oracle multimedia dicom api: next-generation platform for ... · oracle multimedia dicom: native...
Post on 15-May-2020
54 Views
Preview:
TRANSCRIPT
Oracle Multimedia DICOM API: Next-Generation Platform for Medical Imaging Solutions
Melli AnnamalaiPrincipal Product Manager, Oracle Multimedia and Medical ImagingDongbai GuoDICOM Architect and Senior Development Manager
<Insert Picture Here>
Program Agenda
• Medical Image Data Management Challenges
• Oracle Multimedia DICOM– Advantages of database infrastructure
– Native DICOM Support
• Use Cases– Image-enabled Electronic Health Records– Clinical Trials Image Management
– Medical Research
• Summary
What are Medical Images?
• X-Rays, CTs, MRIs, Pathology images, etc.
• DICOM (Digital Imaging and Communications in Medicine) is the universal standard data format– Images– Video
– Structured Reports
• Non-DICOM formats for legacy and other medical images– TIFF, JPEG, etc.
Medical Images Ubiquitous in Healthcare Applications
• Image data is pervasive across all departments and clinical specialties
• Key component of evidence-based medicine
• Image data used in a variety of applications
Courtesy: Louis Humphrey
Medical Image Data Management Challenges
• Share images across multiple systems: interoperability
• Save costs– Storage (hardware)
– Development, deployment and maintenance (labor)
• Scalability– Handle massive increase in data volume (image volume doubles
every 18 to 20 months*)
– Scale to large information networks
• Security and privacy while sharing and accessing images
• Protect against data loss
• Performance– Fast access to local, remote, and archived medical images
* Frost & Sullivan
Oracle Multimedia DICOM for Managing Medical Images
Oracle Database 11g
Patient id Patient Name
DICOM
John1005
…….
…….
Patient DOB
02-20-48
• Oracle Database 11g provides native and complete support for storing and managing medical images– Standard and Enterprise editions
• Address medical image data management challenges with database infrastructure
• Store and manage images together with other patient data in the database
Oracle Database 11g
Oracle Multimedia DICOM for Managing Medical Images
• Standards based open programmable platform for sharing and interoperability– Vendor-neutral DICOM storage and DICOM data model – Data accessible for secondary use– Future proof: Protect against changes in DICOM standard and
modalities– One database, one data model, multiple applications
• Low costs even as volume grows– Flexible choice for hardware – Eliminates maintenance of specialized code for image stores– Manage medical images with DBA and database developer skills– Information Lifecycle Management: Transparently move data to
cheaper disks with DICOM tags based transfer policy
Oracle Multimedia DICOM for Managing Medical Images
• Scalable to petabyte and exabyte medical image archives
• Guaranteed security and privacy – Uniform security policies across all types of data and all tiers of data
– Access control at the data level
– Audit access
– Anonymize for privacy, Ex: mask patient names
• Performance: As fast as underlying hardware device– Ex: Read 850 CT images/sec, Write 550 CT images/sec
• Reduces Risk– Protect images against loss with backup and recovery
• High Availability
Oracle Multimedia DICOM: Native DICOM Support
• Reduce development costs – APIs for rapid development and deployment
• (zero code) Metadata extraction– Extract all 2028 standard tags– Extract vendor-specific extensions– Fast, accurate, search based on any DICOM attribute
• (zero code) Automated DICOM data validation– Check conformance to application requirements and verify data quality– Easily develop workflows that are conformance aware
• (zero code) Include DICOM content in reports and display in browsers– Convert DICOM to web-friendly formats like JPEG– Create thumbnails
• (zero code) Create DICOM images from legacy non-DICOM images
Oracle Multimedia DICOM: Native DICOM Support
• Future Proof: Model driven architecture
• Changes managed through configuration documents in the event of– New version of DICOM standard
– New modality– New vendor-specific metadata
• No change in application, no downtime, no database upgrade
Electronic Health Records
• Electronic Health Records (EHRs) reduce healthcare costs– 360 degree view of all patient data + images as an integrated
whole– Longitudinal: Throughout the lifetime of the patient
– Latitudinal: Across all healthcare disciplines and specialties
• Link all patient information and patient images– Access from any location in multi-site hospitals
– Access across national and regional networks
