Nessys (Nuclear Segmentation) Tutorial

nessys
segmentation

#1

A video tutorial is now available for our Segmentation tool Nessys : Nuclear Envelope Segmentation System

Here is a link to the tutorials
password: ‘nessys’

This is version 1 so comments are welcomed :yum:. (I know it goes a bit fast)


#2

Comments from Sally after watching version 1 of Nessys video tuturial. Some of these are just suggestions that may not be necessary- perhaps would be over-explaining- but I’m just putting them here just for consideration.

000:20 typo “croped” should be cropped

00:37 you moved the slide to change brightness (or contrast?)- might be good to explain that this is what you are doing.

01:07 typo “smooting” should be smoothing

01:56 Perhaps have an additional explanation when you are pointing to the dim cells with no maxima eg “we find no maxima on these dim cels so we must decrease the maxima threshold”

03:53 You click on ‘create best guess”. Perhaps explain what this does? Does it re-segment based on the new criteria?

04:31 Here you move to other z sections. I think it would have helped if you’d explained at the start of the training-of-classifier that we are focusing on only one z section so we can ignore ‘out of focus’ shapes. I think this may help explain why you seemed to be ignoring some non-segmented shapes before now.

05:30 I wonder it it is just worth pointing out here that you can edit the classifier after saving it… although this might not be necessary since there is a big button sitting there saying “edit current classifier”.

06:30 Here you tell us that some ‘merged’ nuclei are resolved when you do the area to volume correction. Would it help to note this earlier, during the training of the classifier? eg perhaps something saying “if you still detect some errors from merged nuclei, note that these are often resolved at the next stage when we link the 2D slices into 3D areas”

07.02 ‘split tolerance’… selected ‘intensity only’. It is not explained at all why you chose this option, but perhaps it is ok if this is explained in the pdf that you refer to.

07:57 missing word “…opened Fiji” instead of opened IN Fiji.


#3

just while I notice it, in the word document pdf

“Split Tolerance : The split will be performed if the score for a particular volume is an outlier with
respect to the scoe identified”

scoe should be “score”


#4

Thanks! If you see other typos or if something else could be improved, please let me know.