Method and System for Interoperability of Medical Imaging Clouds Based on DICOM Protocol
By deploying image front-end gateways and DICOM protocol processing in medical institutions, a multi-level indexing system is constructed, and a unified software platform and end-to-end encryption technology are provided. This solves the problems of cross-institutional sharing of medical image data and mutual recognition of examination results, realizes multi-role collaborative diagnosis and treatment and data security, and improves the efficiency and security of medical resource utilization.
Patent Information
- Application Number
- CN202611316144.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-28
- Publication Date
- 2026-11-17
AI Technical Summary
The existing methods for accessing and sharing medical imaging data suffer from problems such as limited functionality, difficulties in cross-institutional data interaction, lack of multi-role collaborative diagnosis and treatment functions, failure to align with policies on mutual recognition of examination and test results, and insufficient security.
By deploying image front-end gateways in medical institutions, using the DICOM protocol for standardized image data processing and encrypted transmission, constructing a multi-level indexing system, providing a unified software platform to support multi-role access control, realizing mutual recognition of image consultation and examination results, and employing end-to-end encryption technology to ensure data security.
It has enabled standardized management and cross-institutional sharing of medical imaging data, provided differentiated functional services, supported online consultations and diagnostic reports, achieved mutual recognition of examination results, and improved the efficiency and security of medical resource utilization.
Smart Images

Figure CN123120216A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of medical image information transmission and storage management technology, specifically a medical image cloud interconnection method and system based on the DICOM protocol. Background Technology
[0002] With the rapid development of digital medical imaging technology, the Digital Imaging and Communications in Medicine (DICOM) standard has become a globally accepted standard for medical image information exchange. Currently, medical institutions at all levels widely use Picture Archiving and Communication Systems (PACS) based on the DICOM protocol to manage and store image data.
[0003] However, existing methods for accessing and sharing medical imaging data have many shortcomings. Currently, most hospitals use QR codes or embedded web links to provide access to imaging results, requiring patients to enter their ID number or the last six digits of their mobile phone number for authentication before viewing the imaging data. This model has the following drawbacks:
[0004] First, its functionality is limited, supporting only basic online operations. Current methods can only perform basic operations such as browsing images, adjusting window width and level, zooming in and out, horizontal movement, and taking screenshots, which cannot meet the needs of in-depth image data processing in clinical diagnosis and academic exchange.
[0005] Secondly, each hospital system operates independently, lacking a unified standard. Different medical institutions use PACS systems from different manufacturers. Due to the inconsistent image management information data models and interface standards adopted by each manufacturer, cross-institutional data exchange faces difficulties in interface adaptation and increased complexity in development and expansion. When patients visit different hospitals, they often need to undergo repeated imaging examinations, which not only increases medical costs but also wastes medical resources.
[0006] Third, there is a lack of multi-role collaborative diagnosis and treatment capabilities. Existing technologies cannot realize functions such as online consultation application, writing consultation reports, and publishing imaging diagnostic reports, which limits the cross-regional sharing and efficient utilization of high-quality medical resources.
[0007] Fourth, there is a lack of effective alignment with national policies on mutual recognition of test results. According to the "Guiding Opinions on Further Promoting Mutual Recognition of Test Results among Medical Institutions" issued by the National Health Commission and other departments, by the end of 2025, all closely integrated medical consortia must achieve mutual recognition of all test results, and all secondary and above public hospitals must enable cross-institutional access to test results. However, current technology lacks effective technical solutions and implementation pathways to align with these mutual recognition policies.
[0008] To address the aforementioned issues, some improvements have emerged in existing technologies. For example, a patent has proposed an image cloud interconnection method that creates image examination request resources through an FHIR server and receives and saves image data through a DICOM gateway. Another solution proposes a remote query and retrieval system based on DICOM object storage, which enables remote retrieval of image data through a front-end module, a central service registration module, and a cloud service query and archiving module. However, these solutions still have the following shortcomings: (1) they do not realize differentiated function presentation and permission management for multiple roles; (2) they lack effective connection with the national policy on mutual recognition of examination and testing results; (3) they do not provide a complete end-to-end solution for consultation application, consultation opinion writing, and image diagnosis report publication; and (4) they lack a security guarantee mechanism for the entire process of image data acquisition, transmission, storage, and retrieval.
[0009] Therefore, there is an urgent need for a medical imaging cloud interconnection method and system based on the DICOM protocol to solve the problems existing in the above-mentioned technologies.
[0010] The information disclosed above in this background section is only for enhancing the understanding of the background technology of this invention, and therefore may include prior art that is not known to those skilled in the art. Summary of the Invention
[0011] The purpose of this invention is to provide a method and system for interconnecting medical imaging clouds based on the DICOM protocol, so as to solve the problems in the prior art.
