Automated cellular structure extraction in biological images with applications to calcium imaging data

Abstract

Recent advances in experimental methods in neuroscience enable measuring in-vivo activity of large populations of neurons at cellular level resolution. To leverage the full potential of these complex datasets and analyze the dynamics of individual neurons, it is essential to extract high-resolution regions of interest, while addressing demixing of overlapping spatial components and denoising of the temporal signal of each neuron. In this paper, we propose a data-driven solution to these challenges, by representing the spatiotemporal volume as a graph in the image plane. Based on the spectral embedding of this graph calculated across trials, we propose a new clustering method, Local Selective Spectral Clustering, capable of handling overlapping clusters and disregarding clutter. We also present a new nonlinear mapping which recovers the structural map of the neurons and dendrites, and global video denoising. We demonstrate our approach on in-vivo calcium imaging of neurons and apical dendrites, automatically extracting complex structures in the image domain, and denoising and demixing their time-traces.

Rapid fluctuations in functional connectivity of cortical networks encode spontaneous

“…With the rapidly expanding efforts to link large-scale patterned cortical signals with sensory encoding, movement, and task performance 2,8,9,18,36,38 , several distinct strategies have been developed to analyze the spatiotemporal organization of network activity, including singular variable decomposition and non-negative matrix factorization 8,39 . Here, we show that functional parcellation of cortical regions 21 followed by diffusion embedding of time-varying correlations using a Riemannian geometry provides a robust means to quantify dynamic functional connectivity that accurately encodes fluctuations in behavior. Notably, we did not find similarly strong representation of sensory information in mesoscopic network activity, although this may reflect a lack of behavioral relevance for the stimuli as presented here.…”

“…After normalization and hemodynamic correction of imaging data (Supplemental Figure 1, see Methods) 1,20 , we segmented the cortex into functional parcels using a graph theory-based approach that relies on spatiotemporal co-activity between pixels (LSSC, Figure 1b) 21 . To compare similar cortical regions across experiments, we identified the LSSC parcel whose center of mass was closest to the center of mass for areas defined by the anatomy-based Allen Institute Common Coordinate Framework (e.g., VISp, RSPd, SSp, MOs, Figure 1b) 22 .…”

Data Processing of Functional Optical Microscopy for Neuroscience

“…Motion correction is typically the first step in the analysis pipeline, to register all frames in the imaging stack such that the neuronal components to be extracted occupy the same spatial footprint in all frames. Denoising can be applied as a preprocessing step either temporally 33 or spatially 68 to improve the detection of ROIs. Normalizing per-pixel time-traces, e.g., by z-scoring, can enhance dim cells, and improve cell detection.…”

“…This process, however, can be improved by modest noise filtering as a preprocessing stage. A number of methods (with example applications) are used across the literature, including median or lowpass filtering, 28,68 downsampling, PCA projection, 61,68 z-scoring, wavelet denoising, 33 other hierarchical models, 101 and deep learning-based denoising. 30,102 All these approaches make different noise and signal model assumptions and should be used judiciously.…”

Neural anatomy and optical microscopy (NAOMi) simulation for evaluating calcium imaging methods

“…For instance, while imaging deeper into scattering tissue with TPM can benefit from decreasing the excitation numerical aperture (NA) [15], it is unknown how this benefit interacts with other optical or experimental design choices, such as adaptive optics [25,26] or dendritic imaging [27,28]. Additionally, while many algorithms have been designed to extract the neural activity traces and spatial profiles from TPM data [29][30][31][32][33][34][35][36][37][38][39][40][41][42][43], few options exist to assess the fidelity of the inferred segmentation beyond comparisons to manually annotated data [24,44,45].…”