Wow, another really unique puzzle from tevemadar! I can't wait to get started.
Comment has been collapsed.
25 identified so far. I'm having some fun along the way. :)
Comment has been collapsed.
Bump for fully solved! I'm a visual thinker so I really enjoyed this puzzle.
Comment has been collapsed.
No, the segmentation was made by my girlfriend, and even she has not worked with the rat "himself" (the rat was a male).
I coded the viewers, and the converter for creating meshes. I am a software engineer.
(But yes, we work in research, and there are people around us who do actual rat experiments)
Comment has been collapsed.
Yay! Bump for fully solved!
I have to admit that all this thing is a bit scary at first, but once you are in, it is pretty enjoyable :)
Thanks for your different and interesting puzzles Tevemadar :3 (You have to know that I have a folder with your name with all your puzzles)
Comment has been collapsed.
Hi, the atlas has been created with a tool called ITK Snap. It is open source, and apparently uses a fixed color table by default, link, but the text inside says "some randomized colors". So I think it does not follow any of the colorblind-aware schemes.
The puzzle is about recognizing 3D "objects" by browsing 2D slices of them. The 3D part is toggleable, so neighboring structures can be told apart for sure. However, in the 2D slices there can be borders with unfriendly color combinations. If you move the mouse cursor over the atlas slice, and the label display changes without crossing a "visible" border, then yes, there is a problem.
If you happen to have a nice palette with 81 RGB entries, I can create an other version of the puzzle next week (this week I am on vacation and I could not do the necessary coding on a tablet).
Comment has been collapsed.
I am back in action.
Some things about 's'
out of the 12, the first one and the last one are used in the puzzle (this may be a hint;
their names happen to be 'septal region' and 'supraoptic decussation', but that is not really a new thing - I assume everyone solving the puzzle has a table containing the mentioned structures in alphabetical order)
Comment has been collapsed.
Bump for an interesting and unique puzzle! Don't be scared off by all of the text in the topic description: this is tevemadar's easiest puzzle to date, in my opinion. if you can read a map, you can complete this puzzle.
And check out the fabulous prizes shown in the random screenshot! It pains me to say that because I really, really want to win Gauntlet. :)
Comment has been collapsed.
12? I only found 10. Eight are given in the clues above (including the example clue), one more as an example for a secondary puzzle, and the tenth as the solution to that secondary puzzle. What am I missing? (Based on the screenshot, it's probably two Gauntlet giveaways, but I don't see where I might have overlooked the two final clues in the puzzle...)
Comment has been collapsed.
10 Comments - Last post 7 minutes ago by lostsoul67
93 Comments - Last post 15 minutes ago by Glas
33 Comments - Last post 26 minutes ago by Axelflox
1,831 Comments - Last post 35 minutes ago by Axelflox
15 Comments - Last post 1 hour ago by vlbastos
386 Comments - Last post 5 hours ago by adam1224
207 Comments - Last post 7 hours ago by sensualshakti
5 Comments - Last post 23 seconds ago by BlackbeardXIII
10,792 Comments - Last post 1 minute ago by Cruse
693 Comments - Last post 1 minute ago by Cruse
17 Comments - Last post 1 minute ago by zeroxxx
192 Comments - Last post 38 minutes ago by antidaz
2 Comments - Last post 1 hour ago by aquatorrent
58 Comments - Last post 1 hour ago by Lexbya
EDIT 14/07/2015: opening this very page is enough for solving the rat brain part of Recycling puzzle. No evil Flash Player things are needed (just noticed a thread right now)
In the previous puzzle 8 people voted for "I really wait for some puzzle about the anatomy of the rat brain". And I have not voted, especially not 8 times. So everything below is their fault.
Edit: some people say they find the puzzle pretty solvable, and chances are that some of them are not actual neuroscientists. I do not know if they are from the 8 or not.
As you probably all know, the Waxholm Space atlas of the Sprague Dawley rat brain, the first open access atlas for any kind of rat brain is already available for some time (INCF). Besides the raw data, there are online and offline (this one is brand new) viewers for the MRI, DTI volumes and the actual segmentation (practically: color coding for the different brain regions).
There is also a set of meshes for the segmentation, however they are not released yet.
