Dicom image routing method and system based on port multiplexing

Abstract

The application relates to the technical field of data transmission control, and discloses a DICOM image routing method and system based on port multiplexing. By binding a shared DICOM port and listening to a connection request, multiple backend DICOM receiving nodes can share a unified access port, a command set and a data set are selectively parsed according to a command type, a structured session context is constructed, routing judgment is upgraded from a traditional coarse-grained mode based on a source IP and a source port to a fine-grained mode based on protocol semantics, a routing decision module adopts a multi-field rule matching and index acceleration mechanism, can improve rule screening efficiency and reduce matching delay in a scene with more concurrent connections, and a routing optimization module dynamically evaluates in combination with running state information and transmission performance information of candidate target DICOM receiving nodes, so that a communication bottleneck can be avoided, in which a fixed transmission path is long-term concentrated on part of nodes.

Description

Technical Field

[0001] This invention relates to the field of data transmission control technology, and more specifically, to a DICOM image routing method and system based on port multiplexing. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] In hospital information systems, medical imaging equipment typically transmits image data to backend systems via the DICOM protocol. Existing image transmission methods mainly include a multi-port, multi-service instance mode and a static relay gateway mode. The former usually opens independent ports and deploys corresponding service instances for each target system, while the latter forwards data through a unified gateway according to preset rules. With the continuous increase in hospital backend systems and the growing complexity of business scenarios, these methods are increasingly unable to meet the requirements of port resource management, ease of device access, routing flexibility, and unified operation and maintenance.

[0004] The main technical problems with existing technologies are as follows: The multi-port, multi-service instance mode uses a deployment method where one target system corresponds to one port or one service instance. Adding a backend system typically requires adding a port, adjusting firewall policies, and modifying image device configurations, resulting in high deployment, maintenance, and upgrade costs. The static relay gateway mode primarily relies on limited network layer information such as source IP and source port for traffic distribution. The routing rules are static and coarse-grained. When the same device needs to distribute images to different target systems for different scenarios, it still requires adding ports or service instances, limiting the overall system flexibility, scalability, and dynamic optimization capabilities. Furthermore, in scenarios with a large number of concurrent image devices, fixed transmission paths can easily create communication bottlenecks at some target nodes, leading to increased network latency, limited service response, and the inability to automatically recover once image transmission is interrupted, thus affecting the continuity and stability of system access. Summary of the Invention

[0005] To address the aforementioned problems, this invention proposes a DICOM image routing method and system based on port multiplexing, which solves the problems of cumbersome manual configuration, inability to self-heal transmission interruptions, and lack of fine-grained dynamic routing in the prior art.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: The first aspect of this invention provides a DICOM image routing system based on port multiplexing, comprising: The DICOM port listening module is configured to bind to a shared DICOM port and listen for TCP connection requests, receiving DICOM protocol data. The DICOM protocol parsing module is configured to selectively parse command sets and datasets in DICOM protocol data according to command type, and extract route-related tags according to preset parsing rules to generate a structured session context based on semantic information. The routing decision module is configured to perform matching processing on the generated structured session context based on multi-field rule matching and index acceleration mechanism, and filter candidate target DICOM receiving nodes; The routing optimization module is configured to obtain the running status information and transmission performance information of the candidate target DICOM receiving nodes, and dynamically evaluate the candidate target DICOM receiving nodes based on the routing-related fields in the session context of the current DICOM association request to determine the final target DICOM receiving node.

[0007] Further technical solutions include preset parsing rules such as tag whitelist parsing rules, on-demand parsing rules based on command type, and key field parsing rules; Alternatively, the DICOM protocol parsing module includes a session identification and first packet multiplexing unit, a PDU parsing unit, a command set on-demand parsing unit, a whitelist tag parsing unit, a private tag parsing unit, and a session context construction unit; The session identification and first packet multiplexing unit is used to identify DICOM-related requests in the first data packet based on DICOM protocol data. The PDU parsing unit is used to parse the DICOM protocol data and extract the calling AE header and the called AE header for the identified DICOM association request; The whitelist label parsing unit is configured with preset label whitelist parsing rules. For the dataset part that needs to be parsed, it extracts routing-related standard labels according to the preset label whitelist. The private tag parsing unit is configured to selectively parse private tags in the dataset of DICOM protocol data based on a pre-configured private tag dictionary; The session context building unit is configured to generate the corresponding session context based on the parsing results, including the title of the calling AE, the title of the called AE, the route-related standard tags, and the route-related fields in the private tags.

[0008] A further technical solution, the DICOM protocol parsing module, also includes a fault-tolerant processing unit, which is configured to retain the basic routing fields when some fields are missing.

[0009] Further technical solutions, including a routing decision module, include: The index filtering unit is used to filter a subset of candidate rules from the configured set of routing rules based on the structured session context and a preset high-discrimination field. The rule filtering unit is configured to perform conditional filtering on a subset of candidate rules by joint matching of multiple fields, filter out the hit routing rules that meet the matching conditions and sort them, select one or more routing rules with the highest sorting results as routing decisions, and filter out one or more candidate target DICOM receiving nodes. The cache hit unit is configured to maintain a cache of routing results for repeated session feature combinations within a preset time range, and to directly output the corresponding routing results when the cache usage conditions are met, so as to determine the candidate target DICOM receiving node.

[0010] Further technical solutions include obtaining one or more of the following in the routing optimization module: average transmission time, failure rate, current load, and historical success rate; Routing-related fields include one or more of the following: the title of the calling AE, the title of the called AE, Modality, inspection location, Study description, Series description, device manufacturer, device model, and key private tag values.

[0011] Further technical solutions also include a routing rule management module, which provides a set of routing rules to the routing decision module and is configured to store and maintain the set of routing rules; Alternatively, it may also include a forwarding and storage module, connected to the routing optimization module, configured to establish a DICOM association based on the final target DICOM receiving node, and forward the corresponding DICOM image data to the final target DICOM receiving node for storage and / or processing.

