Towards Automated Hypothesis Testing in Neuroscience

摘要

Scientific data generation in the world is continuous. However, scientific studies once published do not take advantage of new data. In order to leverage this incoming flow of data, we present Neuro-DISK, an end-to-end framework to continuously process neuroscience data and update the assessment of a given hypothesis as new data become available. Our scope is within the ENIGMA consortium, a large international collaboration for neuro-imaging and genetics whose goal is to understand brain structure and function. Neuro-DISK includes an ontology and framework to organize datasets, cohorts, researchers, tools, working groups and organizations participating in multi-site studies, such as those of ENIGMA, and an automated discovery framework to continuously test hypotheses through the execution of scientific workflows. We illustrate the usefulness of our approach with an implemented example.