Скачать презентацию MIRC Clinical Trials Medical Imaging Resource Center Скачать презентацию MIRC Clinical Trials Medical Imaging Resource Center

4ea0fd41d9a0a27f875a98bbf4457b64.ppt

  • Количество слайдов: 36

MIRC Clinical Trials Medical Imaging Resource Center MIRC Clinical Trials Medical Imaging Resource Center

RSNA 2006 MIRC Courses Overview of the RSNA MIRC Software Installing MIRC on Your RSNA 2006 MIRC Courses Overview of the RSNA MIRC Software Installing MIRC on Your Laptop Using MIRC for Clinical Trials Hands ON: Using the RSNA Teaching File

MIRC Project Objective Support the production, storage, indexing and distribution of medical imaging resources: MIRC Project Objective Support the production, storage, indexing and distribution of medical imaging resources: – Teaching files – Scientific & technical documentation – Research images & datasets – Clinical trials data

MIRC Concepts The key concepts: – MIRC is a community of cooperating libraries on MIRC Concepts The key concepts: – MIRC is a community of cooperating libraries on the Internet. – The libraries cooperate by supporting a common query mechanism. – The query mechanism allows users to find information wherever it is stored.

The RSNA MIRC Software Query Service Storage Service Document index and storage Author services The RSNA MIRC Software Query Service Storage Service Document index and storage Author services DICOM services File Service Public and private file cabinets

MIRC Site Services Submit Service Zip Service Input Queue Admin Service HTTP Import TCE MIRC Site Services Submit Service Zip Service Input Queue Admin Service HTTP Import TCE Service OS File System Storage Service MIRC Index DICOM Export Object Processor XMLServer HTTP Export DICOM Import Query Service MIRC Anonymizer DICOM Import File Service Author Services Database Export MIRC Anonymizer

DICOM Service Receives DICOM objects from PACS, workstations, modalities, … Automatic document generation – DICOM Service Receives DICOM objects from PACS, workstations, modalities, … Automatic document generation – Images grouped into documents by study – Template driven – Automatic JPEG generation from DICOM – Programmable anonymization (de identification and re identification)

DICOM Service Automatic trial patient ID generation Connection to other DICOM Services for clinical DICOM Service Automatic trial patient ID generation Connection to other DICOM Services for clinical trials – Support for encrypted communications (SSL) DICOM import / DICOM export HTTP(S) import / HTTP(S) export Database export

Database Interface Links to an externally developed database Provides classes for interrogating: – DICOM Database Interface Links to an externally developed database Provides classes for interrogating: – DICOM objects – Metadata files: • XML • ZIP (with manifest) Anonymization for DICOM and XML objects

MIRC Clinical Trial Data Flow DICOM HTTP FC Database Internet HTTP / HTTPS DICOM MIRC Clinical Trial Data Flow DICOM HTTP FC Database Internet HTTP / HTTPS DICOM PI DICOM FC Field. Center Application: - Image reception - Anonymization - Transmission to PI MIRC Site: - Image reception - Image Storage - Transfer to PACS and/or workstations

Anonymization & Pseudonymization Anonymization & Pseudonymization

Real World Data Modification • Anonymization – Removal of PHI – Replacement of PHI Real World Data Modification • Anonymization – Removal of PHI – Replacement of PHI with values that are the same for all patients • Pseudonymization – Replacement of PHI with values that are specific to the patient and the trial but which are not available anywhere else

Pseudonymization Element types – Patient ID – Dates • Birth • Study – UIDs Pseudonymization Element types – Patient ID – Dates • Birth • Study – UIDs – Accession, non unique IDs Central vs distributed remapping

The HIPAA 18 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. The HIPAA 18 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. Name Location; all geographic subdivisions smaller than a state, including street address, city, county, precinct, zip code, and their equivalent geocodes. Dates (all dates related to the subject of the information, e. g. birth dates, admission dates, discharge dates, encounter dates, surgery dates, etc. ) Telephone numbers Fax numbers Electronic mail addresses Social security numbers Medical record numbers Health plan beneficiary numbers Account numbers Certificate / license numbers Vehicle identifiers and serial numbers, including license plate numbers Device identifiers and serial numbers Web Universal Resource Locators (URLs) Internet Protocol (IP) address numbers Biometric identifiers, including finger and voice prints Full face photographic images and any comparable images Any other unique identifying number, characteristic, or code

Anonymizer Scripting • Modality specific • Manufacturer specific • MIRC defined script languages – Anonymizer Scripting • Modality specific • Manufacturer specific • MIRC defined script languages – DICOM – XML • Shared remapping tables – Self consistency

DICOM Anonymization Parameter access @param(@Parameter. Name) Date/time @date(separator) @time(separator) Element access @contents(Element. Name) @contents(Element. DICOM Anonymization Parameter access @param(@Parameter. Name) Date/time @date(separator) @time(separator) Element access @contents(Element. Name) @contents(Element. Name, "regex", "replacement") Element creation and deletion @require() @require(Element. Name, "default value") @empty() @blank(n) @remove() @keep()

DICOM Anonymization Special purpose data modification @initials(Element. Name) @scramble(Element. Name, word 1 skip, word DICOM Anonymization Special purpose data modification @initials(Element. Name) @scramble(Element. Name, word 1 skip, word 1 take, word 2 skip, word 2 take, . . . ) @alphabetichash(Element. Name, max. Chars. Output, max. Words. Input) @numerichash(Element. Name, max. Chars. Output, max. Words. Input) @ptid(siteid, Element. Name, prefix, first, width, suffix) @ptid(@SITEID, Element. Name, @PREFIX, first, width, @SUFFIX) @hashptid(siteid, Element. Name, prefix, suffix) @hashptid(@SITEID, Element. Name, @PREFIX, @SUFFIX) @offsetdate(siteid, Element. Name, basedate) @offsetdate(@SITEID, Element. Name, @BASEDATE) @incrementdate(Element. Name, inc. In. Days) @round(Element. Name, groupsize) @round(Element. Name, @Parameter. Name)

