the processing modules.

[0248] The factor extraction module is used to take the set of anomaly checks as the current set of anomaly checks, call the processing module according to the module execution sequence, extract the current factor set from the current set of anomaly checks, and read the missing checks from the full check result dictionary to obtain the supplementary factor set.

[0249] The condition matching module is used to perform combined matching on the current factor set and the supplementary factor set based on multiple matching conditions included in the available condition configuration, and to determine that the available condition configuration is triggered when the multiple matching conditions are met.

[0250] The consumption update module is used to remove consumed check items according to the participation item index set when the processed record of the module meets the output conditions. The consumed check items are the abnormal check items that actually participated in the combined matching and were removed. The module also updates the consumption status of the cross-module auxiliary attribute in the full check result dictionary, and the consumption status is updated from non-consumed to consumed.

[0251] The information output module is used to perform sorting and deduplication and factor inclusion deduplication on the records processed by the module, and output structured inspection information.

[0252] The above modules can be deployed on the same server or implemented collaboratively by multiple service processes via interfaces. The data acquisition module provides an input snapshot, the permission parsing module provides a configuration snapshot, the factor extraction module and the condition matching module operate under the same processing task identifier, the consumption update module performs atomic updates with version verification on the anomaly check item set, and the information output module performs deduplication and encapsulation after all processing modules have completed. Each module can be implemented by the processor executing computer programs stored in memory.

[0253] Specifically, the data acquisition module's output interface can transmit processing task identifiers, anomaly check item set references, full check result dictionary references, processing module priority queue versions, and condition configuration set versions; the permission parsing module's output interface can transmit available condition configuration identifier sets and module execution sequences; the factor extraction module transmits the current factor set, supplementary factor set, and factor source type to the condition matching module; the condition matching module transmits module processing records, participant item index sets, and expected data versions to the consumption update module; and the consumption update module transmits consumption records and updated module processing record sets to the information output module. By explicitly defining the interface objects, modules do not need to share undeclared temporary variables.

[0254] When deployed on the same server, each module can function as a functional component within the same application process and pass data via memory objects. When deployed across multiple service processes, module output can be passed via serialized data objects and message queues, while data for the same task can be associated with processing task identifiers. Regardless of the deployment method, the consumption update module serves as the sole writer for the current anomaly check item set; other modules only read snapshots of the set to maintain the consistency of the consumption state.

[0255] Optionally, the system can categorize the execution status of processing modules into pending, executing, completed, and incomplete. When the factor extraction module or condition matching module returns an incomplete status, the consumption update module refuses to submit the set of participating item indexes; only after the consumption update is completed will the information output module include the corresponding module's processing record in the deduplication input. Through these status constraints, the system can prevent erroneous intermediate results from entering the structured inspection information and supports retries for incomplete modules based on the processing task identifier.

[0256] In an exemplary deployment, the data acquisition module and the permission resolution module reside in the data access service, the factor extraction module, the condition matching module, and the consumption update module reside in the rule execution service, and the information output module resides in the report data service. The data access service generates input snapshots and configuration snapshots for task T001, the rule execution service sequentially generates module processing records and consumption records, and the report data service performs deduplication and outputs structured inspection information. Each service only exchanges data objects related to the corresponding module's function, thus enabling independent expansion of processing capabilities without altering the chained processing logic.

[0257] In the above implementation methods, the institutional configuration filtering, chained factor extraction, two-dimensional condition group matching, consumption update, and record deduplication are sequentially linked according to the data flow. The permission parsing module first determines the condition configuration and module execution sequence that can enter this task; the factor extraction module forms two types of factors using a dynamic anomaly set and a full dictionary; the condition matching module generates module processing records and participation item indexes; the consumption update module changes the main trigger range visible to subsequent modules; and the information output module compresses the records and forms structured inspection information. The output of the previous mechanism constitutes the explicit input of the next mechanism, thus enabling a traceable and verifiable report data processing closed loop to be completed without generating direct health conclusions for specific target objects.

Claims

1. A chain-based collaborative processing method for health examination data, characterized in that, include: Get the set of anomaly check items, the dictionary of full check results, the priority queue of processing modules, and the set of condition configurations; Based on the target organization identifier, organization processing level, applicable object rules, and information processing level, available conditions are filtered and configured, and a module execution sequence is formed according to the priority queue of the processing modules. The set of anomaly checks is used as the current set of anomaly checks. The processing module is called according to the module execution sequence. The current factor set is extracted from the current set of anomaly checks, and the missing checks are read from the full check result dictionary to obtain the supplementary factor set. Based on the multiple matching conditions included in the available conditions configuration, a combined matching is performed on the current factor set and the supplementary factor set. When the multiple matching conditions are met, it is determined that the available conditions configuration is triggered, and the module processing record and the set of participating item indexes are obtained. When the module processes a record that meets the output conditions, it removes consumed check items according to the participating item index set. The consumed check items are the abnormal check items that actually participated in the combined matching and were removed. The module also updates the consumption status of the cross-module auxiliary attributes in the full check result dictionary, changing the consumption status from unconsumed to consumed. The module processes the records by sorting and deduplication, and factor inclusion deduplication, and outputs structured inspection information.