[0012] A further technical solution also includes a forwarding and storage module, connected to the routing optimization module, configured to establish a DICOM association based on the final target DICOM receiving node, and forward the corresponding DICOM image data to the final target DICOM receiving node for storage and / or processing; Configured to perform the following procedures: Retrieve the lifecycle status information of the recorded DICOM session; Monitor the connection and transmission status during the DICOM image data forwarding process, and monitor the connection status, response status, data transmission progress and / or timeout status in real time to determine whether the current forwarding process is normal; When the target node is detected to be unreachable, transmission fails, or timeout occurs, a retry will be performed according to the preset retry policy. If a retry fails, switch to a backup target node corresponding to the current business scenario to continue forwarding.

[0013] A second aspect of the present invention provides a DICOM image routing method based on port multiplexing, comprising the following steps: Listen for TCP connection requests based on the configured shared DICOM port and receive DICOM protocol data; For the acquired DICOM protocol data, the command sets and datasets in the DICOM protocol data are selectively parsed according to command type, and routing-related tags are extracted according to preset parsing rules to generate a structured session context based on semantic information. Based on multi-field rule matching and index acceleration mechanisms, the generated structured session context is matched and filtered to select candidate target DICOM receiving nodes; Obtain the running status and transmission performance information of the candidate target DICOM receiving nodes. Based on the routing-related fields in the session context of the current DICOM association request, dynamically evaluate the candidate target DICOM receiving nodes to determine the final target DICOM receiving node.

[0014] A further technical solution involves selectively parsing command sets and datasets in the DICOM protocol data according to command type, extracting routing-related tags based on preset parsing rules, and generating a structured session context based on semantic information. This method includes the following steps: The first data packet based on the DICOM protocol data identifies the DICOM association request; For DICOM association requests that have passed the identification process, parse the DICOM protocol data and extract the title of the calling AE and the title of the called AE. Preset command type on-demand parsing rules, selectively parse command sets and corresponding datasets based on DIMSE command type; The preset label whitelist parsing rules are used to extract routing-related standard labels for the dataset parts that need to be parsed, according to the preset label whitelist. Based on a pre-configured private tag dictionary, private tags in the dataset of DICOM protocol data are selectively parsed; Based on the parsing results, the corresponding session context will be generated from the calling AE title, the called AE title, the routing-related standard tags, and the routing-related fields in the private tags.

[0015] A further technical solution, based on multi-field rule matching and index acceleration mechanisms, involves matching the generated structured session context and filtering candidate target DICOM receiving nodes, including the following steps: Based on the structured session context and a preset high-discrimination field, a subset of candidate rules is selected from the configured set of routing rules. The candidate rule subset is subjected to conditional filtering by multi-field joint matching. The matching routing rules that meet the matching conditions are selected and sorted. One or more routing rules with the highest sorting results are selected as routing decisions, and one or more candidate target DICOM receiving nodes are selected. A routing result cache is maintained for recurring session feature combinations within a preset time range, and the corresponding routing result is directly output when the cache usage conditions are met to determine the candidate target DICOM receiving node; the session feature combination includes one or more of the following: calling AE header, called AE header, Modality, and source IP network segment.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention configures the DICOM port listening module to bind to a shared DICOM port and listen for TCP connection requests, enabling multiple backend DICOM receiving nodes to share a unified access port. This avoids adding new ports or deploying independent service instances for each target system, reducing deployment and maintenance costs. The DICOM protocol parsing module selectively parses command sets and datasets according to command type, constructing a structured session context. This improves routing decision-making from the traditional coarse-grained approach based on source IP and source port to a fine-grained approach based on protocol semantics, thereby enhancing the system's adaptability to different examination types, clinical scenarios, and backend systems. The routing decision module employs multi-field rule matching and index acceleration mechanisms, improving rule filtering efficiency and reducing matching latency in scenarios with many concurrent connections. The routing optimization module dynamically evaluates the candidate target DICOM receiving nodes based on their running status and transmission performance information, preventing fixed transmission paths from concentrating on a few nodes and forming communication bottlenecks. This reduces the probability of increased latency, limited response, and image transmission interruptions, thereby improving the system's access continuity, transmission stability, and flexibility.

[0017] The advantages of the present invention, as well as its additional advantages, will be described in detail in the following specific embodiments. Attached Figure Description

[0018] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute a limitation thereof.

[0019] Figure 1 This is a block diagram of a DICOM image routing system based on port multiplexing according to Embodiment 1 of the present invention; Figure 2 This is a flowchart of the DICOM image routing method based on port multiplexing according to Embodiment 2 of the present invention. Detailed Implementation

[0020] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0021] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0022] It should be noted that the terminology used herein is for describing particular embodiments only and is not intended to limit the exemplary embodiments of the present invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof. It should be noted that, without conflict, the various embodiments and features within those embodiments can be combined with each other. The embodiments will now be described in detail with reference to the accompanying drawings.

[0023] Explanation of technical terms DICOM (Digital Imaging and Communications in Medicine) is a standard protocol for medical imaging data. It is used to standardize the format definition, storage method and network transmission method of medical images and related information, thereby enabling data interconnection between imaging equipment and back-end systems. The imaging equipment can be CT, MR, DR, DSA, ultrasound and other equipment, and the back-end systems can be medical information systems such as PACS system, imaging cloud platform, AI-assisted diagnostic platform and other medical information systems.