DICOM Anonymization UID functions @uid(root, Element. Name) @uid(@UIDROOT, Element. Name) @hashuid(root, Element. Name) @hashuid(@UIDROOT, DICOM Anonymization UID functions @uid(root, Element. Name) @uid(@UIDROOT, Element. Name) @hashuid(root, Element. Name) @hashuid(@UIDROOT, Element. Name) Non unique identifier functions @accession(Element. Name) @id(Element. Name) @integer() Conditional functions @if(Element. Name, exists){true clause}{false clause} @if(Element. Name, isblank){true clause}{false clause} @if(Element. Name, matches, "regex"){true clause}{false clause} @quarantine() @skip()

XML Anonymization Comment: #comment Identifier: $name Path: /tag[n]/. . . /tag[n]/…/@attr with wildcards [*] XML Anonymization Comment: #comment Identifier: $name Path: /tag[n]/. . . /tag[n]/…/@attr with wildcards [*] Methods: Remapping: $uid( expression ) Creation: $require( expression ) Destruction: $remove() Debugging: $print( expression ) Expression: path “text” identifier $method() etc. Assignment: identifier = expression path = expression

Mapping / Hashing • Central vs local mapping – Central mapping is more controlled, Mapping / Hashing • Central vs local mapping – Central mapping is more controlled, but it requires that PHI be kept at the PI’s site. – Local mapping is more susceptible to data loss, but has a different security profile. • Hashing – Very convenient in a multi center trial and very secure, but inversion can only be done through a dictionary attack.

Data Formats: Beyond DICOM Data Formats: Beyond DICOM

Clinical Trial Data Files • Images – DICOM, multi media • Text data – Clinical Trial Data Files • Images – DICOM, multi media • Text data – XML • Trial specific file collections – Zip • Undefined file type

Data Object Classes • File. Object • File management methods • Common data access Data Object Classes • File. Object • File management methods • Common data access methods overridden by subclasses: – get. UID – get. Study. UID – get. Description – Subclasses • Dicom. Object • Xml. Object • Zip. Object

Dicom. Object • Encapsulates a DICOM file • Key methods: – get. Mirc. Image Dicom. Object • Encapsulates a DICOM file • Key methods: – get. Mirc. Image – get. Buffered. Image – get. Element. Value – 41 methods in all

Xml. Object • Encapsulates an XML file • Key methods: – get. Document – Xml. Object • Encapsulates an XML file • Key methods: – get. Document – get. Text – get. Document. Element. Name – get. Value – 7 methods in all • Problem: lack of standardization for UID, Study. UID, and Description

Zip. Object • Encapsulates a zip file containing: • An XML file named manifest. Zip. Object • Encapsulates a zip file containing: • An XML file named manifest. xml – MIRC defined schema for key elements • Any other files • Key methods: – get. Manifest. Document – get. Entries – extract. File. Text – 15 methods in all

Managing Software and Configuration Updates Managing Software and Configuration Updates

Lessons from Field Centers • IT support is almost unavailable. • Initial software installation Lessons from Field Centers • IT support is almost unavailable. • Initial software installation requires a human being, but it should be simple. • Software updates should require a person to trigger them. • Anonymizer scripts should be automatically updated.

MIRC Update Service Field. Center Storage SCP HTTP Server Anonymizer Storage SCU S/W Update MIRC Update Service Field. Center Storage SCP HTTP Server Anonymizer Storage SCU S/W Update HTTP(S) MIRC Update Servlet S/W Config 1 Field. Center Sites Config 2 Config n

Field. Center Field. Center

Typical Imaging Dataflow Principal Investigator Field Center Modality PACS Workstations PACS Field. Center Storage Typical Imaging Dataflow Principal Investigator Field Center Modality PACS Workstations PACS Field. Center Storage SCP HTTP Server Anonymizer Storage SCU S/W Update HTTP(S) MIRC Database DICOM

The Field. Center Application Runs at the image acquisition site Receives images and metadata The Field. Center Application Runs at the image acquisition site Receives images and metadata files via the DICOM and HTTP(S) protocols Anonymizes images and metadata files Transmits to the PI MIRC site via HTTP(S) Automatically links to the PI MIRC site for remapping tables and software updates

The Field. Center Application The Field. Center Application

The RSNA MIRC Software Available on the RSNA MIRC site (mirc. rsna. org) – The RSNA MIRC Software Available on the RSNA MIRC site (mirc. rsna. org) – Click the Downloads link in the header

The RSNA MIRC Software Runs on Windows, Mac, Linux, Solaris Three components: Java / The RSNA MIRC Software Runs on Windows, Mac, Linux, Solaris Three components: Java / Tomcat / MIRC All components are free MIRC Users Group meetings: 11/27 11/30 1: 30 3: 30 12: 15 E 262 E 257

Questions/Answers & Help The MIRC Wiki – http: //mircwiki. rsna. org RSNA Forums: – Questions/Answers & Help The MIRC Wiki – http: //mircwiki. rsna. org RSNA Forums: – http: //forums. rsna. org – User Comments, Questions and Answers – RSNA MIRC Software Announcements – MIRC Implementers’ Corner Subscribe on the Forums site