|
Author |
Thread Statistics | Show CCP posts - 17 post(s) |
|
HPA Illuminator
State War Academy Caldari State
10
|
Posted - 2016.03.11 06:58:53 -
[1] - Quote
Terminal Insanity wrote:Im pretty sure you selected the correct one... its just gonna take time for people to get good at this, and for the system to weed out the bad ones =p was i was right in selecting the two categories for this one? Previously i was only selecting Nucleus on samples like this because i felt it was the prominent feature, but now as ive went though some of these and getting more comfortable, im thinking this should be both Nucleus and Cytoplasm? http://i.imgur.com/iYVeStn.pngI went with Cytoplasm and not Vesicles because the staining in the red part was more blurry, not really defined dots Im also curious what the difference between an "Abnormal sample" and "cell to cell variation"/"not identifiable" would be
Difficult one. It's really a borderline regarding the cytoplasm. Nucleus, absolutely correct. Cytoplasm... maybe (it's definitely not vesicles, you're correct there). It could just be "background" staining since it's so weak. So, I can understand why it has a 50% ratio :) Personally, I would go for only nucleus, but I have looked at *a lot* of images, so I'd actually say it's better to label everything you think matches.
Abnormal sample: you find a distinct pattern in the cells that doesn't match any of the categories. OR it's a broken image
Cell to cell variation: When you see e.g. a strong nucleus staining in a few cells, but some are really weak. Or some cells stain nucleus and some stain mitochondria (or whatever). When things differ visibly between cells.
Not identifiable: You can't distinguish any pattern, the green is just all over the place in the cell and everything is just the same shade of green. |
|
|
HPA Illuminator
State War Academy Caldari State
10
|
Posted - 2016.03.11 07:07:25 -
[2] - Quote
Nevyn Auscent wrote:http://i.imgur.com/tkkrx53.png Try and work that one out for a laugh.
LOL! It's either imaged on top of the cell (= wrong focus), or it's a dying cell. If everything looked like that, it's clearly an "unidentifiable". |
|
|
HPA Illuminator
State War Academy Caldari State
10
|
Posted - 2016.03.11 07:12:13 -
[3] - Quote
Van Dracon wrote:The new feature has caused many inconsistent results. Especialy with basic findings where you are 100% accurate but get a failed result. This has baffled me if I should continue wasting my time in project failure.
For starters the pictures should be increased in size instead of looking through a key hole.
Can we get 3d images if possible.
Why did I fail with one of the results. Please give back users more explanatory information to educate them more on failed results instead of guessing.
I like to help like many others though I believe the new feature needs to be expanded with more features. At this stage until I see improvement I won't be committed to it as I once thought.
I think (not sure) that there's a forum thread that's better for giving feedback regarding the UX/UI. I'll ask ppl to have a look in this forum too though.
3D images is not possible atm as we've so far only acquired 2D images. The reason is that taking 3D just takes too much time (atm 1 image takes around 3s by the microscope, just the actual acquisition, if it were to be 3D it would be a minute or so... that times 150k samples = no doable, unfortunately. It would be awesome to have though!). |
|
|
HPA Illuminator
State War Academy Caldari State
25
|
Posted - 2016.03.12 21:33:39 -
[4] - Quote
Beta Maoye wrote:In this classification result, I realized I was wrong about Cytoplasm for so many cases.
Not sure what you mean? Have added it to our list of possible errors, will double check whether it should have the cell-to-cell variation as a class. Will get back! |
|
|
HPA Illuminator
State War Academy Caldari State
25
|
Posted - 2016.03.13 07:28:33 -
[5] - Quote
Gilbaron wrote:https://i.gyazo.com/54efcad7e2f4666f6ad66988a2e17ceb.png
Was this a question? I would say it's either cell junctions of plasma membrane. I would lean towards plasma membrane because in the upper left and lower right parts you can see some staining outside the cells. Possibly both CJ+PM is correct, that's also an option :) |
|
|
HPA Illuminator
State War Academy Caldari State
26
|
Posted - 2016.03.14 22:05:02 -
[6] - Quote
Helios Anduath wrote:They don't really look like Centrosome because they are not two very well defined spots side by side. I would just have gone MTOC for this.
Based on zoom 2 I'd go for centrsosome and think the others cells were slightly out of focus. |
|
|
HPA Illuminator
State War Academy Caldari State
26
|
Posted - 2016.03.15 07:07:20 -
[7] - Quote
Memphis Baas wrote:I had that one just now; the splash of green overlaps the nucleus on every cell, so it can't really be anything outside the nucleus. I put Nuclear Membrane, Abnormal sample. A Google images for "abnormal nuclear membrane stain" shows a couple examples where the membrane could fold in on itself or maybe shrink, to look slightly like that.
<3
Something happened to cells during prepp, so what you see is not something that is actually specific. I don't think it't the nuclear membrane but rather the plasma membrane that has shrunk (the nucleus still intact).
