Managing dynamic information in large multi-site, multi-species, and multi-discipline consortia is a challenging task for data management applications. Often in academic research studies the goals for informatics teams are to build applications that provide extract-transform-load (ETL) functionality to archive and catalog source data that has been collected by the research teams. In consortia that cross species and methodological or scientific domains, building interfaces which supply data in a usable fashion and make intuitive sense to scientists from dramatically different backgrounds increases the complexity for developers. In this work we have built a multi-species data management system which uses semantic web techniques based on the Neuroimaging Data Model (NIDM ;Figure). We find this approach enables a low-cost, easy to maintain, and semantically meaningful information management system, enabling the diverse research teams to access and use the data.
Citation: Keator, D.B. et al., (2017). A Semantic Cross-Species Derived Data Management Application. Data Science Journal. 16, p.45. DOI: http://doi.org/10.5334/dsj-2017-045
Our work on brain-based biomarkers of dementia in Down’s Syndrome has been selected for an oral presentation at the AD/PDTM 2017, the 13th International Conference on Alzheimer’s and Parkinson’s Diseases and Related Neurological Disorders. In this work we show how amyloid burden in the brain as assessed with Positron Emission Tomography (PET) predicts future clinical transition to dementia.
We’ve published a new paper using the Neuroimaging Data Model (NIDM) we’ve been developing for many years. NIDM was started by an international team of cognitive scientists, computer scientists and statisticians, including PIs of this project, to develop a data format capable of describing all aspects of the data lifecycle, from raw data through analyses and provenance. Our new published work shows how we’ve used NIDM to model mass univariate statistics in neuroimaging.
Maumet, C. et al. Sharing brain mapping statistical results with the neuroimaging data model. Sci. Data 3:160102 doi: 10.1038/sdata.2016.102 (2016).
For more NIDM info see: http://www.sciencedirect.com/science/article/pii/S105381191300596X
I am proud to announce this new P41 biotechnology research resource I am part of. The Center for Reproducible Neuroimaging Computation, seeks to implement a shift in the way neuroimaging research is performed and reported. Through the development and implementation of a FAIR (Findable, Accessible, Interoperable and Reusable, Wilkinson et al., 2016) technology stack that supports a comprehensive set of data management, analysis, and utilization frameworks in support of both basic research and clinical activities, our overarching goal is to improve the reproducibility of neuroimaging science and extend the value of our national investment in neuroimaging research.
Stay tuned for interesting training materials and workshops on reproducibility!
Abstract—Determining the location of crystals within detector
arrays (position profile) is a crucial part of system tuning in
Positron Emission Tomography (PET) scanners. It provides a
basis for the mapping of detected events and the most probable
crystal location to assign the event. Accurate assignments are
crucial for proper coincidence event processing and reconstructed
image resolution. In high resolution imaging systems, image
resolution is of primary importance and proper position profiles
are a critical part of system tuning. The High Resolution
Research Tomograph (HRRT) PET scanner is composed of 936
detector blocks each with 64 dual layer crystals arranged in
8×8 grids. Significant engineer time is spent fixing errors in the
automated position profile estimation software used for system
tuning. We have developed a probabilistic approach to position
profile estimation and applied it to the HRRT PET system. Our
approach is composed of a segmentation model for crosstalk
filtering, a prior over valid position profile configurations, and
a grid partitioning algorithm for crystal location finding. Our
model outperforms the manufacturer supplied position profile
estimation software, yielding a 39% decrease in mean squared
error rate as compared to a gold standard configuration in the
HRRT, while being a general solution applicable to many detector