2. The method according to claim 1, characterized in that, The acquisition of the anomaly check item set, the full check result dictionary, the processing module priority queue, and the condition configuration set includes: The original inspection results were normalized in terms of name and unit according to the standard identification of the inspection items; The inspection results that exceed the corresponding reference range are encapsulated as anomaly check items and written into the anomaly check item set in the original input order; All inspection results that have completed name normalization and unit normalization are written into the full inspection result dictionary with the inspection item standard identifier as the key. For each anomaly check item, save the set index, check item standard identifier, check result value, reference range, and anomaly direction; Processing modules whose processing scopes overlap are written into the processing module priority queue according to the preset business priority; Write the condition identifier, factor field, two-dimensional condition group, minimum number of matching factors, applicable object rules, and information processing level into the condition configuration set.

3. The method according to claim 2, characterized in that, When the module processes a record that meets the output conditions, it removes consumed check items according to the participating item index set. These consumed check items are the abnormal check items that actually participated in the combined matching and were removed. The module also updates the consumption status of the cross-module auxiliary attributes in the full check result dictionary, changing the consumption status from unconsumed to consumed. This includes: Before invoking the current processing module, record the data version and number of elements of the current anomaly check item set; Write the index of the anomaly check item that matches the current processing module and actually participates in the combination matching into the participation item index set; The participating item index set is deduplicated and sorted in descending order to obtain a descending index sequence; Verify that the data version of the current anomaly check item set is consistent with the data version recorded before the current processing module was called; Consumed check items are removed one by one according to the descending index sequence to keep the set index of unconsumed check items valid; Generate consumption records for removed consumed items, including the consumption module identifier, consumption order, and consumption time. When the current processing module uses cross-module auxiliary attributes, the consumption status of the corresponding cross-module auxiliary attributes will be updated from unconsumed to consumed. The set of exception checks after the removal operation is determined as the current set of exception checks for the next processing module.

4. The method according to claim 1, characterized in that, The system performs combined matching on the current factor set and the supplementary factor set based on multiple matching conditions included in the available condition configuration. When the multiple matching conditions are met, the available condition configuration is triggered, resulting in a module processing record and a set of participating item indexes, including: The two-dimensional condition group configured with the available conditions is parsed into multiple outer candidate groups, and each outer candidate group is parsed into multiple inner conditions, wherein the inner conditions are the matching conditions; The inner condition is matched with the current factor set and the supplementary factor set respectively to obtain the inner condition matching result for each inner condition; When all the inner conditions in the same outer candidate group are satisfied, the corresponding outer candidate group is marked as satisfied, and the outer candidate group marking result is obtained. When the outer candidate group marking result of at least one outer candidate group is satisfied, the two-dimensional condition group is marked as satisfied, and the available condition configuration is determined to be triggered; The participation item index set is formed by summing the indexes of the outlier check items that are marked as satisfying the outer candidate group and originate from the current factor set. The module processing record is generated based on the labeling results of the two-dimensional condition group.

5. The method according to claim 4, characterized in that, The step of matching the inner condition with the current factor set and the supplementary factor set respectively includes: Identify the variable reference markers in the inner conditions, and extract the referenced check item standard identifiers from the variable reference markers; The inspection result value, unit of measurement, and reference range are read from the full inspection result dictionary using the referenced inspection item standard identifier as the key. The inspection result value is converted to a standardized inspection result value by converting the standard unit of measurement corresponding to the reference range. The upper limit, lower limit, and boundary inclusion identifier of the reference range are analyzed. Based on the boundary inclusion identifier, the standardized check result value is compared with the upper limit value and the lower limit value respectively to obtain the Boolean matching value of the normal range reference condition; For inner conditions that do not contain the variable reference marker, the corresponding condition identifier is compared with the abnormal direction identifier in the current factor set to obtain the Boolean matching value of the abnormal condition; Write the Boolean match value of the normal range reference condition and the Boolean match value of the abnormal condition into the condition state vector of the corresponding outer candidate group.

