Spectrum of Cyberenabled Radiology Informatics
Dr. Jun Ni, Ph.D. Associate Professor, Dept. of Radiology Carver College of Medicine The University of Iowa
Minnesota Supercomputing Institute for Advanced Research Computing, 2009
June 29, 2009
+ Information Technology
Medical Imaging --- Digital Radiology
Medical Imaging Software
Medical Imaging Hardware
Medical Imaging Display for Diagnostics
Orchestra of Future Health Care System
Patient Record System Patient Drug Delivery Patient Therapy Patient Health Care …
Five Year Federal Strategic Plan for IT Research and Development (NITRD)
Budgeting
$3 billion/per year for 5 years
NSF and NIH involved
Five Year Federal Strategic Plan for IT Research and Development (NITRD)
Cyber-enabled R&D Themes
Emergent Cognition, Beyond Virtual, and Trust and Confidence
Economy Energy Security Environment Health Care Education
Five Year Federal Strategic Plan for IT Research and Development (NITRD)
Emergent Cognition (Health Care)
Systems that use aggregated Web data, anonymized purchase trends, and other appropriate information to identify disease outbreaks Detect unanticipated outcomes in pharmaceutical use Study the influence of complex, interacting factors on health risks and benefits
Five Year Federal Strategic Plan for IT Research and Development (NITRD)
Beyond Virtual (Health Care)
Remote, real-time, secure and confidential patient-doctor interaction with effective examination and diagnostic capabilities
Trust and Confidence (Health Care)
Complex medical devices such as monitors, pacemakers, pumps, nanodevices, and robotic systems assembled into a patient-specific care network that can be verified and validated for formal certification, enabling confident delivery of individually-tailored acute care in a diversity of settings.
Discipline Identity
Imaging Informatics Medical Imaging Informatics (MII), Radiology Informatics One of medical informatics disciplines Data: medical images Technical driven: teleradiology/telemedicine
Job market: 70,000 on demand Iowa --- Hawkeye Radiology Informatics Minnesota --- Mayo Clinic and UMN?
Radiology Characteristics
Radiology is a data-intensive science and technology-driven specialty in medicine Radiologists leads imaging informatics Frontier in cancer diagnostics Proliferated applications:
Oncology, cardiology, dermatology, surgery, gastroenterology, obstetrics, gynecology and pathology, and other medical fields
Strong digital requirement and IT engagement
Digitization In Medical Sciences
Digital Pathology Digital Radiology P e t a b y t e s E-Health Initiatives/Linkages
40,000 BCE cave paintings bone tools
Electronic Medical Record
3500 writing 0 C.E. 1450 printing 1870 electricity, telephone transistor 1947 computing 1950 Late 1960s Internet
Digital Cardiology
paper 105
1993 The Web 1999 2003
Source: UC Berkeley, School of Information Management and Systems.
Theme --- Aims and Goal
What is current scope of MII?
A subspecialty of radiology that aims to improve medical imaging related discovery and technical services within the healthcare enterprise
Accuracy (methodology) Efficiency (workflow) Usability (feasibility or applicability) Reliability (accessibility) Sustainability Cost/performance
Its ultimate goal to improve health care systems
Data Subject
Study how medical images (within radiology and throughout medical enterprise) is
Acquired Archived Retrieved Processed Analyzed Enhanced Visualized Exchanged
Crossing
A multidiscipline Intersection with other fields:
Medical science Computer and information science Biomedical engineering Electrical Engineering Biological and physiological sciences Medical physics
History
Duration: more than 20 years ago
The concepts in late 70s Int. Society for Optical Engineering (SPIE) conference in 1981, Dr. Paul Capp from University of Arizona PACS in 1992 by SPIE addressed photo-electronic radiology
Cultural vision: Computer Applications in Radiology Revolution of practical radiology with rapid IT development
Demands: data explosion
Major Data Components
PACS (Picture Archiving and Communication System)
Four technical components:
Acquisition and Digitalization (AD) Networking and Communication (NC) Display and Visualization (DV) PACS Analysis and Processing
AD
NC
DV
AP
PACS Service Workflow
Digitalization Image acquisition devices Communication networks Storage archive units Data management system Workflow
Display workstations
Database
Computer processors
Relationship
Organizational Health Care System (HCS)
Health Care System Health Information System (HIS) RIS ePR (eMR)
Digital Imaging System (DIS) or PACS
Other digital clinic systems
PACS Challenges
