Computation/Statistics

点击翻译按钮获取中文摘要

Abstract – Computation and Statistics Core 002 Our Center proposes to have a Computation and Statistics Core intended to provide scientific support to both human and animal research projects. This Core will provide four services to the program team, and also disseminate our technology and generated data to the neuroscience community in general. The first service will be to provide exceptionally high quality acquisition and preprocessing protocols for brain imaging datasets, as well as to introduce new technical advances in the area. Human and rodent brain imaging data processing are complicated and require continued vigilance for proper assessment of quality of collected data, preprocessing of data, and proper selection and application of data analysis methods and tools. Separate pipelines will be provided for rodent and human data preprocessing, segmentation, registration to standard atlases, and advice will be provided for design and processing of collected data. This service will be used by Project 1 for human and rodent brain imaging data. The second service will be to provide advice and statistical tools for experimental design and data analysis to both our internal team members at the neuroscience community at large. The service will help ensure proper blinding of collected data, predetermining of group sizes based on power calculations. The third service will aid researchers in the design and implementation of multi-parameter statistical regression and Bayesian models of human and animal data, especially when combining multi-modal observations. Again, these services will be provided to all four projects as well as the broader scientific community. The fourth service will be disseminating our brain imaging techniques and collected data to the scientific community through our OpenPain portal, and also contacting other groups to encourage them to post their fMRI data on OpenPain, from studies that are already published and were funded through NIH grants, both human and rodent. Baliki and Griffith will lead this Core jointly. Dr. Baliki is a young scientist with extensive expertise in fMRI data analyses methods and in advancing these techniques. Dr. Griffith has expertise is human and behavioral data and has worked with the Apkarian group for many years in the capacity of statistical consultation and analyses. They will also oversee other technical staff involved in organizing OpenPain. The datasets and associated methods and software will provide valuable resources for characterizing human and rodent brain function, and their relationship to behavior.