– Access by physicians from work, home, other hospitals, access by patients
• Oracle Multimedia DICOM platform for image-enabled EHR
Image-enabled EHR: the i-engineers health engine
• Image-enabled “health engine”– Integrate images from PACS systems with patient health record– Use Oracle Multimedia DICOM to
• Extract DICOM metadata for physicians’ access and for driving workflow
• Convert to web-friendly formats for display in browsers and reports
• Customer case study: Spitalnetz Bern Hospital (Bern Health Network, Switzerland)– Universal archive for all content (images and patient data)– Provides access to complete patient health records for
• Physicians in hospital • Primary care physicians in clinics outside the hospital• 2000+ users
The health engine - Architectural Overview
Oracle Application Server
The Health Engine
Oracle Multimedia DICOM Support
DICOM Node
JDBC
DICOMHTTP(S) / SOAP
ASM
Courtesy: AdNovum and the i-engineers
<Insert Picture Here>
“The unique features of Oracle Multimedia in Oracle Database 11g
together with the i-engineers ‘health engine’ application helped us to
significantly reduce costs and improve quality and security of our
clinical document processing.”
Peter GerberCIO, Spital Netz Bern Hospital
Switzerland
Customer Drivers and Business Needs
• Images are part of clinical trial evidence, ex: Tumors, MRIs, X-Rays– Images are an essential part of the Clinical Trials and
Pharmaceutical Research process
• Images are scattered across pharma departments and Contract Research Organizations working on the clinical trial
• Image sharing and transfer is not managed– Use FTP, courier, CDs/DVDs/printouts
– Not secure, data is lost, data quality is compromised– Identifying metadata tags are lost/incorrect
Clinical Research at AstraZeneca
• Problem: – Exponential growth in clinical trials images (nearly 100
terabytes in one year) from X-Rays to microscopic images – Difficult to manage, index, access, and share
• Oracle Multimedia DICOM enables:– A centralized clinical image repository for rapid access,
data integrity, streamlining of compliance, multiple workflows
– A system as scalable and robust for managing images as for other clinical data
– Security to restrict access to authorized users– Data validation– A platform to collect, store and view images throughout
the enterprise
Clinical Research at Novartis
• Problem: Clinical trial images managed by third party Contract Research Organizations (CROs)– Poor Quality Management– Difficult Cost Control – Direct access to images limited and expensive for pharmas
• IBIS solution on Oracle 11g Multimedia DICOM enables– Efficient, lower-cost solution– Improved data quality– Security to restrict access to authorized users– Compliance– Data validation– Quick deployment of standard Oracle Database-based solution
Clinical Research Image Management: IBISimg
Oracle Multimedia
DICOMSite
CRO
Site
User Systems Adm. Reports
Presentation
Authentication
Data Export
Integration Module
Target Systems
CDMS
Study Management
Systems
Data Acquisition Interface
Rules Engine
DICOM Image Files
Study Data
DICOM Image Data
Security
Archive
Backup
Migration
Synchronized metadata with images
Data Archive
IBISimg
Image/data Viewer
Courtesy: IBISWorks
Clinical Research at University of Wisconsin
• Problem: – Pointers to images on file system where invariably lost
– Infrastructure of multiple systems for multiple studies impractical to continue
• Oracle Multimedia DICOM infrastructure– Provided a solution that is configurable and robust
– All studies housed in the same database instance and accessed using the same software
– Enabled workflow: Data stored in two temporary tables for processing before being moved to archive tables
– Non-DICOM (TIFF) images coming in converted into DICOM
• Data will come in from 50 studies (5000 images in each study)
• Ophthalmology images (YBR photometric interpretation)
Multi-disciplinary Medical Research: BioGrid Australia
• Simplifies research– Integrates health research data collated by clinicians in
different institutions
– Links data from multiple institutions in disparate silos
– 12 years of MRI images was scattered on 1000 tapes, taking weeks to months for retrieval
• Multi-terabyte Archive– Consolidates 32 databases from 14 institutions across 4
Australian states– Intranet based solution with 7 million images
– Image retrieval in seconds
– Continues to grow
Courtesy: Naomi Rafael, BioGrid Australia
BioGrid Australia
Authorised researchers query the federated research
repository & download data to their workstations for
computer aided diagnosis.