[0024] Example 1 In one or more of the technical solutions disclosed in the embodiments, such as Figure 1 As shown, a DICOM image routing system based on port multiplexing is disclosed. The system connects multiple medical imaging devices to a single shared DICOM port or a few shared DICOM ports, including: a DICOM port monitoring module, a DICOM protocol parsing module, and a routing decision module. The DICOM port listening module is configured to bind to a shared DICOM port and listen for TCP connection requests, receiving DICOM protocol data. The DICOM protocol parsing module, connected to the DICOM port listening module, is configured to selectively parse command sets and datasets in the DICOM protocol data according to command type, and extract route-related tags according to preset parsing rules to generate a structured session context based on semantic information. The routing decision module, connected to the DICOM protocol parsing module, is configured to receive the structured session context output by the DICOM protocol parsing module, and perform matching processing on the generated structured session context based on multi-field rule matching and index acceleration mechanism to filter candidate target DICOM receiving nodes; The routing optimization module, connected to the routing decision module, is configured to obtain the running status information and transmission performance information of the candidate target DICOM receiving node, and dynamically evaluate the candidate target DICOM receiving node based on the routing-related fields in the session context of the current DICOM association request to determine the final target DICOM receiving node. In the above implementation, by configuring the DICOM port listening module to bind to a shared DICOM port and listen for TCP connection requests, multiple backend DICOM receiving nodes can share a unified access port. This avoids adding new ports or deploying independent service instances for each new target system, reducing deployment and maintenance costs associated with hospital network port management, firewall policy adjustments, and redundant configuration of imaging equipment. Simultaneously, the DICOM protocol parsing module selectively parses command sets and datasets according to command type, extracting semantic information such as patient, examination, equipment, and service identifiers from the DICOM protocol data. This allows for the construction of a structured session context, enabling routing decisions to shift from traditional methods based on... The coarse-grained approach of source IP and source port is upgraded to a fine-grained approach based on protocol semantics, thereby enhancing the system's adaptability to different examination types, clinical scenarios, and backend systems. Furthermore, the routing decision module adopts a multi-field rule matching and index acceleration mechanism, which can improve rule filtering efficiency and reduce matching latency in scenarios with many concurrent connections. The routing optimization module performs dynamic evaluation by combining the running status information and transmission performance information of candidate target DICOM receiving nodes, which can avoid the long-term concentration of fixed transmission paths on a few nodes and the formation of communication bottlenecks, reduce the probability of increased latency, limited response, and image transmission interruption, thereby improving the continuity, stability, and flexibility of data transmission.

[0025] Furthermore, the port-multiplexed DICOM image routing system also includes a forwarding and storage module connected to the routing optimization module, which establishes a DICOM association based on the final target DICOM receiving node and forwards the corresponding DICOM image data to the final target DICOM receiving node for storage and / or processing; Furthermore, the port-multiplexed DICOM image routing system also includes a session management and exception handling module, which monitors the forwarding status of DICOM image data and performs retries and / or switches to a backup target node when the target node is detected to be unreachable, transmission fails, or times out.

[0026] In some embodiments, in the field of medical image transmission, the DICOM port listening module can receive TCP connection requests from multiple imaging devices; it listens for TCP connection requests from multiple medical imaging devices on a preset shared DICOM port to form PDU data; wherein, all of the multiple medical imaging devices configure the same proxy system address and the shared DICOM port as DICOM transmission targets; The DICOM port listening module is used to receive TCP connections initiated by various devices, identify and receive DICOM association requests, establish an instant session context, and hand over the established connections to subsequent modules for processing.

[0027] Optionally, a shared DICOM port can be used, such as the standard port 104, port 11112, or a dynamically allocated high-order TCP port; it can continuously listen for TCP connection requests from various devices in a multi-threaded or asynchronous I / O manner. In some embodiments, the preset parsing rules include tag whitelist parsing rules, on-demand parsing rules based on command type, and key field parsing rules.

[0028] Optionally, the DICOM protocol parsing module includes a session identification and first packet multiplexing unit, a PDU parsing unit, a command set on-demand parsing unit, a whitelist label parsing unit, a private label parsing unit, a session context construction unit, and a fault tolerance processing unit; The session identification and first packet multiplexing unit is used to identify DICOM-related requests in the first data packet based on DICOM protocol data. Specifically, the session identification and first packet multiplexing unit analyzes the PDU type and structure of the first data packet in the DICOM protocol data to identify whether the connection request received through the shared port is a DICOM association request. If it is a DICOM association request, the identification is deemed successful, and the data is transmitted to the PDU parsing unit for subsequent parsing. After receiving a TCP connection from a device on the shared port, the DICOM protocol parsing module reuses the first data packet that the DICOM port listening module has already pre-read, analyzes the PDU type and structure of the data, and identifies whether the connection is a DICOM associated request. This allows for the initial determination of the protocol type without repeatedly reading network traffic and avoids performing invalid parsing on non-DICOM connections.

[0029] The PDU parsing unit is used to parse DICOM protocol data and extract the calling AE header and the called AE header for a DICOM association request that has been successfully identified. The command set on-demand parsing unit is configured to parse command sets and corresponding datasets selectively based on preset command types and on-demand parsing rules. Specifically, for the identified DICOM association request, the command set structure is parsed to determine whether to continue parsing the corresponding data set. Data set parsing is triggered on demand to avoid processing unnecessary data and improve processing performance. After the session is established, the command set is parsed according to the received DIMSE command type (e.g., C-STORE, C-MOVE, C-FIND, etc.) to obtain information such as command type, SOP Class UID, and message ID. Based on this, it is determined whether further parsing of the corresponding data set is needed, thereby constructing a command type-driven on-demand parsing mechanism.

[0030] The whitelist label parsing unit is configured with preset label whitelist parsing rules. For the dataset part that needs to be parsed, it extracts routing-related standard labels according to the preset label whitelist. Specifically, for the portion of the dataset that is determined to require parsing, only key standard labels related to routing decisions are parsed according to a pre-configured label whitelist. The key standard labels in the label whitelist may include, but are not limited to: Modality tags, such as (0008,0060); Study and Series related identifiers, such as Study Instance UID, Series Instance UID, etc.; Standard labels related to departments, examination types, and business lines, such as Body Part Examined, Study Description, Series Description, etc. This embodiment uses a preset tag whitelist parsing rule to avoid full traversal parsing of pixel data and tags unrelated to routing, thereby extracting the key data required for routing with less parsing overhead and improving the lightweight nature and processing performance of DICOM data extraction.

