With recent advances in high-throughput Electron Microscopy (EM) imaging it is now possible to image an entire nervous system of organisms like Drosophila melanogaster. One of the bottlenecks to reconstruct a connectome from these large volumes (œ 100 TiB) is the pixel-wise prediction of membranes. The time it would typically take to process such a volume using a convolutional neural network (CNN) with a sliding window approach is in the order of years on a current GPU. With sliding windows, however, a lot of redundant computations are carried out. In this paper, we present an extension to the Caffe library to increase throughput by predicting many pixels at once. On a sliding window network successfully used for membrane classification, we show that our method achieves a speedup of up to 57×, maintaining identical prediction results.

Differentiable simulations of optical systems can be combined with deep learning-based reconstruction networks to enable high performance computational imaging via end-to-end (E2E) optimization of both the optical encoder and the deep decoder. This has enabled imaging applications such as 3D localization microscopy, depth estimation, and lensless photography via the optimization of local optical encoders. More challenging computational imaging applications, such as 3D snapshot microscopy which compresses 3D volumes into single 2D images, require a highly non-local optical encoder. We show that existing deep network decoders have a locality bias which prevents the optimization of such highly non-local optical encoders. We address this with a decoder based on a shallow neural network architecture using global kernel Fourier convolutional neural networks (FourierNets). We show that FourierNets surpass existing deep network based decoders at reconstructing photographs captured by the highly non-local DiffuserCam optical encoder. Further, we show that FourierNets enable E2E optimization of highly non-local optical encoders for 3D snapshot microscopy. By combining FourierNets with a large-scale multi-GPU differentiable optical simulation, we are able to optimize non-local optical encoders 170× to 7372× larger than prior state of the art, and demonstrate the potential for ROI-type specific optical encoding with a programmable microscope.

Identifying the input-output operations of neurons requires measurements of synaptic transmission simultaneously at many of a neuron’s thousands of inputs in the intact brain. To facilitate this goal, we engineered and screened 3365 variants of the fluorescent protein glutamate indicator iGluSnFR3 in neuron culture, and selected variants in the mouse visual cortex. Two variants have high sensitivity, fast activation (< 2 ms) and deactivation times tailored for recording large populations of synapses (iGluSnFR4s, 153 ms) or rapid dynamics (iGluSnFR4f, 26 ms). By imaging action-potential evoked signals on axons and visually-evoked signals on dendritic spines, we show that iGluSnFR4s/4f primarily detect local synaptic glutamate with single-vesicle sensitivity. The indicators detect a wide range of naturalistic synaptic transmission, including in the vibrissal cortex layer 4 and in hippocampal CA1 dendrites. iGluSnFR4 increases the sensitivity and scale (4s) or speed (4f) of tracking information flow in neural networks in vivo.

Identifying the input-output operations of neurons requires measurements of synaptic transmission simultaneously at many of a neuron’s thousands of inputs in the intact brain. To facilitate this goal, we engineered and screened 3365 variants of the fluorescent protein glutamate indicator iGluSnFR3 in neuron culture, and selected variants in the mouse visual cortex. Two variants have high sensitivity, fast activation (< 2 ms) and deactivation times tailored for recording large populations of synapses (iGluSnFR4s, 153 ms) or rapid dynamics (iGluSnFR4f, 26 ms). By imaging action-potential evoked signals on axons and visually-evoked signals on dendritic spines, we show that iGluSnFR4s/4f primarily detect local synaptic glutamate with single-vesicle sensitivity. The indicators detect a wide range of naturalistic synaptic transmission, including in the vibrissal cortex layer 4 and in hippocampal CA1 dendrites. iGluSnFR4 increases the sensitivity and scale (4s) or speed (4f) of tracking information flow in neural networks in vivo.

The Importance Weighted Auto Encoder (IWAE) objective has been shown to improve the training of generative models over the standard Variational Auto Encoder (VAE) objective. Here, we derive importance weighted extensions to Adversarial Variational Bayes (AVB) and Adversarial Autoencoder (AAE). These latent variable models use implicitly defined inference networks whose approximate posterior density qφ(z|x) cannot be directly evaluated, an essential ingredient for importance weighting. We show improved training and inference in latent variable models with our adversarially trained importance weighting method, and derive new theoretical connections between adversarial generative model training criteria and marginal likelihood based methods. We apply these methods to the important problem of inferring spiking neural activity from calcium imaging data, a challenging posterior inference problem in neuroscience, and show that posterior samples from the adversarial methods outperform factorized posteriors used in VAEs.

The mouse embryo has long been central to the study of mammalian development; however, elucidating the cell behaviors governing gastrulation and the formation of tissues and organs remains a fundamental challenge. A major obstacle is the lack of live imaging and image analysis technologies capable of systematically following cellular dynamics across the developing embryo. We developed a light-sheet microscope that adapts itself to the dramatic changes in size, shape, and optical properties of the post-implantation mouse embryo and captures its development from gastrulation to early organogenesis at the cellular level. We furthermore developed a computational framework for reconstructing long-term cell tracks, cell divisions, dynamic fate maps, and maps of tissue morphogenesis across the entire embryo. By jointly analyzing cellular dynamics in multiple embryos registered in space and time, we built a dynamic atlas of post-implantation mouse development that, together with our microscopy and computational methods, is provided as a resource.

Simultaneous recordings of the activity of large neural populations are extremely valuable as they can be used to infer the dynamics and interactions of neurons in a local circuit, shedding light on the computations performed. It is now possible to measure the activity of hundreds of neurons using 2-photon calcium imaging. However, many computations are thought to involve circuits consisting of thousands of neurons, such as cortical barrels in rodent somatosensory cortex. Here we contribute a statistical method for stitching" together sequentially imaged sets of neurons into one model by phrasing the problem as fitting a latent dynamical system with missing observations. This method allows us to substantially expand the population-sizes for which population dynamics can be characterized---beyond the number of simultaneously imaged neurons. In particular, we demonstrate using recordings in mouse somatosensory cortex that this method makes it possible to predict noise correlations between non-simultaneously recorded neuron pairs.

We present a method combining affinity prediction with region agglomeration, which improves significantly upon the state of the art of neuron segmentation from electron microscopy (EM) in accuracy and scalability. Our method consists of a 3D U-net, trained to predict affinities between voxels, followed by iterative region agglomeration. We train using a structured loss based on MALIS, encouraging topologically correct segmentations obtained from affinity thresholding. Our extension consists of two parts: First, we present a quasi-linear method to compute the loss gradient, improving over the original quadratic algorithm. Second, we compute the gradient in two separate passes to avoid spurious gradient contributions in early training stages. Our predictions are accurate enough that simple learning-free percentile-based agglomeration outperforms more involved methods used earlier on inferior predictions. We present results on three diverse EM datasets, achieving relative improvements over previous results of 27%, 15%, and 250%. Our findings suggest that a single method can be applied to both nearly isotropic block-face EM data and anisotropic serial sectioned EM data. The runtime of our method scales linearly with the size of the volume and achieves a throughput of ~2.6 seconds per megavoxel, qualifying our method for the processing of very large datasets.