Transcript MIRC in the WHIMS Clinical Trial

WHIMS/ACCORD and MIRC
• Multi-institutional MR Studies of Cerebrovascular
Disease
– Protocol development
– Data acquisition QC
– Data archival
• Data distribution
– Data analysis
Action to Control Cardiovascular Risk in Diabetes
ACCORD-MIND
Memory in Diabetes
September, 2004
Lenore J. Launer
Laboratory of Epidemiology, Demography, Biometry,
National Institute on Aging
Diabetes and AD
RR
N. Manhattan, NY
AD
Dementia with stroke
Rotterdam, NL
AD
1.3
3.4
1.9
AD w/ CVD
3.0
Rochester, Minn
Hawaii
AD w/ CVD
1.4
Ott et al., 1999; Leibson et al., 1997, Luchsinger, 2001, Peila, 2001
1.9
ACCORD-MIND
• 640 SUBJECTS WILL PARTICIPANT IN THE MRI
COMPONENT
• THE SUBJECTS WILL BE RECRUITED FROM 3 FIELD
CENTERS
– COLUMBIA
– WAKE FOREST
– UNIVERSITY OF MINNESOTA
– CASE WESTERN RESERVE*
• MRI’S WILL BE PERFORMED AT
– Baseline period upon enrollment
– 48-month follow-up.
Women’s Health Initiative Memory
Study (WHIMS)
Effects of Hormone Therapy on Subclinical
Neurological Pathology
(WHIMS-MRI)
WHIMS-MRI Study
• Primary Goal
– Conduct a cross-sectional MRI substudy on approximately
1,450 women from the WHIMS E+P and E-only trials to
document the relative rate of subclinical infarcts associated
with E+P and E-only therapy
WHIMS MRI PROCEDURE
• STANDARDIZED ACQUISITION
– FULL BRAIN COVERAGE
– 1.5 or 3MM SLICE THICK/NO GAP
– SLICE ANGLE AND POSITION ALONG AC/PC
• ON SITE DATA ARCHIVE
• DATA TRANSMISSION TO MRQC CENTER
• Quarterly ACR TEST/SCANNER QC
MRI PROTOCOL- ACR QC TEST
• Gradient Uniformity
• Low Contrast Detectability
• High Contrast Spatial
Resolution
• Slice Thickness
• Signal-Noise (SNR)
• Ghosting Artifacts
WHIMS MRI PROTOCOL
PULSE SEQUENCES
• Series 1 – 3 plane Gradient echo localizer for positioning.
• Series 2 – Sag T1 mid-slice image to demonstrate anatomical location
of the AC/PC for slice angle and position.
• Series 3- Oblique Axial Spin Density/ T2-weighted from the vertex to
skull base parallel to the AC/PC plane.
• Series 4 - Oblique Axial FLAIR T2- weighted images matching slice
positions in Series 3.
• Series 5 - Oblique Axial 3D SPGR T1 weighted-images matching slice
positions in Series 3.
WHIMS MRI PROTOCOL
Sag localizer
Ax PD
Ax Vol T1 SPGR
Ax T2
Ax FLAIR
MIRC Clinical Trial Configuration
DICOM
FC
DICOM
PI
DICOM
FC
HTTP / HTTPS
DICOM
PI
DICOM
FC
WHIMS MRI Study
• Data Analysis
– Tissue segmentation
– Spatial normalization
• ROI labeling
– Structure/Function Correlation
• ROI based
• Voxel based
Data Analysis Flow Chart
Tissue Classification
Skull Stripping
T1 image Volume
Labeled and
Measured
ROI’s
Model
Model based ROI
Labeling (HAMMER)
WHIMS MIRC
• Enter WHIMS patient and study info at MR scanner.
• Send images to your local MIRC site as if it were a
PACS.
• The MIRC site will forward the images to the
principal investigator site(s) automatically.
Image Transmission
Current Methods
– Network based medical image transmission
• MIRC
• ACRIN
• Ftp
– Physical archival/storage media transmission
• CD-ROM (Raw format/Dicom CD)
• DAT tape (Raw format/Dicom)
• MOD
• Radiographic Films
Image Transmission from Imaging Facilities
to MRI Analysis Centers
• Problems with standard File Transmission Protocol (FTP)
• Security! No encryption, not HIPAA-compatible
• Blocked by enterprise firewall
• No imaging data property preservation (time, owner, etc.)
• Problems with other image data methods (CD-ROM, DAT, MOD, Films)
• Incorrect and/or missing data
• Unreadable format
• Transferring data ( lost shipping and delays)
• Damaged archived media
• Filmless institutions/Film is expensive
• Un-anonymized data sent (HIPAA requirements)
MIRC Image Transmission/Retrieval
from FC to Reading Center
– Clinical Trial Personnel
• MRI Reading Center
– PI, Coordinator, IT support
• MRI Field Center
– IT support/Service engineer
– MRI technologist
– Radiologist
WHIMS-MRI RECRUITMENT
(CROSS-SECTIONAL STUDY)
• 1450 SUBJECTS WILL PARTICIPATE IN
THE MRI COMPONENT (3 YEAR PERIOD)
• THE SUBJECTS WILL BE RECRUITED FROM
14 WHI PARTICIPATING CLINICS
WHIMS-MRI Field Centers
•
•
•
•
•
•
•
UNC-Chapel Hill
Columbus
Milwaukee
Minneapolis
New York
Pittsburgh
Penn- MRIQCC
•
•
•
•
•
•
•
Davis
Des Moines
Gainesville
Los Angeles
Nevada
Stanford
Worcester
MIRC Image Transmission Progress
• Since the initial start-up ( 09/04) of the WHIMS trial:
– All 14 FC have been contacted to verify a timeframe for
MIRC installation
– 6 FC completed with installation and transfer
– 8 FC pending transmission
– 2 FC approved for recruitment
– 1 FC recruitment begins 12/1/04
– Completion of all FC for MIRC installation and
recruitment anticipated by 12/31/04
MIRC Image Transmission
• Installation delays- WHIMS-MRI Field Centers
– IT Personnel timeframe
– FC Network/Firewall/Proxy Server issues (varies between sites)
• 1 FC : 1hour installation
• 1 FC : 2 days (proxy server issue)
• 4 FC : 3 days-1 week ( Network permission and protection)
– Transmission/retrieval issues at MR reading center (Penn)
• 1 FC : successful transmission of imaging data (approx. time
5 minutes from FC to MRQC)
• 1 FC : dicom elements re-identification
• Firewall issues at MR reading center (Network permission)
– 2 weeks for access to port and identify problems/issues
Image Transmission using MIRC
• Advantages:
– Automatic network-aware transmission/resending using standard
Secure Internet Protocol (HTTPS).
– User-friendly, requires minimal interference once installed at FC.
– Reduction in delays for sending and receiving of MRI image data
– Improved tracking methods/organization of data
• Disadvantages:
– Time consuming for multi-center trials: Initial set-up , ongoing
communication and installation of individual FC sites.
• Improvements: IT support and/or Network issues are identified at
each site prior to training session and installation time period.
• IT support for clinical research trials that specifically involve the
collection of imaging data among mulitple sites.