[0031] The private tag parsing unit is configured to selectively parse private tags in the dataset of DICOM protocol data based on a pre-configured private tag dictionary; Specifically, for the Private Tag in the dataset, the Private Creator tag is first parsed to establish the correspondence between the private tag and the device manufacturer or application module; then, based on the configurable private tag dictionary, one or more private tags involving business line identifiers, tenant information, AI processing result flags and workflow identifiers are parsed, and the parsing results are normalized to support cross-vendor and cross-scenario extended deployment. Optionally, the private tag dictionary can be configured to expand dynamically without modifying the core system code.

[0032] The session context building unit is configured to generate the corresponding session context based on the parsing results, including the title of the calling AE, the title of the called AE, the route-related standard tags, and the route-related fields in the private tags. Specifically, the parsing results of the PDU parsing unit, the whitelist tag parsing unit, and the private tag parsing unit are uniformly encapsulated into a DICOM session context object. The session context object includes the calling AE title, the called AE title, Modality, Study identifier, Series identifier, key standard tags, and selected private tags, which are used to provide a unified structured input to the downstream routing decision module.

[0033] The fault-tolerant processing unit is configured to retain basic routing fields to maintain routing decisions when some fields are missing. Specifically, when some fields in the standard label or private label fail to be parsed, are missing, or are unavailable, a hierarchical fault-tolerant strategy is adopted to retain at least the basic fields in the calling AE title, the called AE title, and Modality, so as to ensure that the routing decision module can still complete basic routing judgment based on the retained fields, thereby enhancing the robustness and availability of the system in scenarios with incomplete parsing.

[0034] In some embodiments, the routing decision module is configured to perform matching processing on the generated structured session context based on multi-field rule matching and index acceleration mechanism, including an index filtering unit, a rule filtering unit, a cache hit unit, and a routing result output unit. The index filtering unit is used to filter a subset of candidate rules from the configured set of routing rules based on the structured session context and a preset high-discrimination field. Optionally, the index filtering unit can utilize high-discrimination fields such as the called AE title and Modality to build a multi-level index structure and quickly filter to obtain a subset of candidate rules; The multi-level index structure can adopt a two-level Map or a tree index; the multi-level index structure realizes the index acceleration mechanism, which can avoid sequential traversal of all routing rules, thereby narrowing the subsequent matching range and improving the rule matching efficiency.

[0035] The rule filtering unit is configured to perform conditional filtering on a subset of candidate rules by joint matching of multiple fields, filter out the hit routing rules that meet the matching conditions and sort them, select one or more routing rules with the highest sorting results as routing decisions, and filter out one or more candidate target DICOM receiving nodes. Optionally, in the candidate rule subset, the conditional filtering for multi-field joint matching can be filtered according to conditions such as source IP address, time period, Study / Series identifier, private tag value, etc., and the hit routing rules are sorted based on preset rule priority and the number of matching fields. Furthermore, when multiple matching rules meet the conditions simultaneously, the rule with higher priority is adopted first; when the priorities are the same, the rule with more matching fields is adopted first, thereby improving the controllability and accuracy of routing decisions.

[0036] This embodiment uses multi-field joint filtering, which improves routing accuracy and adaptability in different business scenarios compared to static traffic distribution based on a single field.

[0037] The cache hit unit is configured to maintain a route result cache for repeated session feature combinations within a preset time range, and directly output the corresponding route result when the cache usage conditions are met, so as to determine the candidate target DICOM receiving node. Specifically, for multiple DICOM association requests with the same combination of session characteristics within a short period of time, the routing result caching unit stores the routing result corresponding to the combination of session characteristics; wherein, the combination of session characteristics may include one or more of the following: calling AE title, called AE title, Modality, and source IP network segment.

[0038] When the session feature combination corresponding to a subsequent DICOM association request is consistent with the saved session feature combination, and the relevant parsing fields have not changed, the routing result caching unit directly outputs the saved routing result without having to perform candidate rule filtering, rule sorting and processing again, thereby reducing redundant calculations and improving routing processing efficiency in high-concurrency scenarios.

[0039] Compared to traditional static routing methods that traverse the rule list sequentially and match based solely on the source IP address or a single AE title, the routing decision module in this embodiment utilizes multi-field indexes to quickly locate candidate rules and combines them with priorities for fine-grained matching. This enables fine-grained routing based on semantic information such as the calling AE, the called AE, and Modality. Through multi-field joint matching, index-accelerated filtering, and a cache reuse mechanism, the accuracy of routing decisions is ensured while improving rule matching efficiency and system performance.

[0040] In some embodiments, the routing optimization module, connected to the routing decision module, is configured to obtain the running status information and transmission performance information of the candidate target DICOM receiving node, and dynamically evaluate the candidate target DICOM receiving node based on the routing-related fields in the session context of the current DICOM association request, and determine the final target DICOM receiving node. Optionally, the running status information and transmission performance information obtained in the routing optimization module may include one or more of the following: average transmission time, failure rate, current load, and historical success rate.

[0041] Optionally, the routing-related fields include one or more of the following: the title of the calling AE, the title of the called AE, Modality, inspection location, Study description, Series description, device manufacturer, device model, and key private tag values.

[0042] Optionally, the intelligent routing optimization module can score or predict the probability of candidate target DICOM receiving nodes based on a machine learning model, and output the final target DICOM receiving node under the premise of meeting preset constraints.

[0043] Optionally, the machine learning model may be one or more of the following: gradient boosting tree model, random forest model, and lightweight neural network model; Under the premise of meeting preset constraints, the final target DICOM receiving node is output. Specifically, the candidate target DICOM receiving nodes output by the model dynamic evaluation are verified according to the hard constraints in the routing rule management module. Optionally, the hard constraints include one or more of security policies, compliance requirements, and mandatory routing rules. Only when the candidate target DICOM receiving node meets the hard constraints is it included in the selection range of the final target DICOM receiving node, so as to ensure that the dynamic routing optimization results meet business rules and compliance requirements.

[0044] When the dynamic evaluation model experiences prediction timeout, model anomaly, output confidence below a preset threshold, or other unavailability, the rollback processing unit stops using the dynamic evaluation results and rolls back to the routing results from the routing decision module. This ensures that the system can continue to complete routing processing even in abnormal intelligent optimization scenarios.

