Meeting 53, 2023-11-17
- Bengt Sennblad
Â
When is the meeting? | 17-November-2023 at 10.00 - 11.00 |
Where is the meeting? | Zoom video-conference: http://meet.nbis.se/aiaio |
What is the meeting about? | How to extract dynamic features from fixed cells images? |
Who is the responsible presenter? | Juliette Griffié, DDLS fellow, Stockholm University |
Abstract: Single molecule localisation microscopy (SMLM), with a spatial resolution of 10-30 nm, provides a unique insight into the cellular nanoscale organisation. It relies on the temporal separation of fluorophores' emission, which is collected over typically thousands of imaging frames. Thus this gain in resolution comes at a cost: long acquisition time and increased photo toxicity. As a result, most SMLM data sets consist of images of fixed cells. Here, we will discuss how to reliably extract dynamic features, such as the cell cycle, from these static images relying on unsupervised machine learning. We will demonstrate how generative models, conventionally used to produce synthetic data sets, can be tailored to the conversion of SMLM images to an information-rich point representation, in which otherwise lost features, including live cell information, can be decoded.
Slides: TBA (after seminar)
Â