Fast and robust active neuron segmentation in two-photon calcium imaging using spatiotemporal deep learning

Significance Two-photon calcium imaging is a standard technique of neuroscience laboratories that records neural activity from individual neurons over large populations in awake-behaving animals. Automatic and accurate identification of behaviorally relevant neurons from these recordings is a critical step toward complete mapping of brain activity. To this end, we present a fast deep learning framework which significantly outperforms previous methods and is the first to be as accurate as human experts in segmenting active and overlapping neurons. Such neuron detection performance is crucial for precise quantification of population-level and single-cell–level neural coding statistics, which will aid neuroscientists to temporally synchronize dynamic behavioral or neural stimulus to the subjects’ neural activity, opening the door for unprecedented accelerated progress in neuroscientific experiments. Calcium imaging records large-scale neuronal activity with cellular resolution in vivo. Automated, fast, and reliable active neuron segmentation is a critical step in the analysis workflow of utilizing neuronal signals in real-time behavioral studies for discovery of neuronal coding properties. Here, to exploit the full spatiotemporal information in two-photon calcium imaging movies, we propose a 3D convolutional neural network to identify and segment active neurons. By utilizing a variety of two-photon microscopy datasets, we show that our method outperforms state-of-the-art techniques and is on a par with manual segmentation. Furthermore, we demonstrate that the network trained on data recorded at a specific cortical layer can be used to accurately segment active neurons from another layer with different neuron density. Finally, our work documents significant tabulation flaws in one of the most cited and active online scientific challenges in neuron segmentation. As our computationally fast method is an invaluable tool for a large spectrum of real-time optogenetic experiments, we have made our open-source software and carefully annotated dataset freely available online.

[1]  A. Oppenheim,et al.  Nonlinear filtering of multiplied and convolved signals , 1968 .

[2]  Fernand Meyer,et al.  Topographic distance and watershed lines , 1994, Signal Process..

[3]  Liam Paninski,et al.  Efficient and accurate extraction of in vivo calcium signals from microendoscopic video data , 2016, eLife.

[4]  Lorenzo Torresani,et al.  Learning Spatiotemporal Features with 3D Convolutional Networks , 2014, 2015 IEEE International Conference on Computer Vision (ICCV).

[5]  David Pfau,et al.  Simultaneous Denoising, Deconvolution, and Demixing of Calcium Imaging Data , 2016, Neuron.

[6]  A. Gamal,et al.  Miniaturized integration of a fluorescence microscope , 2011, Nature Methods.

[7]  Kilian Q. Weinberger,et al.  Densely Connected Convolutional Networks , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[8]  Andrea Giovannucci,et al.  NoRMCorre: An online algorithm for piecewise rigid motion correction of calcium imaging data , 2017 .

[9]  Mario Dipoppa,et al.  Suite2p: beyond 10,000 neurons with standard two-photon microscopy , 2016, bioRxiv.

[10]  Samy Bengio,et al.  Understanding deep learning requires rethinking generalization , 2016, ICLR.

[11]  Pengcheng Zhou,et al.  CaImAn an open source tool for scalable calcium imaging data analysis , 2019, eLife.

[12]  E. Pnevmatikakis,et al.  NoRMCorre: An online algorithm for piecewise rigid motion correction of calcium imaging data , 2017, Journal of Neuroscience Methods.

[13]  H. Sebastian Seung,et al.  Automatic Neuron Detection in Calcium Imaging Data Using Convolutional Networks , 2016, NIPS.

[14]  M. Stryker,et al.  Modulation of Visual Responses by Behavioral State in Mouse Visual Cortex , 2010, Neuron.

[15]  Terrence J. Sejnowski,et al.  An Information-Maximization Approach to Blind Separation and Blind Deconvolution , 1995, Neural Computation.

[16]  Attila Losonczy,et al.  SIMA: Python software for analysis of dynamic fluorescence imaging data , 2014, Front. Neuroinform..

[17]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[18]  Weijian Yang,et al.  In vivo imaging of neural activity , 2017, Nature Methods.

[19]  Rafael Yuste,et al.  Multi-scale approaches for high-speed imaging and analysis of large neural populations , 2016, bioRxiv.

[20]  Liam Paninski,et al.  Fast online deconvolution of calcium imaging data , 2016, PLoS Comput. Biol..

