This week, at a pre-conference workshop for the Flux Congress, a group of neuroimaging researchers came together to discuss the current state of infant MRI. There were talks regarding data collection and analysis from researchers involved in large scale studies such as the Baby Connectome Project and the Developing Connectome Project, but also from individual labs and smaller research groups who are attempting to tackle these problems. 

The most reassuring aspect of this workshop was hearing from experts that issues of registration to standard spaces and the use of atlases were some of the major barriers in this field. These are exactly the issues I have struggled with when analyzing MRI data from toddlers. In a breakout session, I asked about the best processing pipelines for this age group, and the response I got was that we don’t have a standard processing pipeline yet. I was advised to try out different pipelines and find the a method that is specifically tuned for my data – visually scrutinizing every image at every stage of processing. One powerful message was that there is no “good enough” in this field – you need to find the ideal processing pipeline for your study. 

When asked which pipelines other researchers use, some reported manually registering every subject’s brain to a template, manually segmenting gray and white matter, and manually labeling regions of interest; others stated that we shouldn’t reinvent the wheel and instead should try to use the existing pipelines and provide feedback to the developers about issues that arise. 

In the end, the workshop introduced more questions than answers, but also confirmed that I am not the only one asking these questions. The future goal of FIT’NG is to act as a network of people who are all working together to apply what we know about neuroimaging to studies of fetal, infant, and toddler brain development. I am excited to be a part of this group and I hope that it moves the field forward and makes these days of “banging our heads against a wall” worth it.