[0012] To achieve the above objectives, the present invention provides the following technical solution: a method for interconnecting medical imaging clouds based on the DICOM protocol, comprising the following steps:
[0013] Step S1: Deploy image front-end gateways in each medical institution, establish communication connections with the PACS systems of each medical institution through the DICOM protocol, acquire medical image examination data in real time, and perform standardized processing on the image data, including parsing the DICOM file header information to extract patient information, examination information, sequence information and image information, performing format verification and integrity verification on the image data, performing anonymization and desensitization processing, and compressing it using a lossless compression algorithm;
[0014] Step S2: Upload the standardized DICOM image data to the cloud medical imaging data center for storage through a secure encrypted channel, and construct a multi-level index system, which includes a patient index, an examination index, a sequence index, and a full-text search index;
[0015] Step S3: Build a unified software platform that supports patients, clinicians, and radiologists to access the system through terminals. Identify user roles based on user authentication information and assign corresponding functional permissions and data access permissions based on roles.
[0016] Step S4: In response to the image consultation request initiated by the first user through the software platform, verify the identity and permissions of the initiating user, retrieve the corresponding DICOM image data from the cloud medical imaging data center according to the patient identifier and examination identifier contained in the consultation request, package the retrieved DICOM image data and related clinical data to generate a consultation case and send it to one or more target terminals, establish a multi-party real-time communication channel to support the consultation participants to view and discuss images synchronously, and receive and publish consultation opinions and / or image diagnosis reports;
[0017] Step S5: Connect to the national and local medical institution examination and test result mutual recognition platform to obtain mutual recognition rules and mutual recognition item list. For each image examination data stored in the cloud medical imaging data center, conduct mutual recognition assessment based on institutional qualifications, equipment quality control, examination timeliness and image quality factors. Add mutual recognition mark to the image examination data according to the assessment results and display mutual recognition status information when accessing the data.
[0018] Step S6: End-to-end encryption technology is used to ensure data security during image data transmission. DICOM image data stored in the cloud is encrypted and stored, and data access and operation behaviors are logged for auditing and traceability.
[0019] Preferably, the image data standardization processing in step S1 further includes:
[0020] For DR, CT, and MR image data, a lossless compression algorithm based on predictive coding is used for compression. The lossless compression algorithm includes:
[0021] Pixel-by-pixel prediction is performed on the pixel data of the DICOM image, and the pixel prediction value is... Calculate using the following formula:
[0022]
[0023] in, This is the pixel value in the same column of the previous row. The pixel values of the previous column in the same row. The value is the pixel value of the previous row and column, α, β, and γ are prediction coefficients, and α+β+γ=1;
[0024] Calculate the true pixel value p {i,j} Compared with the predicted value The difference e {i,j} =p {i,j} - For the difference e {i,j} Entropy encoding is performed to generate a compressed image data stream.
[0025] Preferably, the multi-level index system in step S2 adopts a hierarchical hash index structure, wherein:
[0026] The patient index uses the hash value of the patient's unique identifier (PID) as the index key, i.e. ,in Use the SHA-256 hash function;
[0027] The check index uses the combination of the check instance UID hash and the patient index key as the index key, i.e. ,in" " indicates a string concatenation operation;
[0028] The full-text search index uses a TF-IDF-based inverted index structure for terminology in image report text. Its weight Calculate using the following formula:
[0029]
[0030] in, For the term In the report word frequency in , For the total number of reports, For included terms The number of reports.
[0031] Preferably, step S3, which assigns corresponding functional permissions and data access permissions based on roles, specifically includes:
[0032] A role-based access control model is adopted, defining a role set R = {r p ,r c ,r r},in p Indicates the patient role, r c Indicates the role of a clinician, r r Indicates the role of a radiologist;
[0033] Define a permission set P, and the patient role permission set P. rp={Query personal images, browse images, download images, share images}, the set of permissions for the clinical physician role is P. rc ={Query authorized patient images, initiate consultation requests, view consultation opinions}, the radiologist role permission set is P rr ={Query authorized patient images, initiate consultation request, view consultation opinions, receive consultation requests, review images online, write consultation opinions, publish consultation opinions, write diagnostic reports, and publish diagnostic reports};
[0034] Data access permissions adopt an attribute-based access control policy. The access authorization value A(u,d) for user u to data object d is determined by the following formula:
[0035] A(u,d)=1 if u's role permissions include the access operations required by d and u's attributes satisfy d's access conditions; otherwise, A(u,d)=0.
[0036] User u is allowed to access data object d if and only if A(u,d)=1.
[0037] Preferably, step S4, establishing a multi-party real-time communication channel to support simultaneous image review and discussion among consultation participants, specifically includes:
[0038] Establish a multi-party real-time communication channel based on WebRTC, with the consultation host terminal acting as a signaling server to establish P2P connections with the terminals of each participating party;
[0039] During the synchronized video viewing process, the video operation events of the host terminal are captured in real time and serialized into a sequence of operation instructions, which are then broadcast to the terminals of each participating party through a real-time communication channel.
