Make talk for OHBM Educational Session

Title of educational course: Reusing Public Neuroimaging Datasets

Title of talk: The devil is in the details: accessing phenotypic data for brain-behaviour relationships

Deadline to co-presenters: 17th May 2018

Date of presentation: 17th June 2018

Abstract: Open neuroimaging datasets are fantastic resources for the community to develop new methods and validate existing analysis pipelines. But when the brain images are accompanied by additional information about the individuals in the dataset, we enter the world of population neuroscience (Paus, 2010), of being able to link deep imaging phenotypes to individual measures of behaviour over time (Poldrack et al, 2015), or of linking in vivo brain measures to posthumous measures of gene expression (Whitaker et al, 2016; Vertes et al, 2016). This section of the course will focus on understanding the increased privacy and ethical requirements for sharing phenotypic data. Planning a study that uses “other people’s data” is often not a free lunch; in particular each individual study will store data in slightly different ways and there is high variability in how easy these information are to access and re-use. We’ll discuss the importance of finding metadata to accompany the data itself, and of maintaining a healthy skepticism of data you did not collect yourself. We will also brainstorm - using an interactive slido voting poll - ways in which users can feedback errors or the location of missing information to the owners of the dataset to ensure that bugs are caught and the documentation improved. The end will be a call for participatory collaborative research, encouraging users to become contributors to the excellent shared resources so that everyone in the field of neuroimaging benefits from their existence.

WhitakerLab / WhitakerLabProjectManagement

Make talk for OHBM Educational Session #133