RMH –LRR(Inc WH Data)
PMC -LRR
Austin -LRR
Institute specific data loaded into institute
specific Local Research Repository
De-identified data
Oracle Database
11g
Extract T
ransform &
Load
Public Data Sources
GenBank
UniProt LocusLink
PubMed
PeterMac
Oncologyn=1,265
uArrayn=80 test
Tissue Bankn=1,764
Austin
Oncologyn=965
Biomarkersn=0
Tissue Bankn=1,663
Melbourne
Diabetesn=21,048
Surveillancen=3,541
Familial Can=3,758
Tissue Bankn=454
Oncologyn=3,600
Epilepsyn=1,823
Biomarkersn=50
Growing research data
Nightly updates
Researchers’ feedback is
input back to the system
BioGrid Australia
• Chose Oracle Multimedia DICOM because:– Simplified IT requirements through native processing of
DICOM images
• Accelerated development and deployment
– Preserved patient privacy by anonymizing DICOM images without losing the link between the image and patient record
– Reduced storage requirements
– Superior functionality at lower cost– Re-use implementation for future projects
– Skilled programmers readily accessible
<Insert Picture Here>
“Oracle Database 11g gives us functionality that allows us to preserve patient
confidentiality while allowing experts to undertake valuable research that can lead to
greater insight into epilepsy and other conditions.”
Naomi RafaelTechnology and Systems
Manager, BioGrid Australia
Diagnostic Workstations
Diagnostic Workstations
Computed Radiography
Digital Modality Metadata,
reports in the database
DICOM Images on the file system
•File systems need specialized
•Security
•Backup
•Recovery
•…
•Metadata management
DICOM Application
Server
•DICOM Protocol
•Light-weight processing
Database Image Storage in a PACS System
Diagnostic Workstations
Diagnostic Workstations
Computed Radiography
Digital Modality
DICOM Application
Server
•DICOM Protocol
•Light-weight processing
Database Image Storage in a PACS System
Oracle
Multimedia DICOM
Diagnostic Workstations
Diagnostic Workstations
Computed Radiography
Digital Modality
DICOM Application
Server
•DICOM Protocol
•Light-weight processing
Database Image Storage in a PACS System
Oracle
Multimedia DICOM
Diagnostic Workstations
Diagnostic Workstations
Computed Radiography
Digital Modality
Metadata, reports and
DICOM images in the database
DICOM Application
Server
•DICOM Protocol
•Light-weight processing
Database Image Storage in a PACS System
Oracle
Multimedia DICOM
Diagnostic Workstations
Diagnostic Workstations
Computed Radiography
Digital Modality
Metadata, reports and
DICOM images in the database
•Metadata management
•Integrated search
•Validation
•Anonymization
•As fast as file systems
DICOM Application
Server
•DICOM Protocol
•Light-weight processing
Database Image Storage in a PACS System
Oracle
Multimedia DICOM
Diagnostic Workstations
Diagnostic Workstations
Computed Radiography
Digital Modality
Metadata, reports and
DICOM images in the database
•Database Supported
•Security
•Availability
•Backup
•Recovery
•…for all data
•Scalable: Grid, ASM, fast incremental backup …..
•Metadata management
•Integrated search
•Validation
•Anonymization
•As fast as file systems
DICOM Application
Server
•DICOM Protocol
•Light-weight processing
Database Image Storage in a PACS System
Oracle
Multimedia DICOM
Oracle Secure Enterprise Search
• Search across multiples of content: Patient notes, research papers, medical images
Oracle Universal Content Management
• Add DICOM content into Oracle Universal Content Management
• Extract DICOM metadata using Oracle Multimedia DICOM mid-tier APIs during check-in of content
• Use metadata for search in Oracle Universal Content Management
• Unable to share patients’ medical images between systems, labs, hospitals.