[0040] Each participating terminal receives and parses the operation instruction sequence, and synchronously executes the same image operation in the local image browsing component to achieve real-time synchronization of multi-terminal image display;
[0041] The data structure of the operation instruction sequence is C={t,op,params,uid}, where t is the timestamp, op is the operation type, params is the set of operation parameters, and uid is the user identifier.
[0042] Preferably, the mutual recognition assessment in step S5 specifically includes:
[0043] For imaging data E, its mutual recognition score Calculate using the following formula:
[0044]
[0045] in, The assessment score is based on the institution's rating within the mutual recognition medical institution whitelist. The equipment quality control score is determined by whether the equipment has passed the quality control inspection. To check the timeliness score, =e -λ·Δt , where Δt is the number of days from the inspection date to the current date, and λ is the time decay coefficient; Image quality is scored based on the image's contrast, resolution, and noise level, using a pre-defined machine learning model. , , , The corresponding weight coefficients, and satisfying + + + =1;
[0046] like ≥T accept T accept If the preset mutual recognition acceptance threshold is met, the image examination data is marked as "mutually recognized"; otherwise, it is marked as "not mutually recognized" or "conditionally mutually recognized".
[0047] Preferably, the end-to-end encryption technology in step S6 specifically includes:
[0048] A secure transmission channel based on the SSL / TLS protocol is established between the image front-end gateway and the cloud data center. The RSA asymmetric encryption algorithm is used for key negotiation, and the AES-256 symmetric encryption algorithm is used to encrypt the transmission of DICOM image data.
[0049] For DICOM image data stored in the cloud, a key generation method based on DICOM file header information is used to generate a storage encryption key, wherein the storage encryption key K... store Generate using the following formula:
[0050] K store =PBKDF2(Password,Salt,iterations,keylen)
[0051] Where Password is the cryptographic material generated based on the combination of the patient ID and examination UID in the DICOM file, Salt is the randomly generated salt value, iterations is the number of iterations, and keylen is the key length;
[0052] The encryption keys for different image data are independent of each other, and the keys are stored separately from the encrypted image data. Access to the keys requires additional identity authentication and permission verification.
[0053] Preferably, the method further includes an image data integrity verification step:
[0054] After the DICOM image data is uploaded to the cloud data center, an integrity check value is calculated for the image data. This integrity check value uses a Merkle hash tree-based verification structure. For a dataset containing n DICOM files... Its root hash value Calculate using the following formula:
[0055]
[0056] Where H is the SHA-256 hash function. Let be the hash value of the i-th DICOM file;
[0057] The root hash value The image data is stored in association with the image data. When the image data is retrieved, the root hash value of the current data is recalculated and compared with the stored root hash value. If the two match, the verification is successful; otherwise, it is determined that the image data has been tampered with and an alarm is issued.
[0058] Preferably, the image diagnostic report in step S4 is generated using a structured report template, which includes the following standardized fields:
[0059] Report header fields: Report number, patient information, examination information, report type, report date;
[0060] The imaging findings field contains standardized descriptive entries for tissues categorized by anatomical location and sequence type. Each entry includes a location identifier, a description of the signs, and a severity score.
[0061] Diagnostic conclusion field: includes primary diagnosis, differential diagnosis, and recommendations;
[0062] During the report generation process, the system automatically fills in the report header fields based on the DICOM tags of the image data. Radiologists complete the writing of image findings and diagnostic conclusions by selecting standardized descriptive entries and entering free text.
[0063] Once the report is completed, the system will automatically perform a report integrity check. The check rules include: whether the required fields are complete, whether the diagnostic conclusions are logically consistent with the imaging findings, and whether the terminology used conforms to the standards. Only after the check passes can the report be submitted and published.
[0064] This invention also provides a medical imaging cloud interconnection system based on the DICOM protocol, comprising:
[0065] The image acquisition layer, deployed in various medical institutions, includes a DICOM communication module, a data parsing module, a data desensitization module, and a data compression module. It is used to establish a communication connection with the PACS system of each medical institution through the DICOM protocol, acquire medical image examination data in real time, and perform parsing, desensitization, and compression processing on the image data.
[0066] The cloud storage layer, deployed in the cloud, includes an image storage module, an index building module, and a data encryption module. It is used to store and manage processed DICOM image data and build a multi-level index system.
[0067] The business logic layer, deployed in the cloud, includes a user management module, a permission management module, a consultation management module, a report management module, and a mutual recognition management module, which are used to realize user authentication, role and permission allocation, consultation process control, diagnostic report generation, and mutual recognition and evaluation of examination results;
[0068] The terminal access layer includes a patient-side APP, a clinician-side APP, and a radiologist-side APP, which are used to provide differentiated user interfaces and functional services according to different roles. The functional services include image query, download and sharing on the patient-side, consultation request initiation on the clinician-side, and online consultation, consultation opinion writing and diagnostic report publication on the radiologist-side.
