Light-field deep learning enables high-throughput, scattering-mitigated calcium imaging

Light field microscopy enables volumetric, high throughput functional imaging. However, the computational burden and vulnerability to scattering limit light field's application to neuroscience. We present a strategy for volumetric, scattering-mitigated neural circuit activity monitoring. A physics-based deep neural network, LNet, is trained with two-photon volumes and one-photon light fields. A processing pipeline uses LNet to extract calcium activity from light-field videos of jGCaMP8f-expressing neurons in acute cortical slices. The extracted time series have high signal-to-noise ratios and reduced optical crosstalk compared to conventional volume reconstruction. Imaging 100 volumes per second, we observed putative spikes fired at up to 10 Hz and the spatial intermingling of putative ensembles throughout 530 x 530 x 100-micron volumes. Compared to iterative algorithms, LNet LFM cuts light-field video processing time from hours to minutes and hence advances the goal of real-time, scattering-robust volumetric neural circuit imaging for closed-loop and adaptive experimental paradigms. ### Competing Interest Statement The authors have declared no competing interest.