MICCAI 2020

3 Erdem V. and Amin N. 11 DAILY Mo n d a y building problem, the team are studying the viability of neuronal positions in the worm body across different animals. For the signal extraction part, they are trying to understand how neural activity is giving rise to behavior. That is a fundamental problem in systems and computational neuroscience . The team had a ‘ eureka ’ moment when they used a very new area of computational research called optimal transport . They applied it to a problem that it had not been applied to before – segmenting neurons – and it yielded excellent results . “This problem is more challenging than the usual problems people deal with and there are several reasons for this,” Amin explains. “The images you get do not have a perfect spatial resolution. As As each of these tasks are so involved, they have decided to dedicate a paper to each of them! “Each of these papers are like an atomic part of analysis and data extraction from videos,” Amin tells us . “ A similar pipeline is developed for humans . For example, fMRI data analysis, where you start with a bunch of images from the brain, build an atlas, and then go ahead and try to extract signals. We thought that some of the methods we were developing were applicable to other model organisms, so we wanted to keep it as generic as possible so that the community could at least use some of the building blocks of our method on other types of problems or animals.” In the context of neuroscience , people usually start with simple questions and simple model organisms. In the atlas