2020-07-29 –, Red Track
The field of neuroinformatics requires collaboration from highly skilled experts from many diverse fields.
We will briefly introduce the problem space and show how JuliaNeuroscience addresses this.
Neuroscience relies on the development of highly technical mathematical and computational techniques. These techniques are subsequently used and empirically validated and interpreted with the assistance of neuroscientists. This dynamic results in an additional step to the classic two-language problem, communication between fields that often have a very different knowledge base. Often this communication gap is managed by a common coding language. However, the increasing demand for incorporating multiple modalities and larger sample sizes into data analyses is incompatible with most languages used by neuroscientists (e.g., Matlab, R, Python). Consequently, neuroscientists are often using outdated techniques and those developing newer techniques struggle to reach broader adoption for empirically testing and improving their work. We will show how Julia can help solve the three-language problem through a common programmatic language and the interaction between generic code bases across different fields by introducing the NeuroCore package and related resources.
We will begin by providing a high-level overview of how NeuroCore provides a generic API and access to flexible types. We will provide examples illustrating how this design can be easily extended and incorporated by others. We will then provide brief examples of current and planned uses of NeuroCore, including plotting, graph theory, file format access, and data wrangling. Finally, we will show how this design will facilitate natural integration into other components of the Julia ecosystem.
I'm an MD/PhD candidate investigating predictive biomarkers for developmental psychopathology. I use magnetic resonance imaging, neurophysiology, and psychometrics to perform my research.