[0069] The security layer includes an identity authentication module, a transmission encryption module, a storage encryption module, and an audit log module, which are used to provide full-process security protection for data transmission, data storage, and data access behavior.
[0070] Compared with the prior art, the present invention has the following beneficial effects:
[0071] 1. This invention effectively solves the interface compatibility problem between PACS systems from different manufacturers by deploying standardized image front-end gateways in various medical institutions and uniformly adopting the DICOM 3.0 protocol for image data acquisition, transmission and storage, thereby realizing standardized management and cross-institutional sharing of medical image data.
[0072] 2. This invention provides differentiated functional services for patients, clinicians, and radiologists through a unified software platform. Patients can query, download, and share imaging data online; clinicians can query and initiate consultation requests online; and radiologists can conduct online consultations, write consultation opinions, and publish imaging diagnostic reports, thus realizing closed-loop management of the entire process from imaging examination to diagnostic report.
[0073] 3. This invention, through its built-in mutual recognition assessment module, connects to national and local mutual recognition platforms to automatically assess and label imaging examination data, providing clear mutual recognition status information for medical institutions and patients, and effectively supporting the implementation of policies for mutual recognition of examination and testing results.
[0074] 4. By enabling cross-institutional sharing and mutual recognition of medical imaging data, this invention avoids repeated imaging examinations for patients when visiting different medical institutions, thus saving medical resources and reducing the medical expenses burden on patients. Through online consultation and remote diagnostic reporting functions, patients in primary healthcare institutions can conveniently obtain diagnostic services from radiologists in higher-level hospitals, which helps alleviate the problem of uneven distribution of medical resources. Attached Figure Description
[0075] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.
[0076] Figure 1 This is a flowchart illustrating the medical imaging cloud interconnection method based on the DICOM protocol of the present invention.
[0077] Figure 2 This is a schematic diagram of the overall architecture of the medical imaging cloud interconnection system based on the DICOM protocol of this invention. Detailed Implementation
[0078] To enable those skilled in the art to better understand the technical solutions of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.
[0079] Example 1: Image Data Acquisition and Standardization Processing
[0080] In this embodiment, the PACS system of a tertiary hospital is used as an example to illustrate the process of image data acquisition and standardization.
[0081] First, an image front-end gateway device is deployed on the hospital's intranet. The front-end gateway is configured with DICOM communication parameters, including the device AETitle (Application Entity Title), interface port, and IP address. The front-end gateway establishes a connection with the hospital's PACS system via the DICOM C-STORE service, receiving DICOM image data pushed by the PACS system.
[0082] Upon receiving DICOM image data, the front-end gateway performs the following processing steps:
[0083] Step 1: DICOM File Parsing. The front-end gateway reads the DICOM file header and parses each data element in the file according to the DICOM 3.0 protocol standard. Key information obtained from the parsing includes:
[0084] Patient module information: PatientName, PatientID, PatientBirthDate, PatientSex.
[0085] Check module information: StudyInstanceUID, StudyDate, StudyTime, StudyDescription, Modality.
[0086] Sequence module information: SeriesInstanceUID (sequence instance UID), SeriesNumber (sequence number), SeriesDescription (sequence description);
[0087] Image module information: SOPInstanceUID (SOP instance UID), Rows (number of image rows), Columns (number of image columns), PixelSpacing (pixel spacing).
[0088] Step 2: Data Anonymization. The parsed patient identity information is anonymized. Specifically, the patient name is replaced with an encrypted anonymous identifier, and the patient ID is replaced with a unique anonymous ID generated internally by the system. Necessary examination information and imaging data are retained for clinical diagnosis. The anonymization process uses an irreversible hash encryption algorithm to ensure that the original patient information cannot be recovered from the anonymized data.
[0089] Step 3: Image Data Compression. The pixel data in the desensitized DICOM image data is compressed. Different compression strategies are adopted based on the image modality:
[0090] For DR, CT, and MR image data, a lossless compression algorithm based on predictive coding is used for compression. Specifically, pixel-by-pixel prediction is performed on the pixel data of the DICOM image, and the pixel prediction value is... Where α=0.5, β=0.3, γ=0.2. Calculate the true pixel value p. {i,j} Compared with the predicted value The difference e {i,j} =p {i,j} - The differences are entropy encoded to generate a compressed image data stream. This compression method ensures that the image details required for diagnosis are not lost.
[0091] For DR and CR images, near-lossless compression can be used to achieve a higher compression ratio while ensuring diagnostic quality.
[0092] The compressed DICOM image data is uploaded to the cloud-based medical imaging data center via an SSL / TLS encrypted channel.
[0093] Example 2: Multi-role permission management and function presentation
[0094] This embodiment details the role-based access control and functional differentiation mechanism.
[0095] The system employs a role-based access control (RBAC) model. When registering through the software platform (APP), users must select a role type and submit corresponding qualification verification materials. After the system administrator approves the application, the corresponding role permissions are assigned to the user.
