Third, deep learning applications are often limited by accessibility to computing power (high-end GPU). Second, an interactive user interface for deep learning, especially in the context of cell tracking, is lacking ( Kok et al., 2020 Wen et al., 2021). Because the quality and quantity of the training data are crucial for the performance of deep learning, users must invest significant time and effort in annotation at the start of the project ( Moen et al., 2019).
First, pre-trained models can be inadequate for new tasks and the preparation of new training data is laborious. In spite of its powerful performance, deep learning remains challenging for non-experts to utilize, for three reasons.
As deep learning is data-driven, it is adaptable to a variety of datasets once an appropriate model architecture is selected and trained with adequate data ( Moen et al., 2019). Recent progress in deep learning has led to significant advances in bioimage analysis ( Moen et al., 2019 Ouyang et al., 2018 Weigert et al., 2018). ELEPHANT yields accurate, fully-validated cell lineages with a modest investment in time and effort. 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). Successive prediction-validation cycles enrich the training data, leading to rapid improvements in tracking performance. This enables users to implement cycles of incremental learning starting from a few annotated nuclei. ELEPHANT provides an interface that seamlessly integrates cell track annotation, deep learning, prediction, and proofreading. We present ELEPHANT, an interactive platform for 3D cell tracking that addresses these challenges by taking an incremental approach to deep learning. Moreover, annotation, training, prediction, and proofreading currently lack a unified user interface. Its use in cell tracking is limited by the scarcity of annotated data for the training of deep-learning models. Deep learning is emerging as a powerful approach for bioimage analysis.