Reference Image Database to Evaluate Therapy Response (RIDER)
1. Executive Summary
The RIDER database is a targeted data collection for the purpose of generating an initial consensus on how to harmonize data collection and analysis for quantitative imaging methods as applied to measure the response to drug or radiation therapy. The long term goal is to provide a resource to permit harmonized methods for data collection and analysis across different commercial imaging platforms, as required to support multi-site clinical trials, using imaging as a biomarker for therapy response. Thus the database should permit an objective comparison of methods for data collection and analysis as a national and international resource as described in the first RIDER white paper report (2006):
All the image data are DICOM compliant and the image, annotations and meta data formats meets all the requirements for caBIG and the NBIA. The data collection has two phases as described below.
2. RIDER PILOT PROJECT (2005-2007)
This data collection was originally supported under supplemental funding for the LIDC U01 project and focused on the collection of longitudinal studies using X-ray CT for monitoring the response to therapy. The data came primarily from MDACC and several of the LIDC academic sites. The data is not annotated. However it contains a sub-set of 30 longitudinal cases that are annotated using the RECIST criteria.
3. RIDER Contracts: 2007-Beyond
NCI has exercised a series of contracts with specific academic sites for collection of repeat "coffee break" and longitudinal phantom and patient data for a range of imaging modalities (currently CT, PET CT, DCE MRI, DW MRI) and organ sites (currently lung, breast, and neuro). The goals are as follows:
- Develop a consensus on requirements for quantity assurance methods for longitudinal studies using phantom data as applied to each modality above,
- Develop a consensus on the stability of imaging platforms using repeat and longitudinal phantom measurements over the time period that therapy would be exercised,
- Develop a consensus on methods to measure the minimum change that can be measured using repeat and longitudinal patient or volunteer studies
- Provide access to results of measurements performed on these databases by each academic site to encourage a comparison on methodologies,
- Provide consensus-based juried publications to encourage a broad acceptance of the methods described above.
- Provide a resource for NCI research networks that address quantitative imaging such as the Quantitative Imaging Network QIN): http://grants.nih.gov/grants/guide/pa-files/PAR-08-225.html
The databases are being provided with a timely goal of 18 months time frame, namely from the time the contract was initiated and the published results within less than 2 years. The methods for data collection and analysis including results are described in the new Combined RIDER white Paper Report (Sept 2008):
The RIDER project will be replaced when the QIN initiative is fully implemented as this research network will be tasked to create database resources collected from phase 1-3 clinical trials, where clinical outcomes will be included in the meta data: http://grants.nih.gov/grants/guide/pa-files/PAR-08-225.html
3.1 ACADEMIC SITE COLLECTIONS: Focus on methods for data analysis
- Repeat CT Measurements: Human subjects: Lung
- Download the related lesion notes: MSKCC RepeatCT Lesion notes for RIDER.xls
- Download the related publication: Zhaob-RepeatCT Radiology2009.pdf
UNIVERSITY OF WASHINGTON
- Repeat measurements: PET/CT phantoms
- Longitudinal PET/CT human studies: Lung
- Repeat human subject studies: Neuro
- Dynamic Contract Enhanced studies: DCE MRI:
- Diffusion weighted imaging: DWI MR
- Diffusion tensor imaging: DT MRI.
UNIVERSITY OF MICHIGAN
- Repeat measurements: Human subjects: Breast
- DCE MRI
- ISMRM 2009 poster demonstrates how each of the "coffee break" exams were used as an estimate of each patient's null hypothesis, i.e. distribution associated with no change, and thus supports the estimate of the null's 97.5 percentile for subsequent estimation of early response to neoadjuvant chemotherapy on an individual patient basis.
- Repeat measurement: Human studies: Neuro and Liver
- DCE MRI
3.2 ACADEMIC SITE COLLECTIONS: Focus on harmonized methods for data collection
MULTI SITE STUDY (Total of 20 sites): Organized under a contract with the University of Michigan ( Under development: 2009)
- Repeat measurements
- Diffusion Weighted Imaging (DWI)
- Phantom measurements
- Human subjects
- White Paper: http://www.neoplasia.com/pdf/manuscript/v11i02/neo081328.pdf
4. SOCIETY DATA COLLECTIONS
RSNA QIBA AND NCI (RIDER and IRAT) (2009-beyond: under development)
- Multi Site DCE MRI phantom studies using a modified ADNI phantom (http://www.loni.ucla.edu/ADNI/ )
5. OTHER FEDERAL AGENCIES
FDA CDRH: Partly funded by NCI and NIBIB
- Phantom images expected to be loaded to NBIA by end of 2009.
- NIST has NCI NBIA data and data from other data sources
- See Biochange 2008