[0096] (a) Patient-side function implementation
[0097] Patients log in to the system after completing real-name authentication using their mobile phone number and ID card number. The main functions of the patient-side interface include:
[0098] Examination Record Inquiry: Patients can view a list of all their historical imaging examination records on the "My Examinations" page. The list is arranged in reverse chronological order by examination date and displays the examination institution, examination type, examination date, and report status.
[0099] Image Viewing: After clicking on a specific examination record, the patient enters the image browsing page. The system loads the corresponding DICOM image data from the cloud and displays the medical images on the mobile device using HTML5 or a native rendering engine. The patient can perform operations such as adjusting window width and level, zooming, panning, rotating, and measuring.
[0100] Image download: Patients can download their DICOM image data to local storage. The downloaded files are in standard DICOM format or a common image format.
[0101] Image sharing: Patients can generate sharing links or QR codes for image data, set the sharing validity period and access permissions, and send the link to designated clinicians or medical institutions.
[0102] (II) Implementation of Clinical Physician Terminal Functions
[0103] Clinical physicians log in to the system after passing the medical practitioner qualification certification. The main functions of the clinical physician's side include:
[0104] Patient image retrieval: Clinicians can enter patient identification information, including name, ID number or medical record number, and, with patient authorization or system authorization, retrieve all imaging examination data of that patient in the system.
[0105] Consultation request initiation: The clinician selects the imaging examination data of the patient who needs to be consulted, fills out the consultation request form, including the patient's brief medical history, suspected clinical diagnosis, purpose of consultation, etc., selects the target consultation department or designated radiologist, and submits the consultation request.
[0106] (III) Functional Implementation for Radiologists
[0107] Radiologists log in to the system after passing the medical practitioner qualification certification and radiology professional qualification certification. The main functions of the radiologist's side include:
[0108] Consultation reception and processing: Radiologists can view the received consultation requests in the "Pending Consultations" list. After clicking, they can access the complete DICOM imaging data and clinical information for professional image interpretation and diagnosis.
[0109] Consultation opinion writing: After reviewing the images, radiologists write consultation opinions online, including imaging findings, diagnostic conclusions, and recommendations.
[0110] Consultation Opinion Publication: After the radiologist completes the consultation opinion and it is reviewed, the consultation opinion can be published with one click. The consultation opinion will be pushed to the initiating party (clinician) and the patient in real time.
[0111] Image diagnostic report writing and publication: Radiologists can write formal image diagnostic reports online, including the complete process of report template selection, report content filling, report review and publication.
[0112] Example 3: Cross-institutional access to and collaborative consultation of imaging data
[0113] This embodiment details the implementation process of cross-institutional image retrieval and collaborative consultation.
[0114] Scenario Description: Patient Zhang underwent a cranial MRI at People's Hospital of City A (a secondary hospital). Clinician Li suspected an intracranial space-occupying lesion and requested a consultation from Wang, a neuroimaging expert at the Affiliated Hospital of Medical University of City B (a tertiary hospital).
[0115] Implementation steps:
[0116] Step 1: Initiating a consultation request. Clinician Li logs into the system, clicks the "Request Consultation" button on the MRI examination record page of patient Zhang, fills out the consultation request form, including the patient's chief complaint, present medical history, positive signs of physical examination, preliminary diagnosis, etc., selects the department of the target consultation expert Wang, and submits the consultation request.
[0117] Step 2: System Routing and Notification. After receiving the consultation request, the system performs the following operations:
[0118] (1) Verify the authority and qualifications of clinician Li;
[0119] (2) Retrieve MRI DICOM image data of patient Zhang from the cloud;
[0120] (3) Package the imaging data, clinical data, and consultation request form into a consultation case package;
[0121] (4) Based on the target department and expert information, the consultation case package is pushed to the terminal of radiologist Wang;
[0122] (5) Notify Wang of new consultation assignments via push notification or SMS.
[0123] Step 3: Image Review and Interpretation. Radiologist Wang logs into the system and sees Zhang's consultation request in the "Pending Consultations" list. Wang clicks to enter the consultation details page, and the system loads Zhang's MRI DICOM image data from the cloud. Wang uses the image viewing tool for professional image interpretation, which allows for multi-sequence comparison browsing, 3D reconstruction, and fine adjustment of window width and level.
[0124] Step 4: Writing and Publishing Consultation Opinions. After reviewing the images, Wang wrote his consultation opinion online, including the imaging findings, diagnostic conclusions, and further suggestions. After completing the writing, Wang clicked the "Publish" button, and the consultation opinion was sent in real time to clinician Li and patient Zhang.
[0125] Step 5: Diagnostic Report Generation. If a formal imaging diagnostic report is required, Mr. Wang can use the report template, fill in the complete diagnostic report content, and have it approved by the department before publication. The published diagnostic report has legal effect and can serve as the official basis for clinical diagnosis and treatment.
