Building Together: Open Tools, AI Segmentation, and Community Infrastructure for Volume EM data analysis

Presentation from IMC21 in Liverpool, given on Thursday 3rd September in the Enabling Technologies - Machine Learning in Biological Imaging session. Title Building Together: Open Tools, AI Segmentation, and Community Infrastructure for Volume EM data analysis Abstract Volume electron microscopy is generating ultrastructural datasets of unprecedented scale and complexity, but the tools and expertise needed to extract biological meaning from these data have struggled to keep pace. I'll begin by outlining some AI-based tools we have developed for analysing volume EM datasets, highlighting some of the challenges encountered when developing such workflows. Next, I will go on to describe the work of CCP-volumeEM (https://www.ccp-volumeem.ac.uk/), a UKRI funded community project for volume EM data analysis, which brings together developers, facility staff, and life scientists to share best practice and drive sustainable software development. This community effort aims to reduce the entry barrier for the reproducible use and development of the latest data science and AI tools by the volume EM community, and beyond. In addition to a range of software development efforts, I will describe our community consolidation and training initiatives, which are key elements in ensuring the best tools are in the hands of end-users.