[0045] It is feasible to periodically retrain or update the parameters of the machine learning model offline by recording each routing decision and subsequent transmission results, so that the routing strategy has the ability to evolve adaptively.

[0046] To address the problem that existing technologies rely solely on static traffic distribution based on network layer information such as source IP address and source port, resulting in coarse-grained routing that struggles to adapt to the refined routing needs of the same medical imaging equipment across different business scenarios, this embodiment employs a combined technical solution of DICOM semantic-level parsing, multi-field joint matching, index acceleration, result caching, and intelligent routing optimization. By parsing routing-related fields in DICOM associated requests, command sets, and datasets, it extracts business semantic information such as the calling AE title, the called AE title, Modality, Study / Series identifier, examination site, description information, and private tags. This information is then combined with conditions such as source IP address and time period for joint matching. Furthermore, multi-level indexing narrows the range of candidate rules, and routing result caching reduces redundant calculations. Finally, it dynamically evaluates and optimizes candidate target nodes by incorporating historical transmission time, failure rate, and current load metrics. Through these technical means, the system can not only distribute traffic based on device origin but also achieve fine-grained dynamic routing based on specific business semantics, improving matching efficiency, system throughput, and adaptability to complex hospital information scenarios while ensuring routing accuracy.

[0047] In some embodiments, the port-multiplexing-based DICOM image routing system further includes a routing rule management module, which is used to provide a set of routing rules to the routing decision module and is configured to store and maintain the set of routing rules. One possible implementation is that the routing rules maintained in the routing rule management module can be configured in a structured manner. Each routing rule includes at least the matching conditions, the target DICOM receiving node, and priority control information for rule conflict handling; Matching conditions are used to describe under what conditions the current DICOM association request should match this rule. Matching conditions can be configured based on a single field or a combination of multiple fields. Optional fields that can be used as matching criteria include, but are not limited to, the title of the called AE, the title of the called AE, Modality, source IP address, source IP network segment, time period, Study identifier, Series identifier, standard tag value, and private tag value, etc. By using the above fields individually or in combination, the system can distinguish DICOM association requests according to device source, business type, time scenario, inspection category, or specific business identifier.

[0048] The target DICOM receiving node indicates the forwarding destination of the image after a rule is matched. Different rules can correspond to different target nodes, which can be local PACS, regional PACS, cloud platforms, AI-assisted diagnostic platforms, specialty business systems, etc. Thus, the system can automatically send images to the corresponding backend systems based on the source, type, or business attributes of different examination data.

[0049] To ensure a definite routing result even when multiple rules simultaneously meet the conditions, priority and weight information are configured in the routing rules. Priority is used to select the order of different rules, while weight is used for further decision-making when priorities are the same or close. With the help of priority and weight configuration, the system can implement stable and controllable rule-hitting logic in complex business environments, avoiding uncertain routing results caused by conflicts between multiple rules.

[0050] In addition to basic matching and forwarding information, routing rules can also carry optional processing actions. These actions indicate whether additional processing needs to be performed on the image before forwarding, such as compression, anonymization, tag completion, or repackaging, thereby improving the system's adaptability to different target systems and different business requirements.

[0051] In terms of implementation, the routing rule management module prioritizes providing the ability to add, modify, and delete rules, and supports rule management through a configuration interface or application programming interface (API). Furthermore, this module also supports online rule updates, enabling rule changes to be completed without stopping system operation when adding backend systems, adjusting business flow, or optimizing matching strategies. This significantly reduces system maintenance costs and improves the flexibility of business deployment and adjustments.

[0052] The routing rule management module in this embodiment does not only store rules, but also serves as the rule control center for the entire DICOM image routing system. On the one hand, it supports fine-grained routing through multi-field configurable matching conditions; on the other hand, it ensures that routing decisions have scalability, controllability, and dynamic maintenance capabilities through unified management of target nodes, priorities, weights, and optional processing actions.

[0053] In some embodiments, the forwarding and storage module, connected to the routing optimization module, is configured to establish a DICOM association based on the final target DICOM receiving node and forward the corresponding DICOM image data to the final target DICOM receiving node for storage and / or processing; it can be configured to perform: Single-target forwarding mode is used to send images to only one target system. In this mode, the system sends the current DICOM image data to a single target DICOM receiving node based on the final routing result, to meet the needs of regular image archiving, storage, or single business processing. This mode is suitable for scenarios where images only need to enter a single backend system, such as sending them only to a local PACS system for unified storage management.

[0054] The multi-target forwarding mode is used to simultaneously send the same image to multiple target systems based on routing results, enabling backup and multi-service synchronization. In this mode, the system sends the same DICOM image data to multiple target DICOM receiving nodes simultaneously according to preset configurations, thereby achieving parallel distribution of image data. Through this method, the system can complete delivery to multiple systems simultaneously after a single reception, such as sending to a local PACS, a cloud archiving platform, and an AI-assisted diagnostic platform, to meet the needs of data backup, business collaboration, and parallel processing across multiple scenarios.

[0055] Furthermore, the forwarding and storage module is also configured to process the data before forwarding, such as performing necessary compression, anonymization, tag completion, or repackaging on the DICOM data before forwarding, to meet the target system's requirements for data format and privacy protection.

[0056] In some embodiments, the session management and exception handling module is used to monitor the forwarding status of DICOM image data, and to perform retries and / or switch to a backup target node when the target node is detected to be unreachable, transmission fails, or times out. Specifically, it is configured to perform the following procedures: Retrieve the lifecycle status information of the recorded DICOM session; Monitor the connection and transmission status during the DICOM image data forwarding process, and monitor the connection status, response status, data transmission progress and / or timeout status in real time to determine whether the current forwarding process is normal; When the target node is detected to be unreachable, transmission fails, or timeout occurs, a retry will be performed according to the preset retry policy. If a retry fails, the system will switch to a backup target node that corresponds to the current business scenario to continue forwarding. During the retry and / or backup target node handover process, record one or more of the following: the time of the exception, the failure data identifier, the reason for the failure, the number of retries, the target node information, and the handover result, in order to generate an exception handling record; The anomaly handling records, along with corresponding historical transmission success rates, average transmission time, failure rates, and other operational metrics, are provided to the intelligent routing optimization module for subsequent training, updating, and parameter adjustment of the routing optimization model.

