The Data Sets - Neocortical Layer 6 Axons
Folder Contents:
This is 1 of the 5 datasets used for the DIADEM Grand Challenge Qualifiers Round.
This dataset consists of 34 separate mouse neocortical layer 6 axons all contained within the same 6 image stacks. Each image stack represents one tile for a montage that contains all 34 separate axonal fibers. Thus, all image stacks are to be used for both training and scored projections. Manual reconstructions are provided for 14 of these 34 axons, which are to be used for training (see 'General README'). Contestants will trace the other 20 projections which will be officially scored for the Qualifiers Round.
Experimental Procedures:
Dataset owner: V De Paola, MRC Clinical Sciences Center, Imperial College London
Species: Mouse
Strain: C57/B6, Thy1 transgenic mouse
Nervous System Region: Neocortical layer 6 axons
Fiber type: Axons
Labeling Method: Membrane-targeted GFP in transgenic mouse Line 15
Image Acquisition Method: 2-photon lasers scanning microscopy in vivo
Tracing Method: Neurolucida (Colchester, VT)
Objective Lens: 40x water
See De Paola et al., 2006, Neuron Vol 49(6) for general information related to the data
Z distance between successive images within an image stack: 3.40 pixels
Download Instructions:
1. Download and extract 'Neocortical Layer 6 Axons v2.rar'.
2. Folder will contain 2 subfolders: 'Image Stacks' and 'Manual
Reconstructions'.
3. Unlike other datasets, image stack numbers do not correspond
to reconstruction numbers, since all reconstructions use the same
image stacks (though a given image stack may or may not contain
portions of each of the 34 neocortical axonal fibers).
4. 'Image Stacks' folder will contain 6 subfolders ('01' though
'06') each containing 1 of the 6 image stacks.
5. 'Manual Reconstructions' folder will contain 14 SWC files:
'NC_1' through 'NC_14'.
6. Each SWC file is a manually-traced digital reconstruction
traced in 3 dimensions (X,Y, and Z).
7. Unlike other datasets, SWC files will only automatically align
with image stack '06', which is spatially the top leftmost image
stack tile in XY. Instructions are provided under 'Individual
Stack Information' below for how to use translation values for
proper alignment.
8. Z coordinates must also be translated to align properly,
except for aligning with image stack '06' (also see 'Individual
Stack Information' below).
9. Translation coordinate numbers correspond to the position of
the top left corner for each image stack.
10. Image stack tiles may overlap with each other. Overlapping
structures should only be traced once.
11. Image stacks may contain overlapping projections that should
not be traced.
12. Starting coordinates are given below that clearly mark the
root of the correct axonal projections to reconstruct.
13. Assuming that contestants reconstruct from each image stack
separately, finished reconstructions that cross multiple image
stacks must be connected into single trees as is done for the
provided manual reconstructions.
14. Alternatively, software such as VIAS
(http://mssm.edu/cnic/tools-vias.html)
can be used to tile and
integrate stacks into unified image stacks. If this method is
used, be sure to align image stacks in Z as well as in X and Y.
If there is any discrepancy in relative stack positions compared
to translation values provided below, it is advised to still use
the translation values for alignment since they were used for the
manual reconstructions. Overlapping structures found in
neighboring image stacks can be used to aid in alignment.
15. Special note for this dataset: External branches (i.e. branches that end in a termination) with path lengths of less than 17 pixels will be ignored in the scoring process. Thus, they can be traced or ignored by the contestant without affecting the score, provided that the traced branches remain less than 17 pixels in length.
Manual Reconstruction NC_1 (light blue) is superimposed and offset (for visualization purposes) over image stack NC_6. Arrow (red) points to the root of the reconstruction.
Individual Stack Information (all image stacks are 512x512 pixels in XY): The values below describe how far to move image stacks when they load at (X,Y,Z) = (0,0,0) at the upper left corner of the image stack. If using Neuromantic, which does not have an option to translate image stacks, the reconstructions must instead be moved by the negative of each image stack's X,Y,Z values below to align properly with that image stack.
