International standard specification
Description:
Preferred ID: ILX:0105592 Type: term ID# ILX:0105592 Score: 23642.018
BIRN-INCF Derived Data Project
Description: The Derived Data Working Group is focused on understanding, addressing and solving the known difficulties of sharing and combining data across multiple sites after post-processing and/or data analysis. This WG is working closely with researchers at UC Irvine, Columbia, Duke, MIT, MRN and INCF on these goals:-Develop provenance and derived data extraction scripts for common neuroimaging analysis software and pipelines-Extend tools and specifications for international community requirements-Create/curate an initial neuroimaging terminology and ontology for derived data and common neuroimaging concepts-Augment/improve derived data import and query functionality for multi-site derived data federationsNeuroimaging data often involves many aspects of image and signal processing on the raw data which takes considerable time and expertise. Making such derived data available in a structured and well documented way is of interest to the international neuroimaging community. The availability of large multi-site datasets such as the Function BIRN schizophrenia datasets and the 1000 Functional Connectomes Project data provides researchers with well-documented and structured raw data. The addition of a similarly well-documented and structured form for derived data is the focus of this working group. Although the Function BIRN had made strides in developing the XCEDE XML schema for the structured storage and interchange of neuroimaging derived data, being able to extract both the derived results and the provenance of what happened to the data during the analytical manipulations directly from the software and pipelines performing the computations is missing. Without this piece, documenting derived data provenance and results becomes a burden on the researchers who are more interested in addressing scientific hypotheses than in documenting who the result was obtained for other users.To address the needs for providing derived data to the public community we have formed the derived data working group. This group brings together researchers with extensive experience with neuroimaging studies and specifically with large-scale data analysis and management using the more prevalent software packages in the field, e.g. SPM and FSL. This working group leverages the experience of the group both within BIRN and in communication with other researchers through the INCF to identify information needed for a provenance description to be useful. Face-to-face meetings to discuss needs and possible approaches were held in 2011 and 2012 at UC Irvine in January, at OHBM in June 2011, at the INCF Standards for Data Sharing task force meeting in September, and a fourth meeting in November at the annual meeting for the Society for Neuroscience.Once these details are established and there is consensus among the community, we will address the format for which this information will be stored. Initially we are focusing on XML schemas that integrate with the existing XCEDE XML schema (W3C Prov model) which describes neuroimaging data sets in general. After identifying the needed pieces of information and their storage format we will both work with the developers of the software packages to integrate the provenance schemas into their distributions or where not possible, write scripts to extract the information. All products of the derived data group are open-source and freely available to the community through the BIRN public wiki and INCF GitHub repositories.
Preferred ID: NLX:149615 Type: term ID# ILX:0101318 Score: 1.4979838
About
The SciCrunch Infrastructure was developed as a cooperative data platform to be used by diverse communities in making data more FAIR.
Contact Us
scicrunch.org