1 / 21

Grid Based Infrastructure for Distributed Medical Imaging

Grid Based Infrastructure for Distributed Medical Imaging. Carl Kesselman ISI Fellow Director, Center for Grid Technologies Information Sciences Institute Research Professor Computer Science Viterbi School of Engineering University of Southern California

flower
Télécharger la présentation

Grid Based Infrastructure for Distributed Medical Imaging

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Grid Based Infrastructure for Distributed Medical Imaging Carl KesselmanISI FellowDirector, Center for Grid TechnologiesInformation Sciences Institute Research Professor Computer Science Viterbi School of Engineering University of Southern California Joint work with Stephan G. Erberich, Ann Chervenak, Robert Schuler, Laura Pearlman, Jonathan C. Silverstein

  2. Problem • How to share medical images across federated medical environment • Issues • Minimal disturbance of enterprise environment • Co-existence with existing medical imaging tools and user interfaces • Privacy/security requirements • Federation across multiple sites

  3. The Medicus Solution • Medical Imaging and Computing for Unified Information Sharing (MEDICUS) • Exploit existing medical imaging standards in local enterprise • Gateway into standard Grid service for federation • Security • Data discovery • Data movement Globus MEDICUS Proto-Project @ http://dev.globus.org/wiki/Incubator/MEDICUS

  4. Digital Imaging and Communicationsin Medicine (DICOM) • Defines image format • Standard header (metadata) and image formats • Simple communication protocol for image access and publication • store, find, get, move • Used by existing medical imaging systems • Picture Archiving and Communications Systems (PACS)

  5. Python Runtime C Runtime Java Runtime Open Source Grid Software Globus Toolkit v4 www.globus.org Data Replication CredentialMgmt Replica Location Grid Telecontrol Protocol Delegation Data Access & Integration Community Scheduling Framework WebMDS Reliable File Transfer CommunityAuthorization Workspace Management Trigger Authentication Authorization GridFTP Grid Resource Allocation & Management Index Security Data Mgmt Execution Mgmt Info Services CommonRuntime

  6. Major Components of Medicus • DICOM Grid Interface Service • OGSA web service to translate between DICOM and Grid operations • OGSA-DAI • Meta-catalog • Data Replication Service (DRS) • Data replication/data discovery • Utilized RLS and GridFTP for disovery, replica management and data movement • Grid Security Infrastructure • Security, authorization

  7. The Grid is the PACS • Meets image exchange needs • Not limited to research use (e.g. BIRN, caBIG) • Single architecture for Clinical and Research use • Federate image references (Meta Catalog) - IHE XDS model • X.509 authentication security model + SAML assertions • Hide Grid workflow from user if possible, e.g. DICOM workflow • Meets image storage needs • FT and DR by replicas • PACS-Grid-PACS too slow for clinical use • Integrate hospital PACS • Data integrity by CRC checksums

  8. Medicus System Design

  9. DGIS: Image publicationDICOM C-STORE Operation Globus MEDICUS Proto-Project @ http://dev.globus.org/wiki/Incubator/MEDICUS

  10. DGIS: Image DiscoveryDICOM C-FIND Operation Globus MEDICUS Proto-Project @ http://dev.globus.org/wiki/Incubator/MEDICUS

  11. Meta Catalog Service for Medical Images • OGSA-DAI + Data Base (e.g. MySQL, Derby, Oracle, ..) • DICOM meta data • Patient level (e.g. encrypted name, id, etc.) • Study level (e.g. date, time, protocol, etc.) • Series level (e.g. imaging type, modality, etc.) • Image level (e.g. position, level, exposure, etc.) • Keys are DICOM UIDs (Study, Series, Image) • Health meta data • Flexible Annotation, e.g. ICD-9

  12. DGIS: Image DeliveryDICOM C-GET/C-MOVE Operations Globus MEDICUS Proto-Project @ http://dev.globus.org/wiki/Incubator/MEDICUS

  13. MEDICUS Fault Tolerance and Disaster Recovery • Fault Tolerance and Disaster Recovery through replicas • OGSA compliant Replication Location Service (RLS) • Index encrypted DICOM keys (study and series UIDs) • Index which storage has physical representation of series record • Local replica index (RLS) • VO replica index (RLS master)

  14. Protected Health Information • MEDICUS v1 • Single layer GSI security model • X.509 proxy certificate standards based • Typical use case: Closed VO like Healthcare provider network, Military network, research network. • MEDICUS v2 • Add second security layer based on patient identity • Patient Centric Authorization using SAML assertions • Patient advocacy – patient controlled access • Logging of “on behalf actor” at Grid Service • All patient data on the Grid • Typical use-case: SOA of third-party storage, image processing services require no-PHI access to DICOM

  15. I2 Shibboleth – Identify Federation • I2 announcement 01/17/2007: .. Both the US National Science Foundation (NSF) and National Institutes of Health (NIH) are moving in this direction. The report states that "the federation model with the most momentum is Shibboleth". • GridShib using Shibboleth • OASIS standards based SAML assertions • GT4 - X.509 certificates with embedded SAML assertions

  16. Patient Authorized Grid Image Workflow

  17. Globus MEDICUS Use-Cases • Multi-center clinical trials • Children’s Oncology Group Phase-I28 international medical centers (since 09/2003) • NANT Cancer Foundation13 national medical centers (since 12/2005) • Off-site Medical Image Storage • Enterprise PACS / Grid PACS • FT and DR by replication using Globus Data Replication Service (DRS) • Medical Image Federation • Enterprise Hospital VO • Military VO • Community Practices VO • Etc.

  18. MEDICUS use cases: Childrens Oncology Group and Neuroblastoma Cancer Foundation Grids

  19. Summary • MEDICUS vertically integrates existing standards based GT4 components – no research specific layer • Fast and efficient DICOM off-site storage • Integrates with hospital PACS + FT and DR • Transparent image workflow for Physician • Flexible and cost efficient deployment using open-source (~ $500 per TB) • PHI protected at patient level • Single HealthGrid solution for Clinical and Research use of same images

  20. Conclusion • MEDICUS present one piece to HealthGrid puzzle • Modular SOA design ideal for collaborative extension, e.g. image processing web services using DICOM image resources on the Grid • Open-source (Apache license), part of theGlobus Toolkit Development release:You are invited to contribute your field of expertisedev.globus.org/wiki/Incubator/MEDICUS • Roadmap: Standards based PHR, Workstation Grid plug-in, IHE XDS/-I WebServices

  21. Horizon Award Winner 2007 Information Science Institute Acknowledgment http://dev.globus.org/wiki/Incubator/MEDICUS IDEA Award Winner 2007 NIH/NCI Grant: UO1-BA97452

More Related