• Hard to search and manage.
ISSUES
Oracle Differentiator
Berkeley DB and Oracle Database 11gSometimes-Connected Mobile Health Care Devices
• Unable to share patients’ medical images between systems, labs, hospitals.
• Hard to search and manage.
ISSUES
• Centralized management in DICOM-tailored database -Oracle Database Multimedia.
• Digitization and telemedicine
Oracle Differentiator
TREND
Berkeley DB and Oracle Database 11gSometimes-Connected Mobile Health Care Devices
Oracle Database 11g Multimedia
DICOM
• Unable to share patients’ medical images between systems, labs, hospitals.
• Hard to search and manage.
ISSUES
• Centralized management in DICOM-tailored database -Oracle Database Multimedia.
• Digitization and telemedicine
• Doctors needs fast, easy access - local persistence.
• Multiple Data Sources needs embedded database
• Telemedicine - anytime and anywhere?
Oracle Differentiator
TREND CHALLENGES
Berkeley DB and Oracle Database 11gSometimes-Connected Mobile Health Care Devices
Oracle Database 11g Multimedia
DICOM
What’s Noteworthy?
• Data available to physicians on the go - secure
• Improved patient care at a lower cost• ACID transactions protect critical patient data
• Small– Footprint: 904KiB
– 60MiB DICOM files -> 600KiB database file
• High performance, low power• Reduce DICOM server’s traffic
• Synchronize DICOM data with Oracle Database
• Images might be annotated using multiple sets of terms• Might be annotated with
‘mandible’ but query with ‘Jaw’ should find the image
• Semantic relationships enhance keyword only search
Semantic Web Technology Search
Find me all DICOM images that contain the term ‘Jaw’
Semantic Relationships
Anatomical partSNOMED CodeAlias
MaxillaT11170Maxilla and
MandibleT-D1217Jaw Mandible
T-11180
SkullT-11100
Jaw=Maxilla+Mandible=T-D1217=T-11170+T-11180
source of image: www.wikipedia.org
Summary
• No, not really!
• Oracle Multimedia DICOM Infrastructure– With native support for DICOM eliminates the need to
manage DICOM data with special ways– Brings the advantage of the database to managing DICOM
data
– Unified management of image data with other patient data– One database, one data model, multiple applications
– Facilitates rapid development and deployment of a range of healthcare applications
Multimedia/DICOM at OOW 2009 - Sessions
Hilton Hotel Continental Parlor 1/2/3
Build Fast, Secure Web Applications with the PL/SQL Gateway and Oracle Multimedia
2:30 p.m.
Marriott Salon 5Scalable and Efficient Archives 11:30 a.m.
Moscone West L3 Room 3000
health engine: Image-Enabled Patient Health Records with Oracle Multimedia DICOM
11:30 a.m.
Hilton Hotel Golden Gate 2
Oracle Database Multimedia: Fast Performance and Deployment for Media Asset Applications
11:30 a.m.
Title
Tuesday, Oct. 13
LocationDate/Time
Multimedia/DICOM at OOW 2009 - Sessions
Moscone South Room 270
Oracle’s Multimedia DICOM API: Next-Generation Platform for Medical Imaging Solutions
11:45 a.m.
Moscone South Room 270
Overlook II: Mod PL/SQL Development Tips – including Google Maps, Spatial Integration and Sending HTML Formatted Emails with Graphics
1:00 p.m.
Title
Wednesday, Oct. 14
LocationDate/Time
•DEMOgrounds
•Enterprise Multimedia Management and Medical Imaging - Moscone West W-021
•Sun Microsystems (Scalable and Efficient Tiered Archive) – Moscone South - 1101
•LSI Systems (IBM Booth) – Moscone South - 733
top related