This detail allows creating a
boring and tediousnovel and exciting puzzle. A page contains a viewer with the meshes, just the name of the structures have been replaced by a number and a character. There are things which are preserved however: the colour of the structures (so everything uses the same colour in all 3 viewers), and their alphabetical order. Like "alveus of the hippocampus" is 01 and "ventricular system" is 80 (this one is not an actual brain structure anyway, but a set of cavities filled with cerebrospinal fluid). The middle one happens to be "inner ear", number 40 (this is not brain at all, just the canals were nicely visible on the MRI). So the numbers are simply preserving the alphabetical order, and the character next to them is used for building 5-letter codes.For example if someone wrote here "medial lemniscus decussation", "habenular commissure", "commissural stria terminalis", "nucleus of the stria medullaris", "ascending fibers of the facial nerve", the solution would be 45, 35, 12, 49, 05, and the letters next to them could form a giveaway code. These are some of the smallest structures in this segmentation, rather impractical to find anyway.
Finding the largest 10 structures should not be painful though: "neocortex", "molecular layer of the cerebellum", "granule cell level of the cerebellum", "brainstem", "corpus callosum and associated subcortical white matter", "spinal trigeminal tract", "basal forebrain region", "striatum", "olfactory bulb", "glomerular layer of the olfactory bulb"
Here are some less random structures, all of them are visible in the initial slice: "perirhinal area 36", "perirhinal area 35", "fasciculus retroflexus", "dentate gyrus", "lateral entorhinal cortex", "cornu ammonis 2", "mammillothalamic tract", "substantia nigra", "descending corticofugal pathways", "commissure of the superior colliculus"
Some from slicing the brain horizontally (so the initial slice has to be rotated by 90 degrees): "fornix", "presubiculum", "fimbria of the hippocampus", "septal region", "genu of the facial nerve", "bed nucleus of the stria terminalis", "stria terminalis", "middle cerebellar peduncle", "optic tract and optic chiasm", "anterior commissure, posterior part"
Then something from the front: "optic nerve" (its continuation has been used already, then comes) "supraoptic decussation", the middle: "thalamus", and from the back: "central canal", "periventricular gray"
The GA-s are running for a pretty long time. Even the ones saying "1 week left" in my profile now, were saying "2 weeks left" yesterday, so 13 days are left for even those.
Have fun, it is not that terrible.
Users guide:
The slicing tools start in MRI view, the droplist in the top-left corner allows switching to DTI and segmentation. In segmentation view the structure under the mouse cursor is going to be displayed at the bottom of the window. This applies to the main view, which shows a rectangular slice of the brain at arbitrary postion and orientation (and size). The other three views show an overview of the entire brain in perpendicular planes, and the rectangle representing the main view is drawn over them.
All views support dragging: the main view moves in its current plane, the other three either rotates the main view in their plane (drag almost anywhere), or moves the main view in their plane (drag the dot in the middle, becomes yellow).
The online viewer requires Flash and it is slow.
The offline viewer requires 3GB of memory, and in case of 32-bit Windows, the 4GT option has to be enabled.
The mesh viewer is even simpler to use, dragging results in rotation, and the structures can be switched on(checkmark)/off(x)/transparent(empty) individually (or all at once with "Set All"). Transparency is set by the first numerical stepper (starts at 50%). There is support for cutting too, so it is possible to set up similar views in both kind of viewers.
The mesh viewer requires Flash. It also requires hardware support for Shader Model 2 (DirectX9), which should not be a real issue. If it is (software rendering with super slow operation, cutting results in hollow view), Chrome may still help (it uses OpenGL, even on Windows). Funny thing is that debug Windows builds (like the Windows 10 preview) support a full software implementation of DirectX, resulting in extremely slow, but visually perfect operation on whatever old hardware.
Hint0 is just a summary to make things less scary: more than half of the structures (43 out of 80) are mentioned above. 8 of them are fully specified.
Hint1: if I were going to solve this puzzle, I would create an Excel (or whatever) sheet with 80 numbered lines, put the fully specified 8 structures into their position, order the rest of the named structures alphabetically in a different column, and roughly position them between the already known ones. Then start experimenting with the viewers.
Obvious enabler hint: so far every single example in my puzzles was an actual giveaway.
Solution
The mesh generator tool along with a pack of meshes have been released on INCF. Online viewer of the set (with real names) is also available here
OHJJe80hGbtfY15xRg3wCPOFr3PySzmd2jZp9hMd is the sequence of all characters encoded above.
After finding any 3 of the giveaways (not counting the OHJJe example here), one could apply the method described in the example, and enter an Alien: Isolation giveaway.
Random screenshot below:
Comment has been collapsed.