Single cell tracking based on Voronoi partition via stable matching
Chang, Y. H.; Linsley, J. W.; Lamstein, J.; Kalra, J.; Epstein, I.; Barch, M.; Daily, K.; Snyder, P.; Omberg, L.; Heiser, L.; Finkbeiner, S.
Show abstract
Live-cell imaging is an important technique to study cell migration and proliferation as well as image-based profiling of drug perturbations over time. To gain biological insights from live-cell imaging data, it is necessary to identify individual cells, follow them over time and extract quantitative information. However, since often biological experiment does not allow the high temporal resolution to reduce excessive levels of illumination or minimize unnecessary oversampling to monitor long-term dynamics, it is still a challenging task to obtain good tracking results with coarsely sampled imaging data. To address this problem, we consider cell tracking problem as "stable matching problem" and propose a robust tracking method based on Voronoi partition which adapts parameters that need to be set according to the spatio-temporal characteristics of live cell imaging data such as cell population and migration. We demonstrate the performance improvement provided by the proposed method using numerical simulations and compare its performance with proximity-based tracking and nearest neighbor-based tracking.
Matching journals
The top 5 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Collective Evolution Learning Model for Vision-Based Collective Motion with Collision Avoidance 94%
- Agent-based simulation for reconstructing social structure by observing collective movements with special reference to single-file movement 94%
- A new method of Bayesian causal inference in non-stationary environments 94%
Similar papers in this journal
- scPADGRN: A preconditioned ADMM approach for reconstructing dynamic gene regulatory network using single-cell RNA sequencing data 94%
- Friendly-rivalry solution to the iterated n-person public-goods game 94%
- Extended Graphical Lasso for Multiple Interaction Networks for High Dimensional Omics Data 93%
Similar papers in this journal
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.