Integrative functional optical imaging approaches: Optics, microfluidics, and machine learning for neuroscience in organoids and small-animal models

Advances in functional imaging have transformed neuroscience, enabling real-time mapping of neural activity and cellular dynamics. Techniques such as light-sheet microscopy allow whole-brain recordings in model organisms like Caenorhabditis elegans and zebrafish, revealing mechanisms of sensorimotor processing, learning, and neural circuit formation. More recently, the vast complexity of these datasets necessitates machine-learning tools for efficient analysis. Machine-learning-driven approaches improve data quality through denoising, automate segmentation of neurons and tissues, and enable analyses on complex data. By integrating machine learning with advanced imaging, researchers can decode developmental trajectories and neural-network function with unprecedented precision. This review explores the synergy between imaging and computation, highlighting how these innovations drive discoveries in neuroscience.