Have added to checkup list. |
|
|
HPA Illuminator
State War Academy Caldari State
27
|
Posted - 2016.03.16 20:18:03 -
[8] - Quote
PAPULA wrote:Ok so, this game is totally wrong. look at this sample: http://www.netsky.org/eve/notcor2.png As you can see it's wrong, but it's not. I chose Nucleoplasm because it is Nucleoplasm, but game says no it is not. And the "correct answer" was: Nuclear speckles
It's a known error. Really sorry for it, we're working on fixing the samples that incorrect and hope to have fixed it by this week. *crossing fingers* Will let ppl know when done. |
|
|
HPA Illuminator
State War Academy Caldari State
28
|
Posted - 2016.04.01 10:36:47 -
[9] - Quote
Rakin o'Dubois wrote:I got the exact same image and totally agree with your classification. I did the same. I wonder how much screwed the baseline is due to the community consensus approach. Is there anything known on how they avoid those effects? Cheers Rakin
Sorry about that, it's a known error that we will correct asap. |
|
|
HPA Illuminator
State War Academy Caldari State
28
|
Posted - 2016.04.01 10:38:26 -
[10] - Quote
Lovely sample! We don't have ciliated cells (unfortunately), so what you see here are membrane protrusions (which is really unusual to find, most cell membrane stainings don't have them). |
|
|
|
HPA Illuminator
State War Academy Caldari State
28
|
Posted - 2016.04.01 10:47:02 -
[11] - Quote
Hm, do I understand it right that the classification was supposed to be nucleoplasm+ER? That's odd, because it should only be ER. I've submitted it to have the nuc removed (and add variations).
It should only be ER as the weak nuclear staining that can be seen is due to background bleedthrough from ER below the nucleus. The green in the nucleus is way to weak to be a true nuclear staining. |
|
|
HPA Illuminator
State War Academy Caldari State
28
|
Posted - 2016.04.01 10:48:46 -
[12] - Quote
Yllej Gniht wrote:Image no. 100000325Classification Result says: "Nuclear Speckles"; it sure looks like "Nucleoplasm" to me. What's the deal here? Is this faint variation enough to declassify it as "nucleoplasm" and flag it as "nuclear speckles"? If so, some explaining text would be appreciated with those very hard to understand "teaching moments". Thank you! :)
This is also one of the known errors, it should indeed be nucleoplasm and nuclear speckles (at first we thought they were mutually exclusive, which is why this one happened). It will be corrected asap. |
|
|
HPA Illuminator
State War Academy Caldari State
28
|
Posted - 2016.04.01 10:51:09 -
[13] - Quote
Nucleoplasm for sure, and probably cytoplasm too. Or what part do you think was confusing? |
|
|
HPA Illuminator
State War Academy Caldari State
28
|
Posted - 2016.04.01 10:58:50 -
[14] - Quote
Circumstantial Evidence wrote:Sample 100000440 has been defined as 100% Vesicles, and I regret not making this choice, in addition to Rods and Rings. I picked R&R because I saw a few rod-like structures... am I completely wrong, or could this sample contain both R&R and Vesicles? Must complete ring structures be present, in samples with R&R? I've marked the BG views with arrows. Area 1 BG -- RGB || Area 2 BG -- RGBSample 100000264 has been defined as 100% Cytoplasm, but I wonder what the structures that seem to outline some parts of the cells might be, marked with arrows in my Green-only screenshot. I thought they might represent a plasma membrane. 100000264 Green -- 100000264 RGB
The first image was a really tricky sample, and I agree it kinda looks like R&R, but not really. Some type of vesicles have this elongated pattern that you can see here. We'll try to remove it from the training set.
For the second image, I see what you mean, but I don't think it's plasma membrane. The cells you look at is a type of brain cell (called U-251 MG), which is known for having what (at least we call) "membrane blebbing", which is that kind of thick edge at the membrane. If you toggle red/green on off you'd likely see a good overlap of the red/green, although the red might be weaker there. |
|
|
HPA Illuminator
The Human Protein Atlas
29
|
Posted - 2016.04.12 14:51:18 -
[15] - Quote
Thanks for reporting, will look into it! |
|
|
HPA Illuminator
The Human Protein Atlas
29
|
Posted - 2016.04.29 13:32:13 -
[16] - Quote
Raven Dallacort wrote:Could we take a look at this control sample as well please. http://imgur.com/vqg6EmT I believe both should be the correct answer.
There's an ongoing discussion in the group how to deal with nuclear speckles vs. nucleoplasm. We're leaning towards that when nuclear speckles are this dominant, one should not choose nucleus/nucleoplasm as well. |
|
|
HPA Illuminator
The Human Protein Atlas
30
|
Posted - 2016.05.02 11:25:43 -
[17] - Quote
Taira Arois wrote:It seems that control sample 100455950 has been mislabled. The correct answer was Nuclear bodies (few) but there are more than 5 spots in every cell. This probably needs to be changed to Nuclear bodies (many).
Please report in to this reddit thread, where we try collect everything:
https://www.reddit.com/r/Eve/comments/4gtoyg/project_discovery_collecting_incorrect_control/ |
|
|
|
|