[0126] Example 4: Evaluation of Mutual Recognition of Imaging Examination Results
[0127] This embodiment details the implementation process of mutual recognition assessment of image examination results.
[0128] The system connects to national and provincial inspection and testing result mutual recognition platforms and regularly obtains the latest mutual recognition rule base and mutual recognition item list.
[0129] For every piece of medical imaging data stored in the cloud-based medical imaging data center, the system performs a mutual recognition assessment upon data entry and each time it is accessed. The assessment process is as follows:
[0130] Step 1: Basic Information Acquisition. The system extracts the inspection organization information (InstitutionName), inspection equipment information (Manufacturer, ManufacturerModelName), inspection time (StudyDate), and inspection type (Modality) from the DICOM file.
[0131] Step 2: Mutual Recognition Condition Determination. The system determines the following mutual recognition conditions one by one:
[0132] (1) Institutional Qualification Requirements: Check whether the institution is a medical institution on the mutual recognition list. The system maintains the whitelist of mutual recognition medical institutions and updates it dynamically according to the list of mutual recognition medical institutions published by the national and local health departments.
[0133] (2) Equipment quality control conditions: Check whether the equipment has passed the quality control tests stipulated by the state. The system connects to the equipment quality control database to verify and check the quality control status of the equipment.
[0134] (3) Examination time limit conditions: Whether the examination time is within the mutually recognized valid time limit. Different types of examinations have different valid time limits. For example, the valid time limit for routine X-ray examinations is generally 30 days, and the valid time limit for CT examinations is generally 90 days.
[0135] (4) Image quality conditions: Whether the image quality meets the mutual recognition standard. The system uses an automatic image quality assessment algorithm to quantitatively evaluate indicators such as image contrast, resolution, and noise level.
[0136] Step 3: Mutual Recognition Scoring and Identification Generation. For image examination data E, its mutual recognition score... Calculate using the following formula:
[0137]
[0138] In this embodiment, the values of each parameter are as follows: Based on the institution's grade, Grade 3A is worth 100 points, Grade 3B is worth 85 points, and Grade 2A is worth 70 points. Passing the quality control test earns 100 points, while failing earns 0 points. =e -λ·Δt λ is set to 0.005; Evaluation was conducted using a convolutional neural network model, with scores ranging from 0 to 100; weighting coefficients were set to... =0.3, =0.2, =0.2, =0.3. Preset mutual recognition acceptance threshold T accept =70 points.
[0139] like ≥70, marked as "mutually recognized", and the mutual recognition validity period is displayed; if Scores between 60 and 69 are marked as "conditional mutual recognition," and the conditions that are not met are displayed; if If the value is less than 60, it is marked as "not mutually recognized" and the reason for the non-recognition is displayed.
[0140] Step 4: Mutual Recognition Information Display. When a user accesses image data, the system displays the mutual recognition status indicator and detailed information in a prominent position on the image browsing interface to provide a reference for the user's decision-making.
[0141] Example 5: System Deployment and Operation
[0142] This embodiment illustrates the actual deployment and operation of the system.
[0143] The system adopts a three-tier architecture: front-end deployment at the hospital, centralized deployment in the cloud, and distributed deployment at the terminals.
[0144] On the hospital side: Deploy image front-end gateway servers in each participating medical institution, configure DICOM communication parameters, and connect to the hospital's PACS system. The front-end gateway supports the standard DICOM 3.0 protocol and is compatible with systems from mainstream PACS vendors.
[0145] Cloud-based: Deploy cloud-based data centers on public or dedicated healthcare clouds, including distributed storage clusters, application server clusters, database clusters, and message queue clusters. The cloud-based system adopts a microservices architecture, supporting horizontal scaling and high availability deployment.
[0146] Terminals: Patients, clinicians, and radiologists access the system via smartphones or PC browsers. The mobile app supports mainstream mobile operating systems, while the PC version employs a responsive web design to adapt to different screen sizes.
[0147] Operation process:
[0148] During daily operation, the system automatically executes the following processes:
[0149] (1) When new imaging examination data is generated by the PACS system of each medical institution, it is automatically pushed to the front-end gateway;
[0150] (2) After the front-end gateway completes data parsing, desensitization and compression, it automatically uploads the data to the cloud data center;
[0151] (3) After the cloud data center completes data storage and index construction, it automatically triggers the mutual recognition assessment;
[0152] (4) After the assessment is completed, the patient can view the new examination records and imaging data on the APP;
[0153] (5) Clinicians and radiologists may view patient imaging data within the authorized scope according to their permissions;
[0154] (6) The entire process of consultation application, consultation opinion and diagnosis report is carried out in a closed loop within the system.
[0155] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.