6. The method according to claim 1, characterized in that, The process of performing combined matching on the current factor set and the supplementary factor set based on multiple matching conditions included in the available condition configuration, determining that the available condition configuration is triggered when the multiple matching conditions are met, and obtaining the module processing record and the participating item index set, further includes: The standard identifiers of the inspection items in the current factor set are deduplicated to obtain the number of different factors; The number of different factors is compared with the minimum number of matching factors configured by the available conditions. When the number of different factors reaches the minimum number of matching factors, the factor number matching status is marked as satisfied, and the available conditions configuration is marked as triggered. When the factor quantity matching status is not satisfied, the two-dimensional condition group of the available condition configuration is parsed into multiple outer candidate groups, and each outer candidate group is parsed into multiple inner conditions. The inner conditions are matched with the current factor set and the supplementary factor set respectively. The normal range reference condition is used as the inner condition for matching. When all the inner conditions in the same outer candidate group are satisfied, the corresponding outer candidate group is marked as satisfied. When at least one outer candidate group is marked as satisfied, the two-dimensional condition group is marked as satisfied, and the available condition configuration is marked as triggered. When the factor quantity matching status is marked as satisfied or the two-dimensional condition group is marked as satisfied, the available condition configuration is determined to be triggered, and the condition identifier, factor source and processing sorting level are recorded to form the module processing record; When the factor quantity matching state is marked as satisfied, only the check item indexes that originate from the current factor set and participate in the counting of different factor quantities are written into the participation item index set; when the two-dimensional condition group is marked as satisfied, only the check item indexes that originate from the current factor set and actually participate in the matching of the two-dimensional condition group are written into the participation item index set.

7. The method according to claim 1, characterized in that, The configuration of available conditions for filtering based on target organization identifier, organization processing level, applicable object rules, and information processing level, and the formation of a module execution sequence according to the priority queue of the processing modules, includes: The applicable rules are parsed into a universally applicable mode, an exclusion list mode, or a specified list mode. In the all-applicable mode, the object matching status corresponding to the target organization identifier is set to allowed, and a first sorting weight is assigned; In the exclusion list mode, when the target organization identifier does not belong to the exclusion identifier set, the object matching status is set to allowed; In the specified list mode, when the target organization identifier belongs to the specified identifier set, the object matching status is set to allowed, and a second ranking weight greater than the first ranking weight is assigned. Input the organization processing level and the information processing level into a preset level permission matrix to obtain the level matching status; The conditional configuration in which both the object matching status and the level matching status are allowed is determined as the available conditional configuration; Matching conflicts between multiple applicable object rules are resolved in the order that the second sorting weight takes precedence over the first sorting weight.

8. The method according to claim 1, characterized in that, The process involves using the set of anomaly checks as the current set of anomaly checks, calling the processing module according to the module execution sequence, extracting the current factor set from the current set of anomaly checks, and reading missing checks from the full check result dictionary to obtain a supplementary factor set, including: In the first processing round, the dedicated processing module and the configuration processing module are invoked according to the module execution sequence. The anomaly check items matched from the current anomaly check item set are marked as primary triggering factors and written into the current factor set; When the required check item for the available condition configuration does not exist in the current factor set, the corresponding check item is read from the full check result dictionary and marked as a reference factor to form the supplementary factor set; Prohibit check item consumption operations from being triggered solely by a set of supplementary factors marked as reference factors; After the first processing round ends, a second processing round is executed for the available condition configurations that have not generated module processing records; In the second processing round, the two-dimensional condition group containing the normal range reference conditions is parsed based on the dictionary of the full check results, and the newly generated module processing record is added to the candidate record set.

9. The method according to claim 1, characterized in that, The module processes records by performing sorting, deduplication, and factor inclusion deduplication, and outputs structured inspection information, including: Sort the processing records of the multiple modules according to the result identifier and processing sorting level; For module processing records with the same result identifier, retain the module processing records whose processing sorting level meets the preset priority conditions; Construct a corresponding set of factor identifiers from the standard identifiers of the inspection items referenced in each retained module processing record; When the first factor identifier set is included in the second factor identifier set and the corresponding module processing record satisfies the preset substitution relationship, the module processing record corresponding to the first factor identifier set is removed. The result identifiers, evidence item identifiers, matching paths, consumption records, and applicable permissions from the remaining module processing records are encapsulated into structured data objects; The structured data objects are sorted according to the priority queue of the processing module to obtain the structured inspection information; The structured examination information is sent to the health examination report generation interface, which then generates editable report prompts.

10. A chain-based collaborative processing system for health examination data, characterized in that, include: The data acquisition module is used to acquire the set of anomaly check items, the dictionary of full check results, the priority queue of processing modules, and the set of condition configurations. The permission resolution module is used to filter available condition configurations based on the target organization identifier, organization processing level, applicable object rules, and information processing level, and form a module execution sequence according to the priority queue of the processing modules. The factor extraction module is used to take the set of anomaly checks as the current set of anomaly checks, call the processing module according to the module execution sequence, extract the current factor set from the current set of anomaly checks, and read the missing checks from the full check result dictionary to obtain the supplementary factor set. The condition matching module is used to perform combined matching on the current factor set and the supplementary factor set based on multiple matching conditions included in the available condition configuration, and to determine that the available condition configuration is triggered when the multiple matching conditions are met, so as to obtain the module processing record and the participating item index set. The consumption update module is used to remove consumed check items according to the participation item index set when the processed record of the module meets the output conditions. The consumed check items are the abnormal check items that actually participated in the combined matching and were removed. The module also updates the consumption status of the cross-module auxiliary attribute in the full check result dictionary, and the consumption status is updated from non-consumed to consumed. The information output module is used to perform sorting and deduplication and factor inclusion deduplication on the records processed by the module, and output structured inspection information.

Images