01- Number of images: 60; Translation (X,Y,Z): (73,507,-5)
02- Number of images: 33; Translation (X,Y,Z): (526,484,11)
03- Number of images: 44; Translation (X,Y,Z): (952,462,-21)
04- Number of images: 51; Translation (X,Y,Z): (924,3,-19)
05- Number of images: 50; Translation (X,Y,Z): (468,-14,-1)
06- Number of images: 46; Translation (X,Y,Z): (0,0,0)
Manual Reconstruction Starting Coordinates (X and Y in pixels; Z in image sequence number, where top image Z = 0):
NC_1 (X,Y,Z): (1,406,44); Training tree; Manual reconstruction provided
NC_2 (X,Y,Z): (2,143,40); Training tree; Manual reconstruction provided
NC_3 (X,Y,Z): (77,740,20); Training tree; Manual reconstruction provided
NC_4 (X,Y,Z): (77,691,51); Training tree; Manual reconstruction provided
NC_5 (X,Y,Z): (201,1018,48); Training tree; Manual reconstruction provided
NC_6 (X,Y,Z): (560,9,-1); Training tree; Manual reconstruction provided
NC_7 (X,Y,Z): (453,1018,19); Training tree; Manual reconstruction provided
NC_8 (X,Y,Z): (560,230,40); Training tree; Manual reconstruction provided
NC_9 (X,Y,Z): (1,466,25); Training tree; Manual reconstruction provided
NC_10 (X,Y,Z): (109,609,-19); Training tree; Manual reconstruction provided
NC_11 (X,Y,Z): (449,705,54); Training tree; Manual reconstruction provided
NC_12 (X,Y,Z): (365,753,-5); Training tree; Manual reconstruction provided
NC_13 (X,Y,Z): (330,710,-5); Training tree; Manual reconstruction provided
NC_14 (X,Y,Z): (140,1018,39); Training tree; Manual reconstruction provided
NC_15 (X,Y,Z): (435,766,40); Scored tree; Manual reconstruction not provided
NC_16 (X,Y,Z): (520,607,24); Scored tree; Manual reconstruction not provided
NC_17 (X,Y,Z): (318,509,10); Scored tree; Manual reconstruction not provided
NC_18 (X,Y,Z): (1418,971,-9); Scored tree; Manual reconstruction not provided
NC_19 (X,Y,Z): (1114,913,10); Scored tree; Manual reconstruction not provided
NC_20 (X,Y,Z): (457,252,45); Scored tree; Manual reconstruction not provided
NC_21 (X,Y,Z): (444,691,-8); Scored tree; Manual reconstruction not provided
NC_22 (X,Y,Z): (492,279,-1); Scored tree; Manual reconstruction not provided
NC_23 (X,Y,Z): (522,264,-1); Scored tree; Manual reconstruction not provided
NC_24 (X,Y,Z): (947,-13,16); Scored tree; Manual reconstruction not provided
NC_25 (X,Y,Z): (569,329,-1); Scored tree; Manual reconstruction not provided
NC_26 (X,Y,Z): (1458,462,7); Scored tree; Manual reconstruction not provided
NC_27 (X,Y,Z): (1110,4,-1); Scored tree; Manual reconstruction not provided
NC_28 (X,Y,Z): (645,672,47); Scored tree; Manual reconstruction not provided
NC_29 (X,Y,Z): (1250,420,16); Scored tree; Manual reconstruction not provided
NC_30 (X,Y,Z): (939,176,14); Scored tree; Manual reconstruction not provided
NC_31 (X,Y,Z): (1236,65,22); Scored tree; Manual reconstruction not provided
NC_32 (X,Y,Z): (978,3,23); Scored tree; Manual reconstruction not provided
NC_33 (X,Y,Z): (1236,4,-5); Scored tree; Manual reconstruction not provided
NC_34 (X,Y,Z): (994,882,11); Scored tree; Manual reconstruction not provided