Matthew Brett

There is a wide gap between the ideal and the reality of scientific culture in neuroimaging. This culture gap is one reason of many to suggest that a large proportion of published imaging studies are wrong. Reasons for the culture gap include the perverse nature of academic rewards,the lack of systematic education in the foundations of imaging, and the complexity of imaging software.

Dr. Brett spoke about work we done at Berkeley to redesign education in neuroimaging with an emphasis on algorithmic thinking, sound techniques for organizing code and data, and mathematical principles of the analysis. Courses finish with projects that must be reproducible from downloading the data to the figures in the report. Dr. Brett also discussed the problem of culture in neuroimaging software, and argue that we can do more to give our students ownership of their analysis.