Feature request: Tagging exposures with multiple label groups

CryoSPARC Live
1

Hello all,

the possibility to manually annotate exposures in the live pipeline with more than one (currently only “rejected“) annotation type would be quite useful to me. Group specific labelling could then be done with numerical keys “1“, “2”, “3“ …

Best,

Mike

1 Like
2

Hi @msleutel,

We are curious to understand your use case for such a feature. What types of things are you wanting to label? What would be the downstream steps you are interested in using the groups for?

Best,

Kye

3

Can’t speak for @msleutel, but I could see use cases for this - e.g. different particle species (small particles, big particles, no particles, aggregated particles) or behaviors, so one can later track back to see if they correlate with square size or grid region during screening

1 Like
4

I routinely collect datasets on complex, heterogeneous, unpurified samples (e.g. mixed population of bacteria). This often leads to multiple different structural targets within a single dataset that are only sporadically observed (i.e. on a small subset of the images). Using automated picking on the entire dataset often leads to any one target being “snowed“ under during 2D classification. So what I most often end up doing is going through the entire dataset manually to mark images with target X of interest as “rejected“, to then continue to only process the “rejected“ images. If there are X different targets in one dataset (I’ve had cases where X>4), I have to go through the entire dataset X times to define each subset of images corresponding to target i - a laborious and time-consuming process. Having the option to work with different labels would greatly cut down on pre-screening time for such cases.

2 Likes
5

Hi @msleutel and @olibclarke ,

Thanks for your suggestions and feedback, this feature request has been noted.

Best,

Kye

1 Like