Tracking cell lineages in 3D by incremental deep learning

Abstract

Deep learning is emerging as a powerful approach for bioimage analysis. Its use in cell tracking is limited by the scarcity of annotated data for the training of deep-learning models. Moreover, annotation, training, prediction, and proofreading currently lack a unified user interface. We present ELEPHANT, an interactive platform for 3D cell tracking that addresses these challenges by taking an incremental approach to deep learning. ELEPHANT provides an interface that seamlessly integrates cell track annotation, deep learning, prediction, and proofreading. This enables users to implement cycles of incremental learning starting from a few annotated nuclei. Successive prediction-validation cycles enrich the training data, leading to rapid improvements in tracking performance. We test the software’s performance against state-of-the-art methods and track lineages spanning the entire course of leg regeneration in a crustacean over 1 week (504 timepoints). ELEPHANT yields accurate, fully-validated cell lineages with a modest investment in time and effort.

Cellpose 2.0: how to train your own model

“…We have shown in the previous section that good models can be obtained with relatively few training images when starting from the Cellpose pretrained model. We reasoned that annotation times can be reduced further if we used a ‘human-in-the-loop’ approach 6 , 19 , 32 . We therefore designed an easy-to-use, interactive platform for image annotation and iterative model retraining.…”

“…The annotation/retraining process can also be repeated in a loop until the entire dataset has been segmented. This approach has been demonstrated for simple ROI such as nuclei and round cells, which allow for weak annotations such as clicks and squiggles 18 , 19 , but not for cells with complex morphologies that require full cytoplasmic segmentation. For example, using an iterative approach 19 , a 3D dataset of nuclei was segmented in approximately one month.…”

“…This approach has been demonstrated for simple ROI such as nuclei and round cells, which allow for weak annotations such as clicks and squiggles 18 , 19 , but not for cells with complex morphologies that require full cytoplasmic segmentation. For example, using an iterative approach 19 , a 3D dataset of nuclei was segmented in approximately one month. It is not clear whether the human-in-the-loop approach can be accelerated further, and whether it can in fact achieve human levels of accuracy on cellular images.…”

Following cell type transitions in space and time by combining live-cell tracking and endpoint cell identity in intestinal organoids

“…The copyright holder for this preprint this version posted June 29, 2022. ; https://doi.org/10.1101/2022.06. 27.497728 doi: bioRxiv preprint systems [13][14][15][16] . Yet, video microscopy has far been limited in studying spatiotemporal differentiation programs.…”

“…Techniques such as confocal and light-sheet imaging have visualized the dynamics of cell proliferation and collective cell migration in major developmental transitions [11][12][13] . Machine learning has more recently enabled automated tracking of individual cells in time, in developing embryos as well as in diverse organoid systems [13][14][15][16] . Yet, video microscopy has far been limited in studying spatiotemporal differentiation programs.…”