Graphical-Model Framework for Automated Annotation of Cell Identities in Dense Cellular Images

Abstract

25Assigning cell identities in dense image stacks is critical for many applications, for comparing 26 data across animals and experiment conditions, and investigating properties of specific cells. 27Conventional methods are laborious, require experience, and could introduce bias. We present 28 a generalizable framework based on Conditional Random Fields models for automatic cell 29 identification. This approach searches for optimal arrangements of labels that maximally 30 preserves prior knowledge such as geometrical relationships. The algorithm shows better 31 accuracy and more robust handling of perturbations, e.g. missing cells and position variability, 32 with both synthetic and experimental ground-truth data. The framework is generalizable across 33 strains, imaging conditions, and easily builds and utilizes active data-driven atlases, which 34 further improves accuracy. We demonstrate the utility in gene-expression pattern analysis, 35 multi-cellular calcium imaging, and whole-brain imaging experiments. Thus, our framework is 36 highly valuable to a wide variety of annotation scenarios including in zebrafish, Drosophila, 37 hydra, and mouse brains. 38 39 6 define cell specific features (unary potentials) and co-dependent features (pairwise potentials) 104 in the model. The basic model uses several pairwise relationship features for all pairs of cells, 105including binary positional relationships, angular relationship, and the Gromov-Wasserstein 106 discrepancy between cells in the image and an atlas. By encoding these features among all pairs 107 of cells, our fully-connected CRF model accounts for label dependencies between each cell pair 108 to maximize accuracy. Third, identities are automatically predicted for all neurons iteratively, 109 taking into account neurons missing in the image stack ( Supplementary Note 1.4). Duplicate 110 assignments are handled by calculating a label-consistency score for each neuron, removing 111

Fast deep learning correspondence for neuron tracking and identification in C.elegans using synthetic training

“…Using only position information, the method achieves 80.0% accuracy at tracking neurons within an individual and 65.8% accuracy at identifying neurons across individuals. Accuracy is even higher on a published dataset [2]. Accuracy reaches 76.5% when using color information from NeuroPAL.…”

“…An alternative is to take a discriminative modeling approach [26]. For example, recent work [2] has used conditional random fields (CRF) to directly parameterize a conditional distribution over neuron labels, rather than assuming a model for the high-dimensional and complex image data. CRF allows for a wide range of informative features to be incorporated in the model, such as the angles between neurons, or their relative anterior-posterior positions, which are known to be useful for identifying neurons [27].…”

“…Not only is the Transformer well-suited to learning features for the neural correspondence problem, it also obviates the need to straighten [31] the worm in advance. Until now, existing methods have either required the worm to be straightened in preprocessing [2,22] or explicitly transformed them during inference [23,24]. Straightening the worm is a non-trivial task, and it is especially error-prone for complicated poses such as when the worm rolls along its centerline.…”

Automated neuron tracking inside moving and deforming animals using deep learning and targeted augmentation

“…This is possible in animals like C. elegans where all neurons have unique identities, and can be done by building probabilistic models of the positions of all neurons using brain atlases that describe a given posture. For each brain image, one calculates the most likely distribution of unique neurons [17–19]. However, this approach is not easily extended to moving, deforming brains, and has only been applied to immobilized animals.…”

Toward a More Accurate 3D Atlas ofC. elegansNeurons

“…However, the inherent positional variability of neurons in the head and tail is now readily apparent from multicolor C. elegans strains, designed for neural identification, that were used to measure neuron positions and their variability across multiple animals ( Toyoshima et al, 2020; Varol et al, 2020; Yemini et al, 2021 ). These strains and concurrent algorithmic advances demonstrate how fluorescent-protein barcodes can be used to accurately determine neuron identities in volumetric images ( Bubnis et al, 2019; Nejatbakhsh et al, 2020; Mena et al, 2020; Chaudhary et al, 2021; Yu et al, 2021; Wen et al, 2021 ).…”