Hi Mike! Answers below..
So the "virtual microscope" works exactly like a "real" one - when you move the slider it "zooms in" and "zooms out" focusing on deeper & shallower layers of the tissue. That means that when you move the slider up you are seeing deeper layers of the brain vasculature, and as you move the slider back, you are going back to the shallower layers.
The reason we can only view the this way is, because the vessels in the brain make up a 3D network & the two-photon microscope used in the Schaffer-Nishimura Lab scans it layer by layer. E.g., here is a segment of this network represented in a 3D model:
Except in Stall Catchers we can only look at one later at a time. However, by looking at the consecutive layers with the "virtual microscope" (by "playing" the "vessel movie") we can get a sense of what the vessel looks like, what is it's 3D shape, how is it crossing the plane of view, and of course - what's going on inside it.
The Facebook video shows exactly what you pointed out - at any one time we are looking at a cross-section of the tissue, and depending on the 3D shape of the vessel & how it crosses the plane of view, it may look very different.
E.g. here's a vessel seen in it's entire length in one particular layer, since it's position in the tissue matches the plane of view:
But others may be vertical to the plane of view or cross it in several points, as in the Facebook video. Here's another example of a differently shaped vessel & how we see it in the movie: https://www.facebook.com/eyesonalz/videos/968760833221928/
In any case, during the duration of the movie, you should be able to follow the entire vessel within the outline as different parts (layers) of it come in & out of view, and focus on the cross-sections that you see at any one time to determine if there are any stalls. Sometimes that's easier than other times!
That is true only for the calibration movies (the ones that give you an immediate answer & a high score), as those are the only ones previously seen by experts. The rest is totally up to you guys! (See below why we can be confident in crowd answers.)
The red dots on real movies are those places there by other catchers, which ideally creates a learning opportunity to see how other have interpreted the movie. Albeit that means that there can be mistakes too! (Again, see below for explanation why that doesn't disturb the overall research.)
@pietro would be the absolute best person to answer this, as he is the one conjuring the crowd science that makes Stall Catchers work Hopefully, he will chip in too, but in short: we collect multiple answers per vessel to make sure that the "crowd answer" is correct. We have tested this in our initial validation study last year, and it turned out that we need to collect 15-20 answer to be sure that the crowd answer is as accurate as that of experts (you crowd are GOOD at this!). Quick infographic explains this (click on it for full size):
In other words, this makes sure that we can be >95% confident (and that's enough for the purpose of the research) that the collective answers of the crowd are correct. The experts will only ever look at the questionable movies, i.e. the one's where no clear "stalled" or "flowing" answer could be determined by the crowd.
However, we have some exciting developments in this front too! @pietro has been developing new dynamic consensus methods, which make sure that we take each catcher's sensitivity into account. Pietro has just presented these findings in a talk at the citizen science conference we all went to, and it is definitely worth checking out here: https://www.youtube.com/watch?v=8IdojGgzBk4
So, exactly as you pointed out - in theory we could have 10 answers by less experienced users for one movie and thus get a less reliable end result, but then again, we could have 3 answers by our super users for one movie and be extremely confident in their collective answer. The new methods for determining crowd answers will take that into account, measure each catcher's sensitivity and stop collecting answers for any movie as soon as there's enough expert-like answers for it. This way we can reduce the average number of annotations needed per vessel to 4-7, which - obviously - makes best use of YOUR time, makes sure everyone's contributions matter & allow us to analyse data at even greater speeds!
Please let us know if you have any more concerns about this - more on the new methods to appear on the blog soon!
Good point, although both blood clot & stall can eventually loosen up & migrate, in principle. So could a piece of fruit if it is soft enough, for example! But the bottom line here is that something solid (a blood cell) is causing the "straw" (the vessel) to become clogged for an certain amount of time
That's what YOU are helping us find out! read on ...
This question would be best answered by @nn62, Chris_Schaffer & the rest of the S-N Lab, but in principle the questions you ask is exactly what we are trying to answer with this research! In other words, the researchers at S-N Lab think there's a link between stalls & Alzheimer's disease, and that by preventing or reversing stalls you could prevent Alzheimer's or even reduce existing symptoms - in fact, much of this has already been demonstrated in the lab. Many question marks remain, however, like the exact nature of this link & what molecular pathways we could tackle to find a treatment.
For example, the current dataset is looking at the spatial relationship between stalls & amyloid plaques - if they occur closer together than flowing vessels, that could indicate that one is driving the development of the other, and by targeting & disrupting this destructive feedback loop we could prevent stalls & amyloid forming in the first place, or reduce the symptoms for existing Alzheimer's patients. As you may know, amyloid plaques are thought to be one of the most likely culprits behind Alzheimer's disease, as they are neurotoxic & damage the nervous system.
Here's a little more science in the form of infographics we made previously, which explain why the researchers are interested in stalls & how they could be linked to Alzheimer's:
What is amyloid beta protein:
How amyloid plaques are formed:
How amyloid plaques could be linked to stalls (and stalls to amyloid plaques):
Also, you may find this article of mine interesting - it explains the science behind Stall Catchers: http://blogs.discovermagazine.com/citizen-science-salon/2016/04/22/science-behind-wecurealz-citizen-science-alzheimers/
Haha, don't worry about it - and keep them coming !!!