Deep learning-enhanced light-field imaging with continuous validation

Visualizing dynamic processes over large, three-dimensional fields of view at high speed is essential for many applications in the life sciences. Light-field microscopy (LFM) has emerged as a tool fast volumetric image acquisition, but its effective throughput and widespread use biology been hampered by computationally demanding artifact-prone reconstruction process. Here, we present framework artificial intelligence–enhanced microscopy, integrating hybrid light-field light-sheet microscope deep learning–based volume reconstruction. In our approach, concomitantly acquired, high-resolution two-dimensional images continuously serve training data validation convolutional neural network reconstructing raw LFM during extended time-lapse imaging experiments. Our delivers high-quality reconstructions video-rate throughput, which can be further refined based on images. We demonstrate capabilities approach medaka heart dynamics zebrafish activity with rates up to 100 Hz. A algorithm enables efficient video rate. addition, concurrently acquired provide ground truth training, refinement algorithm.