Large-Scale Multi-Hypotheses Cell Tracking Using Ultrametric Contours Maps
Jordão Bragantini, Merlin Lange, Loïc Royer
TL;DR
This work tackles large-scale cell tracking in terabyte-scale 3D+t microscopy by jointly selecting segmentation hypotheses across time using an MTIoU-based ILP. It constructs per-frame hierarchies via hierarchical watershed, then optimizes selections with constraints for appearance, disappearance, and division to recover cell lineages. The approach scales through sparse weight computation, parallelization, and a windowed ILP strategy, and it can fuse multiple segmentation sources via contour-based representations. On public benchmarks, it achieves state-of-the-art results on several Cell Tracking Challenge datasets and demonstrates practical applicability to a 3.4 TB zebrafish embryo dataset, promoting annotation-efficient, scalable cell-tracking workflows.
Abstract
In this work, we describe a method for large-scale 3D cell-tracking through a segmentation selection approach. The proposed method is effective at tracking cells across large microscopy datasets on two fronts: (i) It can solve problems containing millions of segmentation instances in terabyte-scale 3D+t datasets; (ii) It achieves competitive results with or without deep learning, which requires 3D annotated data, that is scarce in the fluorescence microscopy field. The proposed method computes cell tracks and segments using a hierarchy of segmentation hypotheses and selects disjoint segments by maximizing the overlap between adjacent frames. We show that this method achieves state-of-the-art results in 3D images from the cell tracking challenge and has a faster integer linear programming formulation. Moreover, our framework is flexible and supports segmentations from off-the-shelf cell segmentation models and can combine them into an ensemble that improves tracking. The code is available https://github.com/royerlab/ultrack.