Different regional and industrial interpretation, configuration, and implementation Different interfaces and prototypes Different standardization
DICOM, HL7, Other IT standards
Different image digitalization of modalities Different scopes
Image acquisition and management technology Data visualization or image display Network and communications Computer application software
PACS-IT Technical Components
hospital
registration
order exam
waiting room
final report on RIS
Traditional Genetic Radiology Workflow
modality
radiologist review
exam operation
Fetch report to HIS
Radiologist preview Paperwork film package send to office
Basic PACS Design
HIS and Database Imaging Modalities Database Gateway
Acquisition Gateway
PACS Controllers & Archive Server
Display Workstations Clients or users Researchers and physicians
Application Servers
Web Server
PACS Classification
Large scale (multiple module-based):
Local networked Module 1 Module 2 Module 3
Distributed multiple- modules within multiple services units; but with single health organization
PACS Classification
Super scale (enterprise-, cyberinfrastructure-, heterogeneous, distributed grid-based, cross organization, or even globally):
High speed network Module 1 of site A Module 2 of site A
Module 1 of Site B
Module 2 of site B
Generic Workflow
Complexity
Cultural, policy, environment, technique, …
Generic PACS
HIS Database Database Gateway Image Modalities Acquisition Gateway PACS Control & Archive Server Workstations Report
Application Servers
Web Server
Classification
Loosely classified into five models
Home-grow model Two team model Turnkey model Partnership model Application service provider (ASP) model
Home-grow (build own car)
Two-teams (special order)
Turn-key (car dealer sale)
Partnership (renting service)
Sizes of Some Common Medical Images
Modality Nuclear medicine (NM) Magnetic resonance imaging (MRI) Ultrasound (US) Digital subtraction angiography (DS) Digital microscopy Digital color microscopy Color light images Computed tomography (CT) Computed/digital radiography (CR/DR) Digital X-ray Digital mammography One Image (bits) 128 x 128 x 12 (2 bytes) 256 x 256 x 12 (2 bytes) 512 x 512 x 8 (1 byte) 512 x 512 x 8 (1 byte) 512 x 512 x 8 512 x 512 x 24 (3 bytes) 512 x 512 x 24 (3 bytes) 512 x 512 x 12 (2 bytes) 2048 x 2048 x 12 (2 bytes) 2048 x 2048 x 12 (2 4000 xbytes)x 12 (2 5000 bytes) # of image/exam 30-60 60-3000 20-240 15-40 1 1 4-20 40-3000 2 2 4 Size for 1 exam 1-2 MB >8 MB 5-60 MB 4-10 MB 0.25 MB 0.75 MB 3-15 MB >20 MB 16 MB 16 MB 160 MB
NM (128, 128) MRI (256, 256) CT (512, 512) DSA (1024, 1024) CR (2048, 2048) Mammogram (4096, 4096)
PACS Technical Concerns
Data Migration
Back-up archive Fault-tolerance Integration with legacy systems Fast wide-area networks Security
Cyber-enabled, service-oriented PACSs with thin technologies
PACS provider
Clinic and health care units
MII Challenges
Lack generic MII ontology (Philological Issue)
Systematic identification and classification of domain entities and existences, and entity relations (communications) No semantic languages for communications or workflows Loosely-defined terminology No linkage and leverage to knowledge, artificial intelligent (AI), decision making
Philology Ontology Wisdom
Cognitive Sciences Knowledge Decision Making Artificial Intelligent
Information Science Information
Metadata Metadata
Management
Metadata
Data
Data
Data
Data
MII Challenges
No standard protocols (Technical issues)
To facilitate the interoperation and communication among globally-distributed MII resources To deploy concurrent hardware and software solutions To utilize cyber-enabled high-speed networks
Short of education/training programs (Business issue)
To foster the next generation in digital health
New Paradigm or Envisioning
Ontological classification and entity relations Knowledge-driven Artificial intelligent Unprecedented capacity for handling massive data System integration and interoperation among various hospital/clinic systems Expansion of MII domain scope
Medical Imaging Informatics Scopes
Integration
Picture Archiving and Communication System (PACS) and Component Systems Imaging Informatics for the Enterprise Image-Enabled Electronic Medical Records or Electronic Patient Records Radiology Information System (RIS) and Hospital Information Management System (HIS)
Medical Imaging Informatics Scopes
Facilitation
Imaging Facilities Design Digital Imaging
Digital Radiography Modalities: Computed tomography (CT or XCT), Nuclear medicine (NM), Positron emission tomography (PET), Singlephoton emission computed tomography (SPECT), Ultrasound (US), Magnetic resonance imaging (MRI), Digital fluorography (DF), Digital subtraction angiography (DSA), etc.