[21]  Liam Paninski,et al.  OnACID: Online Analysis of Calcium Imaging Data in Real Time , 2017, bioRxiv.

[22]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[23]  Lin Tian,et al.  Functional imaging of hippocampal place cells at cellular resolution during virtual navigation , 2010, Nature Neuroscience.

[24]  Konstantinos Kamnitsas,et al.  Efficient multi‐scale 3D CNN with fully connected CRF for accurate brain lesion segmentation , 2016, Medical Image Anal..

[25]  Sina Farsiu,et al.  Information-Theoretic Approach and Fundamental Limits of Resolving Two Closely Timed Neuronal Spikes in Mouse Brain Calcium Imaging , 2018, IEEE Transactions on Biomedical Engineering.

[26]  Yoshua Bengio,et al.  How transferable are features in deep neural networks? , 2014, NIPS.

[27]  Martin J. Wainwright,et al.  Noisy matrix decomposition via convex relaxation: Optimal rates in high dimensions , 2011, ICML.

[28]  Cordelia Schmid,et al.  Long-Term Temporal Convolutions for Action Recognition , 2016, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[29]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[30]  Matthew Eicholtz,et al.  Fast, Simple Calcium Imaging Segmentation with Fully Convolutional Networks , 2017, DLMIA/ML-CDS@MICCAI.

[31]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[32]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[33]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[34]  Thomas Brox,et al.  3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation , 2016, MICCAI.

[35]  Christine Grienberger,et al.  Imaging Calcium in Neurons , 2012, Neuron.

[36]  Trevor Darrell,et al.  Fully Convolutional Networks for Semantic Segmentation , 2017, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[37]  Rafael Yuste,et al.  Fast nonnegative deconvolution for spike train inference from population calcium imaging. , 2009, Journal of neurophysiology.

[38]  Pier Luigi Dragotti,et al.  ABLE: An Activity-Based Level Set Segmentation Algorithm for Two-Photon Calcium Imaging Data , 2017, eNeuro.

[39]  Stefan R. Pulver,et al.  Ultra-sensitive fluorescent proteins for imaging neuronal activity , 2013, Nature.

[40]  Ruijie Li,et al.  NeuroSeg: automated cell detection and segmentation for in vivo two-photon Ca2+ imaging data , 2017, Brain Structure and Function.

[41]  Quico Spaen,et al.  HNCcorr: A Novel Combinatorial Approach for Cell Identification in Calcium-Imaging Movies , 2017, eNeuro.

[42]  Christof Koch,et al.  A large-scale, standardized physiological survey reveals higher order coding throughout the mouse visual cortex , 2018, bioRxiv.

[43]  Joan Bruna,et al.  Training Convolutional Networks with Noisy Labels , 2014, ICLR 2014.

[44]  James E. Fitzgerald,et al.  Photon shot noise limits on optical detection of neuronal spikes and estimation of spike timing. , 2013, Biophysical journal.

[45]  Toru Aonishi,et al.  Detecting cells using non-negative matrix factorization on calcium imaging data , 2014, Neural Networks.

[46]  Sotirios A. Tsaftaris,et al.  Medical Image Computing and Computer Assisted Intervention , 2017 .

[47]  Mark J. Schnitzer,et al.  Automated Analysis of Cellular Signals from Large-Scale Calcium Imaging Data , 2009, Neuron.

[48]  Fred A. Hamprecht,et al.  Sparse Space-Time Deconvolution for Calcium Image Analysis , 2014, NIPS.

[49]  Dean C. Barratt,et al.  Automatic Multi-Organ Segmentation on Abdominal CT With Dense V-Networks , 2018, IEEE Transactions on Medical Imaging.

[50]  Seyed-Ahmad Ahmadi,et al.  V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation , 2016, 2016 Fourth International Conference on 3D Vision (3DV).

[51]  Nir Shavit,et al.  Deep Learning is Robust to Massive Label Noise , 2017, ArXiv.

[52]  Parashkev Nachev,et al.  Computer Methods and Programs in Biomedicine NiftyNet: a deep-learning platform for medical imaging , 2022 .

[53]  Amiram Grinvald,et al.  Accurate spike estimation from noisy calcium signals for ultrafast three-dimensional imaging of large neuronal populations in vivo , 2016, Nature Communications.