Claims
1. A method for interconnecting medical imaging clouds based on the DICOM protocol, characterized in that: Includes the following steps: Step S1: Deploy image front-end gateways in each medical institution, establish communication connections with the PACS systems of each medical institution through the DICOM protocol, acquire medical image examination data in real time, and perform standardized processing on the image data, including parsing the DICOM file header information to extract patient information, examination information, sequence information and image information, performing format verification and integrity verification on the image data, performing anonymization and desensitization processing, and compressing it using a lossless compression algorithm; Step S2: Upload the standardized DICOM image data to the cloud medical imaging data center for storage through a secure encrypted channel, and construct a multi-level index system, which includes a patient index, an examination index, a sequence index, and a full-text search index; Step S3: Build a unified software platform that supports patients, clinicians, and radiologists to access the system through terminals. Identify user roles based on user authentication information and assign corresponding functional permissions and data access permissions based on roles. Step S4: In response to the image consultation request initiated by the first user through the software platform, verify the identity and permissions of the initiating user, retrieve the corresponding DICOM image data from the cloud medical imaging data center according to the patient identifier and examination identifier contained in the consultation request, package the retrieved DICOM image data and related clinical data to generate a consultation case and send it to one or more target terminals, establish a multi-party real-time communication channel to support the consultation participants to view and discuss images synchronously, and receive and publish consultation opinions and / or image diagnosis reports; Step S5: Connect to the national and local medical institution examination and test result mutual recognition platform to obtain mutual recognition rules and mutual recognition item list. For each image examination data stored in the cloud medical imaging data center, conduct mutual recognition assessment based on institutional qualifications, equipment quality control, examination timeliness and image quality factors. Add mutual recognition mark to the image examination data according to the assessment results and display mutual recognition status information when accessing the data. Step S6: End-to-end encryption technology is used to ensure data security during image data transmission. DICOM image data stored in the cloud is encrypted and stored, and data access and operation behaviors are logged for auditing and traceability.
2. The method for interconnecting medical imaging clouds based on the DICOM protocol according to claim 1, characterized in that: The image data standardization process in step S1 also includes: For DR, CT, and MR image data, a lossless compression algorithm based on predictive coding is used for compression. The lossless compression algorithm includes: Pixel-by-pixel prediction is performed on the pixel data of the DICOM image, and the pixel prediction value is... Calculate using the following formula: in, This is the pixel value in the same column of the previous row. The pixel values of the previous column in the same row. The value is the pixel value of the previous row and column, α, β, and γ are prediction coefficients, and α+β+γ=1; Calculate the true pixel value p {i,j} Compared with the predicted value The difference e {i,j} =p {i,j} - For the difference e {i,j} Entropy encoding is performed to generate a compressed image data stream.
3. The method for interconnecting medical imaging clouds based on the DICOM protocol according to claim 1, characterized in that: The multi-level index system in step S2 adopts a hierarchical hash index structure, wherein: The patient index uses the hash value of the patient's unique identifier (PID) as the index key, i.e. ,in Use the SHA-256 hash function; The check index uses the combination of the check instance UID hash and the patient index key as the index key, i.e. ,in" " indicates a string concatenation operation; The full-text search index uses a TF-IDF-based inverted index structure for terminology in image report text. Its weight Calculate using the following formula: in, For the term In the report word frequency in , For the total number of reports, For included terms The number of reports.
4. The method for interconnecting medical imaging clouds based on the DICOM protocol according to claim 1, characterized in that: The specific steps in step S3, which involve assigning corresponding functional permissions and data access permissions based on roles, include: A role-based access control model is adopted, and a role set R = {r p ,r c ,r r} is defined, wherein p represents a patient role, r c represents a clinician role, and r r represents a radiologist role; Define a set of permissions P, the patient role permission set is P rp = {query own images, browse images, download images, share images}, the clinician role permission set is P rc = {query authorized patient images, initiate consultation application, view consultation opinions}, the radiologist role permission set is P rr = {query authorized patient images, initiate consultation application, view consultation opinions, receive consultation application, online reading, write consultation opinions, publish consultation opinions, write diagnosis report, publish diagnosis report}; Data access permissions adopt an attribute-based access control policy. The access authorization value A(u,d) for user u to data object d is determined by the following formula: A(u,d)=1 if u's role permissions include the access operations required by d and u's attributes satisfy d's access conditions; otherwise, A(u,d)=0. User u is allowed to access data object d if and only if A(u,d)=1.
5. The method for interconnecting medical imaging clouds based on the DICOM protocol according to claim 1, characterized in that: The establishment of a multi-party real-time communication channel in step S4 to support simultaneous image review and discussion among consultation participants specifically includes: Establish a multi-party real-time communication channel based on WebRTC, with the consultation host terminal acting as a signaling server to establish P2P connections with the terminals of each participating party; During the synchronized video viewing process, the video operation events of the host terminal are captured in real time and serialized into a sequence of operation instructions, which are then broadcast to the terminals of each participating party through a real-time communication channel. Each participating terminal receives and parses the operation instruction sequence, and synchronously executes the same image operation in the local image browsing component to achieve real-time synchronization of multi-terminal image display; The data structure of the operation instruction sequence is C={t,op,params,uid}, where t is the timestamp, op is the operation type, params is the set of operation parameters, and uid is the user identifier.