[0057] To address the problem in existing technologies where the lack of an automatic recovery mechanism when the target node is unreachable, transmission fails, or the link times out, leading to interruptions in DICOM image transmission and poor system continuity and stability, this embodiment establishes a session management and anomaly handling module. This module constructs a closed-loop processing mechanism including session state tracking, anomaly detection, automatic retries, backup node switching, and log recording. By continuously maintaining DICOM session lifecycle information, real-time monitoring of the transmission process is achieved. When an anomaly is detected, a preset retry strategy is prioritized to restore transmission. If a retry fails, the system automatically switches to a backup target node to continue image transmission. Simultaneously, the abnormal process and transmission indicators are recorded and fed back to the system. Through these technical means, the system possesses adaptive recovery capabilities for transmission anomalies, avoiding transmission interruptions due to single link failures, thereby significantly improving the continuity, stability, and overall reliability of DICOM image transmission.

[0058] Furthermore, the port-multiplexed DICOM image routing system also includes a security and auditing module, as well as a configuration and monitoring module; The security and auditing module is configured to perform authentication, access control, and log auditing on access, routing, and forwarding processes in a port-multiplexed DICOM image routing system. Specifically, the security and auditing module is configured to authenticate access devices and target DICOM receiving nodes to verify the legitimacy of the access party and the receiver, and to perform access control on different devices, nodes or service requests according to preset security policies to restrict unauthorized access or illegal calls.

[0059] Furthermore, the security and auditing module is configured to record audit information for each DICOM session during processing. This audit information includes one or more of the following: session source, target node, routing results, data transmission status, and anomaly information. Through continuous recording of this information, the system can trace and reconstruct each DICOM image routing and forwarding process.

[0060] Furthermore, the security and auditing module is configured to provide log query and log export capabilities to support subsequent auditing, statistical analysis, and troubleshooting. By retrieving, filtering, and exporting log information, centralized analysis and management of system operating status, routing execution, and abnormal events can be achieved.

[0061] The configuration and monitoring module is configured to perform centralized configuration management and operational status monitoring of the port-multiplexed DICOM image routing system.

[0062] Specifically, the configuration and monitoring module is configured to centrally configure DICOM listening ports, routing rules, private label dictionaries, and parameters related to the intelligent routing model, so as to achieve unified management of key system parameters.

[0063] Furthermore, the configuration and monitoring module is configured to monitor one or more of the following: the operating status of each module of the system, the current number of connections, route hit statistics, and error rate. The monitoring results are then visualized and / or alarmed to facilitate unified observation of the system's operating status and anomaly detection.

[0064] Furthermore, the configuration and monitoring module is also configured to support online updates of rules and configurations, so as to complete configuration adjustments without interrupting system operation, thereby reducing system maintenance costs and improving operational flexibility.

[0065] This embodiment addresses the problem in existing technologies where each addition of a backend system requires adding ports, adjusting firewall policies, and modifying the transmission configuration of medical imaging devices, leading to fragmented device access methods, complex network configurations, and high maintenance costs. This embodiment employs a unified access solution using a shared DICOM port, proxy-side semantic routing, centralized rule management, and a configurable and expandable private tag dictionary. By uniformly configuring the transmission targets of multiple medical imaging devices to the IP address of the proxy system and the shared DICOM port, image data is first aggregated to a unified entry point. The system then performs internal forwarding on the proxy side based on routing-related fields in the DICOM protocol. Simultaneously, the routing rule management module centrally maintains and updates the routing rules online, and the configurable private tag dictionary supports extended parsing for devices from different vendors and in different business scenarios. These technical means reduce the number of network ports and the frequency of firewall policy adjustments, avoiding repeated modifications to device and network configurations due to the addition of backend systems. This results in a more unified device access method, more flexible system deployment and expansion, and significantly reduces the configuration complexity and maintenance costs of the DICOM image routing system in a hospital information environment.

[0066] Example 2 Based on Embodiment 1, this embodiment provides a DICOM image routing method based on port multiplexing, such as... Figure 2 As shown, it includes the following steps: Step 1: Listen for TCP connection requests on the configured shared DICOM port and receive DICOM protocol data; Step 2: For the acquired DICOM protocol data, selectively parse the command set and dataset in the DICOM protocol data according to the command type, and extract route-related tags according to the preset parsing rules to generate a structured session context based on semantic information. Step 3: Based on multi-field rule matching and index acceleration mechanism, perform matching processing on the generated structured session context to filter candidate target DICOM receiving nodes; Step 4: Obtain the running status information and transmission performance information of the candidate target DICOM receiving node. Based on the routing-related fields in the session context of the current DICOM association request, dynamically evaluate the candidate target DICOM receiving node and determine the final target DICOM receiving node.

[0067] In this embodiment, by listening to TCP connection requests and receiving DICOM protocol data through a shared DICOM port, DICOM access from different sources can be uniformly aggregated to a single entry point, reducing the configuration complexity under multi-port access methods. By selectively parsing the command set and dataset in the DICOM protocol data according to command type and extracting routing-related tags according to preset parsing rules, a structured session context based on semantic information can be generated. This enables routing processing to be based on key information at the semantic level of the DICOM protocol, improving the targeting of routing judgments. By matching the structured session context based on multi-field rule matching and index acceleration mechanisms, the screening efficiency of candidate target DICOM receiving nodes can be improved. Furthermore, by obtaining the running status information and transmission performance information of candidate target DICOM receiving nodes and dynamically evaluating them in conjunction with the routing-related fields in the current DICOM associated request session context, the final target DICOM receiving node that is more suitable for the current request can be determined from the candidate nodes, thereby improving the flexibility and processing efficiency of the DICOM image routing method.

