Coupled minimum-cost flow cell tracking for high-throughput quantitative analysis

Padfield, Dirk; Rittscher, Jens; Roysam, Badrinath

doi:10.1016/j.media.2010.07.006

Cited by 132 publications

(129 citation statements)

References 50 publications

Supporting

Mentioning

128

Contrasting

Order By: Relevance

“…We adopt the formulation of the matching problem suggested by [6] where a bipartite graph with coupled edges is set up to accommodate for topology changes.…”

Section: Trackingmentioning

confidence: 99%

“…The model by [6] is modified to honor only the biologically possible topology changes, namely that a cell can only split into two if it is undergoing mitosis. A cell is marked as a mitotic candidate if it in the near-past (15 time points = 75 minutes) has been detected as in the pre-mitotic stage using the detector described in Section 3.…”

Section: Cell Detected Where None Is Present (False Positive)mentioning

confidence: 99%

“…While there exist M 2 pairs of potential split candidates, this number can be heavily reduced by selecting only the top β percent pairs sorted according to mutual distance as proposed by [6].…”

Section: Cell Detected Where None Is Present (False Positive)mentioning

confidence: 99%

“…This is accomplished by a modification of the bipartite graph formulation of the matching problem proposed by [6]. The modifications introduced are 1) restricting topology changes to ensure cell division occur during cell mitosis and 2) acknowledging that segmentation errors are present and minimizing their disruptions to the cell lineage construction.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Pipeline for Tracking Neural Progenitor Cells

Vestergaard

Dahl

Holm

et al. 2013

Medical Computer Vision. Recognition Techniques and Applications in Medical Imaging

View full text Add to dashboard Cite

Abstract. Automated methods for neural stem cell lineage construction become increasingly important due to the large amount of data produced from time lapse imagery of in vitro cell growth experiments. Segmentation algorithms with the ability to adapt to the problem at hand and robust tracking methods play a key role in constructing these lineages. We present here a tracking pipeline based on learning a dictionary of discriminative image patches for segmentation and a graph formulation of the cell matching problem incorporating topology changes and acknowledging the fact that segmentation errors do occur. A matched filter for detection of mitotic candidates is constructed to ensure that cell division is only allowed in the model when relevant. Potentially the combination of these robust methods can simplify the initiation of cell lineage construction and extraction of statistics.

show abstract

“…We adopt the formulation of the matching problem suggested by [6] where a bipartite graph with coupled edges is set up to accommodate for topology changes.…”

Section: Trackingmentioning

confidence: 99%

Section: Cell Detected Where None Is Present (False Positive)mentioning

confidence: 99%

Section: Cell Detected Where None Is Present (False Positive)mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Pipeline for Tracking Neural Progenitor Cells

Vestergaard

Dahl

Holm

et al. 2013

Medical Computer Vision. Recognition Techniques and Applications in Medical Imaging

View full text Add to dashboard Cite

show abstract

“…Then, a deterministic solution, e.g. an optimization technique, links the corresponding targets between frames [3][4][5][6][7]. The performance of these algorithms is often sensitive to the detection algorithm and may degrade in the presence of complex target dynamics and highly cluttered detections resulting from very noisy sequences.…”

Section: Introductionmentioning

confidence: 99%

Multi-Target Tracking With Time-Varying Clutter Rate and Detection Profile: Application to Time-Lapse Cell Microscopy Sequences

Rezatofighi

Gould

et al. 2015

IEEE Trans. Med. Imaging

View full text Add to dashboard Cite

Abstract-Quantitative analysis of the dynamics of tiny cellular and sub-cellular structures, known as particles, in time-lapse cell microscopy sequences requires the development of a reliable multi-target tracking method capable of tracking numerous similar targets in the presence of high levels of noise, high target density, complex motion patterns and intricate interactions. In this paper, we propose a framework for tracking these structures based on the random finite set Bayesian filtering framework. We focus on challenging biological applications where image characteristics such as noise and background intensity change during the acquisition process. Under these conditions, detection methods usually fail to detect all particles and are often followed by missed detections and many spurious measurements with unknown and time-varying rates. To deal with this, we propose a bootstrap filter composed of an estimator and a tracker. The estimator adaptively estimates the required meta parameters for the tracker such as clutter rate and the detection probability of the targets, while the tracker estimates the state of the targets. Our results show that the proposed approach can outperform state-of-the-art particle trackers on both synthetic and real data in this regime.

show abstract

Monitoring of adherent live cells morphology using the undecimated wavelet transform multivariate image analysis (UWT‐MIA)

Juneau

Garnier

Duchesne

2016

Biotech & Bioengineering

View full text Add to dashboard Cite

Cell morphology is an important macroscopic indicator of cellular physiology and is increasingly used as a mean of probing culture state in vitro. Phase contrast microscopy (PCM) is a valuable tool for observing live cells morphology over long periods of time with minimal culture artifact. Two general approaches are commonly used to analyze images: individual object segmentation and characterization by pattern recognition. Single-cell segmentation is difficult to achieve in PCM images of adherent cells since their contour is often irregular and blurry, and the cells bundle together when the culture reaches confluence. Alternatively, pattern recognition approaches such as the undecimated wavelet transform multivariate image analysis (UWT-MIA), allow extracting textural features from PCM images that are correlated with cellular morphology. A partial least squares (PLS) regression model built using textural features from a set of 200 ground truth images was shown to predict the distribution of cellular morphological features (major and minor axes length, orientation, and roundness) with good accuracy for most images. The PLS models were then applied on a large dataset of 631,136 images collected from live myoblast cell cultures acquired under different conditions and grown in two different culture media. The method was found sensitive to morphological changes due to cell growth (culture time) and those introduced by the use of different culture media, and was able to distinguish both sources of variations. The proposed approach is promising for application on large datasets of PCM live-cell images to assess cellular morphology and growth kinetics in real-time which could be beneficial for high-throughput screening as well as automated cell culture kinetics assessment and control applications. Biotechnol. Bioeng. 2017;114: 141-153. © 2016 Wiley Periodicals, Inc.

show abstract

Coupled minimum-cost flow cell tracking for high-throughput quantitative analysis

Cited by 132 publications

References 50 publications

Pipeline for Tracking Neural Progenitor Cells

Pipeline for Tracking Neural Progenitor Cells

Multi-Target Tracking With Time-Varying Clutter Rate and Detection Profile: Application to Time-Lapse Cell Microscopy Sequences

Monitoring of adherent live cells morphology using the undecimated wavelet transform multivariate image analysis (UWT‐MIA)

Contact Info

Product

Resources

About