Data Acquisition Digital imaging Archive/Retrieve Integrity and Security Image Data Compression
Medical Imaging Informatics Scopes
Processing
Image processing and enhancement 2D, 3D, 4D, Visualization and Multi-media Speech, Text, Image (Signature) recognitions Documentations CAD
Medical Imaging Informatics Scopes (Cont.)
Roadmap
Terminology, Methodology, System Dictionary, and Ontology Anatomic, biological, and physiological knowledge-based system Transforming the Radiological Interpretation Process (TRIP) Radiology Informatics Education
Medical Imaging Informatics Scopes (Cont.)
Standardization
Digital Imaging and Communications in Medicine (DICOM) Health Level 7 (HL7) Distributed Computing Protocols (W3C, Web Services)
Medical Imaging Informatics Scopes (Cont.)
Communication
Networking Security (network security and patient confidentiality, legal policies) Workflow and optimal data transformation
Medical Imaging Informatics Scopes (Cont.)
Interaction
Radiological Science
Knowledge refreshing Modality updating Digital equipments Health care policy …
Medical Imaging Informatics Scopes (Cont.)
Informatics
Scientific discovery Data mining Architecture design Knowledge boundaries
Medical Informatics Bioinformatics Health Informatics Nursing Informatics Pathology informatics …
Medical Imaging Informatics Scopes (Cont.)
Interaction (cont.)
Computer Science
Algorithms System architecture Software engineering Networking Database architecture High performance computing and distributed computing
Medical Imaging Informatics Scopes (Cont.)
Interaction (cont.)
Information Science (Information Management System)
Decision making Optimization Operation research MII system management Market and policy
Medical Imaging Informatics Scopes (Cont.)
Interaction (cont.)
Information technology
Information system implementation Hardware/software acquisition System ordering, maintaining and updating Information Collaborative tools (Tele-conference, meeting) Instructional tools Accessible tools …
Medical Imaging Informatics Scopes (Cont.)
Applications
Computer-Aided Detection and Diagnosis (CAD) Computer-aided surgery (CAS) or image-guided surgery (IGS) Services assessment (quality assurance and reliability analysis, cost/performance estimation, system measurement) Object-services-class-based medical imaging system Teleradiology/Telemedicine
Medical Imaging Informatics Scopes (Cont.)
Computation (ongoing or future projects)
Data-mining from medical image Artificial Intelligence in medical imaging Process modeling and simulations Digitalization of human body (mass-phantom system) Computational radiology System radiology
Projects
Medical Imaging & Radiology Informatics (MIRI)
Hawkeye Radiology Informatics (HRI)
http://www.uiowa.edu/~hri/ Radiology Informatics Domain Ontology (RIDO) Formal and logical representations and definitions of concepts, terms, and relationships between MII in radiology and biomedical informatics domains, and other medical science domains. Cyberinfrastructure-enabled Radiology Informatics (CIRI) Radiology Informatics Education and Training (RIET) http://www.uiowa.edu/~hri/education.html
Projects
Parallel Computing in Medical Imaging (PCMI)
http://www.uiowa.edu/mihpclab/projects_pcmi.html
Parallelism of Medical Imaging Processing
CT Reconstruction Segregation Registration Texturing and classification Enhancement Image compression Image data mining …
Projects
Modeling Biotransport in Biophysical System (MBBS)
http://www.uiowa.edu/mihpclab/projects_mbbs.html
Nanothermotheropy (nanoHyperthmia) Tumor growth and dynamics (computational oncology)
Optical Imaging Tomography and Applications (DITA)
http://www.uiowa.edu/mihpclab/projects_oita.html
Image Stereology and Clinic Applications (ISCA)
http://www.uiowa.edu/mihpclab/projects_isca.html
Couple Diffusions for Image Enhancement (DDIE)
http://www.uiowa.edu/mihpclab/projects cdie.html
Acknowledgements
Current sponsors
NIH (HPC medical imaging) NSF (HPC computations in nanotechnology) Intel (HPC) Microsoft (dual operating systems and web services-based workflow) IBM (Cell/BE) NVidia (CUDA for GPGPU cluster)
Reference
ACR (American Collage of Radiology) RSNA (Radiology of North American) SIIM (SCAR, Society of Imaging Informatics in Medicine) CARS (Assisted Radiology and Surgery), International ABII (American Board of Imaging Informatics) offers a national certification program that defines the standard for demonstrated knowledge and competence in medical imaging informatics.
It was founded by SIIM and the American Registry of Radiologic Technologists (ARRT).
Thanks Q&A