[0068] In step 1, TCP connection requests can come from multiple imaging devices; listen for TCP connection requests from multiple medical imaging devices on a preset shared DICOM port to form DICOM protocol data, i.e., PDU data; wherein, all of the multiple medical imaging devices configure the same proxy system address and the shared DICOM port as DICOM sending targets; Step 2 involves selectively parsing the command sets and datasets in the DICOM protocol data according to command type, extracting routing-related tags based on preset parsing rules, and generating a structured session context based on semantic information. This method includes the following steps: Step 21: Identify the DICOM association request based on the first data packet of the DICOM protocol; Step 22: For the DICOM association request that has passed the identification, parse the DICOM protocol data and extract the title of the calling AE and the title of the called AE; Step 23: Preset on-demand parsing rules for command types, and selectively parse command sets and corresponding datasets based on DIMSE command types; Step 24: Preset label whitelist parsing rules. For the dataset part that needs to be parsed, extract the routing-related standard labels according to the preset label whitelist. Step 25: Based on the pre-configured private tag dictionary, selectively parse the private tags in the dataset of the DICOM protocol data; Step 26: Based on the parsing results, generate the corresponding session context from the calling AE title, the called AE title, the routing-related standard tags, and the routing-related fields in the private tags; Step 3, based on multi-field rule matching and index acceleration mechanisms, performs matching processing on the generated structured session context to filter candidate target DICOM receiving nodes, including the following steps: Step 31: Based on the structured session context and a preset high-discrimination field, filter a subset of candidate rules from the configured set of routing rules; Step 32: Perform multi-field joint matching condition filtering on the candidate rule subset, filter the matching routing rules that meet the matching conditions and sort them, select one or more routing rules with the highest sorting results as the routing decision, and filter out one or more candidate target DICOM receiving nodes. Step 33: Maintain a routing result cache for recurring session feature combinations within a preset time range, and directly output the corresponding routing result when the cache usage conditions are met to determine the candidate target DICOM receiving node; the session feature combination includes one or more of the following: calling AE header, called AE header, Modality, and source IP network segment.

[0069] In step 4, the running status information and transmission performance information of the candidate target DICOM receiving nodes are obtained. Based on the routing-related fields in the session context of the current DICOM association request, the candidate target DICOM receiving nodes are dynamically evaluated to determine the final target DICOM receiving node. The candidate target DICOM receiving nodes can be scored or predicted probably based on a machine learning model, and the final target DICOM receiving node is output under the premise of meeting the preset constraints.

[0070] Optionally, the machine learning model may be one or more of the following: gradient boosting tree model, random forest model, and lightweight neural network model; Under the premise of meeting preset constraints, the final target DICOM receiving node is output. Specifically, the candidate target DICOM receiving nodes output by the model dynamic evaluation are verified according to the hard constraints in the routing rule management module. Optionally, the hard constraints include one or more of security policies, compliance requirements, and mandatory routing rules. Only when the candidate target DICOM receiving node meets the hard constraints is it included in the selection range of the final target DICOM receiving node, so as to ensure that the dynamic routing optimization results meet business rules and compliance requirements.

[0071] Further technical solutions also include building a routing rule database for storing and updating routing rules. Each routing rule includes at least the matching conditions, the target DICOM receiving node, and priority control information for handling rule conflicts. Matching conditions are used to describe under what conditions the current DICOM association request should match this rule. Matching conditions can be configured based on a single field or a combination of multiple fields. Optional fields that can be used as matching criteria include, but are not limited to, the title of the called AE, the title of the called AE, Modality, source IP address, source IP network segment, time period, Study identifier, Series identifier, standard tag value, and private tag value, etc. By using the above fields individually or in combination, the system can distinguish DICOM association requests according to device source, business type, time scenario, inspection category, or specific business identifier.

[0072] The target DICOM receiving node indicates the forwarding destination of the image after a rule is matched. Different rules can correspond to different target nodes, which can be local PACS, regional PACS, cloud platforms, AI-assisted diagnostic platforms, specialty business systems, etc. Thus, the system can automatically send images to the corresponding backend systems based on the source, type, or business attributes of different examination data.

[0073] Further technical solutions also include monitoring the forwarding status of DICOM image data, and performing retries and / or switching to a backup target node when the target node is detected to be unreachable, transmission fails, or times out, including the following processes: Step 51: Obtain the recorded lifecycle status information of the DICOM session; Step 52: Monitor the connection status and transmission status during the DICOM image data forwarding process. Monitor the connection status, response status, data transmission progress and / or timeout status in real time to determine whether the current forwarding process is normal. Step 53: When the target node is detected to be unreachable, transmission fails, or timeout occurs, retry is performed according to the preset retry strategy; Step 54: If a retry fails, switch to the backup target node corresponding to the current business scenario to continue forwarding.

[0074] During the retry and / or backup target node handover process, record one or more of the following: the time of the exception, the failure data identifier, the reason for the failure, the number of retries, the target node information, and the handover result, in order to generate an exception handling record; The anomaly handling records, along with corresponding historical transmission success rates, average transmission time, failure rates, and other operational metrics, are provided to the intelligent routing optimization module for subsequent training, updating, and parameter adjustment of the routing optimization model.

[0075] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0076] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A DICOM image routing system based on port multiplexing, characterized in that, include: The DICOM port listening module is configured to bind to a shared DICOM port and listen for TCP connection requests, receiving DICOM protocol data. The DICOM protocol parsing module is configured to selectively parse command sets and datasets in DICOM protocol data according to command type, and extract route-related tags according to preset parsing rules to generate a structured session context based on semantic information. The routing decision module is configured to perform matching processing on the generated structured session context based on multi-field rule matching and index acceleration mechanism, and filter candidate target DICOM receiving nodes; The routing optimization module is configured to obtain the running status information and transmission performance information of the candidate target DICOM receiving nodes, and dynamically evaluate the candidate target DICOM receiving nodes based on the routing-related fields in the session context of the current DICOM association request to determine the final target DICOM receiving node.

