Deep learning for light-field microscopy with continuous validation

Light field microscopy is a powerful tool for fast volumetric image acquisition in biology which requires a computationally demanding and artefact-prone image reconstruction process. I will present a novel framework consisting of a hybrid light-field light-sheet microscope and deep learning-based volume reconstruction, where single light-sheet acquisitions continuously serve as training data and validation for the convolutional neural network reconstructing the LFM volume. Our framework produces video-rate reconstructions; their fidelity can be verified on demand and the network can be fine-tuned as necessary.