Imaging Seminar: Image analysis scientific platforms to track cells in very large images
Event details
| Date | 19.05.2025 |
| Hour | 17:00 › 18:00 |
| Speaker | Dr Jean-Yves Tinevez, Institut Pasteur |
| Location | |
| Category | Conferences - Seminars |
| Event Language | English |
<< Registration below >>
Abstract:
Tracking is an image analysis technique that enables following single objects in movies. The task involves the detection of these objects in all frames of the movie, then linking these detections to reconstruct their trajectory.
Life is all about dynamics and therefore tracking of biological components over time is a big problem in biology. Motors travelling on microtubules; endosomes moving across the cell; cancer cells migrating on coverslips; thousands of cells flowing to build organs; entire organisms exhibiting swarm dynamics: to analyse these types of data and answer associated scientific questions, tracking is indispensable.
Every tracking problem is different due to the differences in underlying biology, microscopy and time scales. No software solution is able to address this diversity of challenges out-of-the-box. Therefore, there is a tremendous need for a flexible, performant and extensible platform to efficiently develop, prototype and deploy tracking solutions.
With collaborators we develop TrackMate and Mastodon, that form such a platform. TrackMate is a flexible and extensible tracking software made to harness the very diverse tracking scenarios encountered in Biology. Mastodon, on the other hand, has the unique capability of scaling to tracking millions of cells in 4D in big image data. Both are designed as central software hubs that allow developers to plug in different tracking algorithms and different analysis routines.
In this presentation, I will discuss the challenges they address, mainly going towards very large samples, integrating AI algorithms in tracking pipelines, and delivering a responsive interface facilitating data exploration, curation and analysis in the context of tracking.
The seminar is followed by an aperitif.
Registration appreciated
Look at our other events here
Abstract:
Tracking is an image analysis technique that enables following single objects in movies. The task involves the detection of these objects in all frames of the movie, then linking these detections to reconstruct their trajectory.
Life is all about dynamics and therefore tracking of biological components over time is a big problem in biology. Motors travelling on microtubules; endosomes moving across the cell; cancer cells migrating on coverslips; thousands of cells flowing to build organs; entire organisms exhibiting swarm dynamics: to analyse these types of data and answer associated scientific questions, tracking is indispensable.
Every tracking problem is different due to the differences in underlying biology, microscopy and time scales. No software solution is able to address this diversity of challenges out-of-the-box. Therefore, there is a tremendous need for a flexible, performant and extensible platform to efficiently develop, prototype and deploy tracking solutions.
With collaborators we develop TrackMate and Mastodon, that form such a platform. TrackMate is a flexible and extensible tracking software made to harness the very diverse tracking scenarios encountered in Biology. Mastodon, on the other hand, has the unique capability of scaling to tracking millions of cells in 4D in big image data. Both are designed as central software hubs that allow developers to plug in different tracking algorithms and different analysis routines.
In this presentation, I will discuss the challenges they address, mainly going towards very large samples, integrating AI algorithms in tracking pipelines, and delivering a responsive interface facilitating data exploration, curation and analysis in the context of tracking.
The seminar is followed by an aperitif.
Registration appreciated
Look at our other events here
Links
Practical information
- General public
- Free
Organizer
- EPFL Center for Imaging