2. The DICOM image routing system based on port multiplexing as described in claim 1, characterized in that, The preset parsing rules include tag whitelist parsing rules, on-demand parsing rules based on command type, and key field parsing rules; Alternatively, the DICOM protocol parsing module includes a session identification and first packet multiplexing unit, a PDU parsing unit, a command set on-demand parsing unit, a whitelist tag parsing unit, a private tag parsing unit, and a session context construction unit; The session identification and first packet multiplexing unit is used to identify DICOM-related requests in the first data packet based on DICOM protocol data. The PDU parsing unit is used to parse the DICOM protocol data and extract the calling AE header and the called AE header for the identified DICOM association request; The whitelist label parsing unit is configured with preset label whitelist parsing rules. For the dataset part that needs to be parsed, it extracts routing-related standard labels according to the preset label whitelist. The private tag parsing unit is configured to selectively parse private tags in the dataset of DICOM protocol data based on a pre-configured private tag dictionary; The session context building unit is configured to generate the corresponding session context based on the parsing results, including the title of the calling AE, the title of the called AE, the route-related standard tags, and the route-related fields in the private tags.

3. The DICOM image routing system based on port multiplexing as described in claim 2, characterized in that, The DICOM protocol parsing module also includes a fault-tolerant processing unit, which is configured to retain the basic routing fields when some fields are missing.

4. The DICOM image routing system based on port multiplexing as described in claim 1, characterized in that, The routing decision module includes: The index filtering unit is used to filter a subset of candidate rules from the configured set of routing rules based on the structured session context and a preset high-discrimination field. The rule filtering unit is configured to perform conditional filtering on a subset of candidate rules by joint matching of multiple fields, filter out the hit routing rules that meet the matching conditions and sort them, select one or more routing rules with the highest sorting results as routing decisions, and filter out one or more candidate target DICOM receiving nodes. The cache hit unit is configured to maintain a cache of routing results for repeated session feature combinations within a preset time range, and to directly output the corresponding routing results when the cache usage conditions are met, so as to determine the candidate target DICOM receiving node.

5. The DICOM image routing system based on port multiplexing as described in claim 1, characterized in that, The running status information and transmission performance information obtained in the routing optimization module include one or more of the following: average transmission time, failure rate, current load, and historical success rate. Routing-related fields include one or more of the following: the title of the calling AE, the title of the called AE, Modality, inspection location, Study description, Series description, device manufacturer, device model, and key private tag values.

6. The DICOM image routing system based on port multiplexing as described in claim 1, characterized in that, It also includes a routing rule management module, which provides a set of routing rules to the routing decision module and is configured to store and maintain the set of routing rules. Alternatively, it may also include a forwarding and storage module, connected to the routing optimization module, configured to establish a DICOM association based on the final target DICOM receiving node, and forward the corresponding DICOM image data to the final target DICOM receiving node for storage and / or processing.

7. The DICOM image routing system based on port multiplexing as described in claim 1, characterized in that, It also includes a forwarding and storage module, which is connected to the routing optimization module and is configured to establish a DICOM association based on the final target DICOM receiving node and forward the corresponding DICOM image data to the final target DICOM receiving node for storage and / or processing. Configured to perform the following procedures: Retrieve the lifecycle status information of the recorded DICOM session; Monitor the connection and transmission status during the DICOM image data forwarding process, and monitor the connection status, response status, data transmission progress and / or timeout status in real time to determine whether the current forwarding process is normal; When the target node is detected to be unreachable, transmission fails, or timeout occurs, a retry will be performed according to the preset retry policy. If a retry fails, switch to a backup target node corresponding to the current business scenario to continue forwarding.

8. A DICOM image routing method based on port multiplexing, characterized in that, Includes the following steps: Listen for TCP connection requests based on the configured shared DICOM port and receive DICOM protocol data; For the acquired DICOM protocol data, the command sets and datasets in the DICOM protocol data are selectively parsed according to command type, and routing-related tags are extracted according to preset parsing rules to generate a structured session context based on semantic information. Based on multi-field rule matching and index acceleration mechanisms, the generated structured session context is matched and filtered to select candidate target DICOM receiving nodes; Obtain the running status and transmission performance information of the candidate target DICOM receiving nodes. Based on the routing-related fields in the session context of the current DICOM association request, dynamically evaluate the candidate target DICOM receiving nodes to determine the final target DICOM receiving node.

9. The DICOM image routing method based on port multiplexing as described in claim 8, characterized in that, The method for selectively parsing command sets and datasets in DICOM protocol data according to command type, extracting routing-related tags based on preset parsing rules, and generating a structured session context based on semantic information includes the following steps: The first data packet based on the DICOM protocol data identifies the DICOM association request; For DICOM association requests that have passed the identification process, parse the DICOM protocol data and extract the title of the calling AE and the title of the called AE. Preset command type on-demand parsing rules, selectively parse command sets and corresponding datasets based on DIMSE command type; The preset label whitelist parsing rules are used to extract routing-related standard labels for the dataset parts that need to be parsed, according to the preset label whitelist. Based on a pre-configured private tag dictionary, private tags in the dataset of DICOM protocol data are selectively parsed; Based on the parsing results, the corresponding session context will be generated from the calling AE title, the called AE title, the routing-related standard tags, and the routing-related fields in the private tags.

10. The DICOM image routing method based on port multiplexing as described in claim 8, characterized in that, Based on multi-field rule matching and index acceleration mechanisms, a method for matching and filtering candidate target DICOM receiving nodes by matching the generated structured session context includes the following steps: Based on the structured session context and a preset high-discrimination field, a subset of candidate rules is selected from the configured set of routing rules. The candidate rule subset is subjected to conditional filtering by multi-field joint matching. The matching routing rules that meet the matching conditions are selected and sorted. One or more routing rules with the highest sorting results are selected as routing decisions, and one or more candidate target DICOM receiving nodes are selected. A routing result cache is maintained for recurring session feature combinations within a preset time range, and the corresponding routing result is directly output when the cache usage conditions are met to determine the candidate target DICOM receiving node; the session feature combination includes one or more of the following: calling AE header, called AE header, Modality, and source IP network segment.

Images