6. The method for interconnecting medical imaging clouds based on the DICOM protocol according to claim 1, characterized in that: The mutual recognition assessment in step S5 specifically includes: For imaging data E, its mutual recognition score Calculate using the following formula: in, The score is used to determine the qualification of the institution, and the value is the grade score of the institution in the mutual recognition medical institution whitelist. The equipment quality control score is determined by whether the equipment has passed the quality control inspection. To check the timeliness score, =e -λ·Δt , where Δt is the number of days from the inspection date to the current date, and λ is the time decay coefficient; Image quality is scored based on the image's contrast, resolution, and noise level, evaluated using a pre-defined machine learning model. , , , The corresponding weight coefficients, and satisfying + + + =1; like ≥T accept T accept If the preset mutual recognition acceptance threshold is met, the image examination data is marked as "mutually recognized"; otherwise, it is marked as "not mutually recognized" or "conditionally mutually recognized".
7. The method for interconnecting medical imaging clouds based on the DICOM protocol according to claim 1, characterized in that: The end-to-end encryption technology in step S6 specifically includes: A secure transmission channel based on the SSL / TLS protocol is established between the image front-end gateway and the cloud data center. The RSA asymmetric encryption algorithm is used for key negotiation, and the AES-256 symmetric encryption algorithm is used to encrypt the transmission of DICOM image data. For DICOM image data stored in the cloud, a key generation method based on DICOM file header information is used to generate a storage encryption key, wherein the storage encryption key K... store Generate using the following formula: K store =PBKDF2(Password,Salt,iterations,keylen) Where Password is the cryptographic material generated based on the combination of the patient ID and examination UID in the DICOM file, Salt is the randomly generated salt value, iterations is the number of iterations, and keylen is the key length; The encryption keys for different image data are independent of each other, and the keys are stored separately from the encrypted image data. Access to the keys requires additional identity authentication and permission verification.
8. The method for interconnecting medical imaging clouds based on the DICOM protocol according to claim 1, characterized in that: The method also includes an image data integrity verification step: After the DICOM image data is uploaded to the cloud data center, an integrity check value is calculated for the image data. This integrity check value uses a Merkle hash tree-based verification structure. For a dataset containing n DICOM files... Its root hash value Calculate using the following formula: Where H is the SHA-256 hash function. Let be the hash value of the i-th DICOM file; The root hash value The image data is stored in association with the image data. When the image data is retrieved, the root hash value of the current data is recalculated and compared with the stored root hash value. If the two match, the verification is successful; otherwise, it is determined that the image data has been tampered with and an alarm is issued.
9. The method for interconnecting medical imaging clouds based on the DICOM protocol according to claim 1, characterized in that: The image diagnostic report in step S4 is generated using a structured report template, which includes the following standardized fields: Report header fields: Report number, patient information, examination information, report type, report date; The imaging findings field contains standardized descriptive entries for tissues categorized by anatomical location and sequence type. Each entry includes a location identifier, a description of the signs, and a severity score. Diagnostic conclusion field: includes primary diagnosis, differential diagnosis, and recommendations; During the report generation process, the system automatically fills in the report header fields based on the DICOM tags of the image data. Radiologists complete the writing of image findings and diagnostic conclusions by selecting standardized descriptive entries and entering free text. Once the report is completed, the system will automatically perform a report integrity check. The check rules include: whether the required fields are complete, whether the diagnostic conclusions are logically consistent with the imaging findings, and whether the terminology used conforms to the standards. Only after the check passes can the report be submitted and published.
10. A medical imaging cloud interconnection system based on the DICOM protocol, characterized in that, include: The image acquisition layer, deployed in various medical institutions, includes a DICOM communication module, a data parsing module, a data desensitization module, and a data compression module. It is used to establish a communication connection with the PACS system of each medical institution through the DICOM protocol, acquire medical image examination data in real time, and perform parsing, desensitization, and compression processing on the image data. The cloud storage layer, deployed in the cloud, includes an image storage module, an index building module, and a data encryption module. It is used to store and manage processed DICOM image data and build a multi-level index system. The business logic layer, deployed in the cloud, includes a user management module, a permission management module, a consultation management module, a report management module, and a mutual recognition management module, which are used to realize user authentication, role and permission allocation, consultation process control, diagnostic report generation, and mutual recognition and evaluation of examination results; The terminal access layer includes a patient-side APP, a clinician-side APP, and a radiologist-side APP, which are used to provide differentiated user interfaces and functional services according to different roles. The functional services include image query, download and sharing on the patient-side, consultation request initiation on the clinician-side, and online consultation, consultation opinion writing and diagnostic report publication on the radiologist-side. The security layer includes an identity authentication module, a transmission encryption module, a storage encryption module, and an audit log module, which are used to provide full-process security protection for data transmission, data storage, and data access behavior.
