Visualization of Incrementally Learned Projection Trajectories for Longitudinal Data

Abstract

Human-induced pluripotent stem cell derived brain organoids provide models to study the human brain and its diseases. The neuronal activities of these organoids are recorded over time by Multi-electrode Arrays, calling for approaches to analyze their longitudinal variations to study the organoids' electrophysiological maturation. We introduce a pipeline (Il-Vis) that incrementally learns and visualizes a progression trajectory representing the organoids' electrophysiological properties. Unlike a static model that runs at the end of the experiment, Il-Vis generates a progression trajectory thus far at each sampling timepoint. Results on simulated data, for which the true progression trajectories are known, verified Il-Vis's ability to capture and visualize the trajectories accurately and relative to each other. Results on brain organoid data revealed useful insights about organoids' electrophysiological maturation stages and response patterns when exposed to Quinolinic Acid (and its blocking antibody), a metabolite contributing to many neuroinflammatory diseases including Alzheimer's.