Beginners' Gallery: right, wrong and weird traces

General
#1

Right now I’ve done less than 100 cubes but every once in a while I run across really odd shapes that make me question my judgement. Hopefully I am one of a minority of EyeWire starters that experience this. Maybe this can be the place in the forum where beginners can post their headscratcher pictures and receive comments from experienced users.

helge-#01:

perhaps_a_bouton8uutw.png689×592 207 KB


helge-#02:
dead_end_115379wsbmi.png678×526 191 KB


helge-#03:
wrap_arounddazsx.png676×522 151 KB


helge-#04:


please move/delete this thread if you find it inappropriate or there is already a different place for submissions.


ps. how about a “submit a snapshot” function for views/cubes to be reviewed, combined with user galleries that contain the image plus classification/comment after review?

#2

This is a great idea!  You might also check out the Help What is This?!?  section.  


The Submit a snapshot function is actually something we’ve been angling for for a while.  It’s on the To Do list for the Devs (which is unfortunately super long).


#3

helge-#05:

flareaaujx.png1278×528 491 KB


red: given
magenta: added
blue: exploring (traces not belonging to the main object)

nothing makes sense anymore. Someone please explain. The “blob” is surrounded by many thin structures. What am I looking at? How are such objects to be segmented? I feel like eyewire is getting harder every day :frowning:

#4

Ok, I had a look at your 5 cubes :slight_smile: If you didn’t add the blue explore-mode part to cube #5 - I would have submitted exactly the same solutions like you did.


What the blob concerns… That should be a synapse, the junction between two neurons. The small black bubbles within that synapse could be vesicles containing the neurotransmitters. And according to the picture in the lower left corner of that wiki entry the flat blue part around the synapse is glia…

Woah, now I feel like such a neuroscientist :smiley:
#5

Thank you so much for pointing me in the right direction. Now I have so many beautiful things to read about :slight_smile:

As to the explore-mode: hopefully I discarded that selection before submitting; I just used the additional colour to make a partial inverse selection. Sorry for the confusion.

ps. stumbling upon this image http://www.glaucoma.org/uploads/cfc-glia-degrade-axons.gif where a glia cell is depicted taking out a plasma-filled vesicle. To which cell should the vesicle be assigned? Assigning it to the neuron would produce a nub, assigning it to the glia will tear a hole in the surface. Are these cases even relevant? Maybe I’m just lost in length scales by an order of magnitude or two.


#6

I’m not a glia expert and glia is seldom of interest to us, but what I can tell from the figure is that the vesicles belong to the glia. Obviously this schematic figure depicts the reverse process of a normal synapse works: the neurotransmitters are transferred from the ganglion cell axons to the glia, and then captured by the vesicle inside the glia. Note that there’s another fully enclosed vesicle on the left inside the glia. The “hole” you are mentioning is the reverse of the process called exocytosis. It is a process that vesicles are “fused” with the cell membrane to pour the neurotransmitters out of neurons. Here, the vesicle should open its surface by fusing itself with the cell membrane to embrace the neurotransmitters. 

#7

Thank you for the explanation. I’ll try and keep  the traces down to what definitely belongs to them. Sometimes it’s even worse than just vesicles attaching to membranes and opening inwards/outwards.

On to the next challenge:
We got ourselves another weird-looking cube here. I trimmed away as much as I could from the trace running back (upper left). It all looks so plausible when I am just judging by the 2D view. Some parts of the structure seem to penetrate the surroundings and “dissolve” (the cell wall boundary vanishes and the texures of “inside” and “outside” look just the same).
I know I should rely on the statistics that’ll put my (mis)judgement in perspective based on what all the better eyewirers out there are doing but this cube looks wrong and I’m not sure whether it wouldn’t have been better for me to just skip it. What is this thing here? Anyone please comment. Thanks!

#8

300+ cubes and there they are still. Those “alright, I quit” moments. What is part of the red “seed” and what isn’t? I’d call it a synapse but it won’t separate properly.

#9

Hey Helge, looks like your seed piece is red, and the added stuff is pink, right?  So on this one you’ve actually incorrectly added something quite large (no biggie, happens from time to time).  See how the part that starts from the red is a thin branch, and that the big thing is really rather bulbous.  They look like two different types of cells, right?  When the branch you’re working on suddenlt seems to turn into another type of cell/changes shape drastically  that’s a really good indication that you’ve added something erroneously.


Check out this screenshot, this is what the correct answer for that cube is.  (also thanks for doing the screenshot so I could see the number, it helped me figure out what was going on.

C

#10

I feel like I’m making a real mash-up of these. Could someone check the 3 or 4 cubes i’ve done. i don’t know how to submit them, but i figure you know how to call them up. i don’t know if i should even be attempting this.

#11

Hi Cestmarrant,


To check your cubes in future we’ll need your cube IDs.  To get those type /debug into chat.  That’ll pull up some information about the cube you’re in, including its ID.  Either post the cube and your question in here, or send it to support@eyewire.org, or if you’re in chat you can ask if any scouts are around, they can see into cubes as well.
#12

it’s been a while, so I might be a little ouf of practice so I present this supposed ladder structure to you. Merger or not, that is the question :slight_smile:


#13

Hi helge,


I had a look at the cube in your screenshot and you are correct. While it may seem that you added a lot of branches and built up a ladder structure, this is a correct tracing. You’ll find in Mystery Cells that most tasks have several “nubs” and branches that need to be added - this is typical of ganglion cells.  Starburst cells are less “branchy” and tend to travel out in a more regular straight pattern.  If you have another cube that you’re concerned if the trace is correct, feel free to post here or send us the screenshot to support@eyewire.com. You can always ask in chat and a scout/scythe/admin can assist you as well.

Best,
M.